{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.870583564906709,
  "eval_steps": 3000,
  "global_step": 39000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.691998863220215,
      "epoch": 0.0004962286621675268,
      "grad_norm": 13.125,
      "learning_rate": 2e-06,
      "loss": 10.8423,
      "mean_token_accuracy": 0.0,
      "num_tokens": 11448.0,
      "step": 5
    },
    {
      "entropy": 10.691888427734375,
      "epoch": 0.0009924573243350536,
      "grad_norm": 12.5625,
      "learning_rate": 4.5e-06,
      "loss": 10.7728,
      "mean_token_accuracy": 0.0,
      "num_tokens": 22646.0,
      "step": 10
    },
    {
      "entropy": 10.691094398498535,
      "epoch": 0.0014886859865025804,
      "grad_norm": 9.0,
      "learning_rate": 7e-06,
      "loss": 10.5544,
      "mean_token_accuracy": 0.021141647751210257,
      "num_tokens": 34011.0,
      "step": 15
    },
    {
      "entropy": 10.681868362426759,
      "epoch": 0.001984914648670107,
      "grad_norm": 5.90625,
      "learning_rate": 9.5e-06,
      "loss": 10.2252,
      "mean_token_accuracy": 0.039476046711206435,
      "num_tokens": 44566.0,
      "step": 20
    },
    {
      "entropy": 10.631160545349122,
      "epoch": 0.002481143310837634,
      "grad_norm": 3.9375,
      "learning_rate": 1.2e-05,
      "loss": 10.0163,
      "mean_token_accuracy": 0.03624146841466427,
      "num_tokens": 56274.0,
      "step": 25
    },
    {
      "entropy": 10.599639701843262,
      "epoch": 0.0029773719730051607,
      "grad_norm": 3.171875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 9.8244,
      "mean_token_accuracy": 0.04011724218726158,
      "num_tokens": 66584.0,
      "step": 30
    },
    {
      "entropy": 10.594005584716797,
      "epoch": 0.0034736006351726877,
      "grad_norm": 2.546875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 9.7609,
      "mean_token_accuracy": 0.04112048838287592,
      "num_tokens": 76948.0,
      "step": 35
    },
    {
      "entropy": 10.580470752716064,
      "epoch": 0.003969829297340214,
      "grad_norm": 2.265625,
      "learning_rate": 1.95e-05,
      "loss": 9.7115,
      "mean_token_accuracy": 0.03858393188565969,
      "num_tokens": 87253.0,
      "step": 40
    },
    {
      "entropy": 10.582981967926026,
      "epoch": 0.004466057959507741,
      "grad_norm": 2.453125,
      "learning_rate": 2.2e-05,
      "loss": 9.6634,
      "mean_token_accuracy": 0.0404777305200696,
      "num_tokens": 98273.0,
      "step": 45
    },
    {
      "entropy": 10.553972816467285,
      "epoch": 0.004962286621675268,
      "grad_norm": 2.453125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.5944,
      "mean_token_accuracy": 0.04187291283160448,
      "num_tokens": 107788.0,
      "step": 50
    },
    {
      "entropy": 10.549640369415282,
      "epoch": 0.0054585152838427945,
      "grad_norm": 2.3125,
      "learning_rate": 2.7e-05,
      "loss": 9.5959,
      "mean_token_accuracy": 0.03857453148812055,
      "num_tokens": 118632.0,
      "step": 55
    },
    {
      "entropy": 10.56945514678955,
      "epoch": 0.0059547439460103215,
      "grad_norm": 2.390625,
      "learning_rate": 2.95e-05,
      "loss": 9.4567,
      "mean_token_accuracy": 0.04739701822400093,
      "num_tokens": 128649.0,
      "step": 60
    },
    {
      "entropy": 10.489003658294678,
      "epoch": 0.0064509726081778485,
      "grad_norm": 2.109375,
      "learning_rate": 3.2e-05,
      "loss": 9.4062,
      "mean_token_accuracy": 0.0514289241284132,
      "num_tokens": 138958.0,
      "step": 65
    },
    {
      "entropy": 10.402282333374023,
      "epoch": 0.0069472012703453755,
      "grad_norm": 2.109375,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.3466,
      "mean_token_accuracy": 0.050636925548315045,
      "num_tokens": 151445.0,
      "step": 70
    },
    {
      "entropy": 10.450450229644776,
      "epoch": 0.007443429932512902,
      "grad_norm": 2.0,
      "learning_rate": 3.7e-05,
      "loss": 9.2601,
      "mean_token_accuracy": 0.0536893866956234,
      "num_tokens": 162239.0,
      "step": 75
    },
    {
      "entropy": 10.420755386352539,
      "epoch": 0.007939658594680429,
      "grad_norm": 2.15625,
      "learning_rate": 3.95e-05,
      "loss": 9.2029,
      "mean_token_accuracy": 0.047282115742564204,
      "num_tokens": 173261.0,
      "step": 80
    },
    {
      "entropy": 10.44428882598877,
      "epoch": 0.008435887256847956,
      "grad_norm": 2.1875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.0727,
      "mean_token_accuracy": 0.050180401653051376,
      "num_tokens": 183857.0,
      "step": 85
    },
    {
      "entropy": 10.376095581054688,
      "epoch": 0.008932115919015483,
      "grad_norm": 2.0625,
      "learning_rate": 4.45e-05,
      "loss": 9.0912,
      "mean_token_accuracy": 0.05270495116710663,
      "num_tokens": 196635.0,
      "step": 90
    },
    {
      "entropy": 10.319845962524415,
      "epoch": 0.00942834458118301,
      "grad_norm": 2.015625,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.901,
      "mean_token_accuracy": 0.05795478150248527,
      "num_tokens": 208700.0,
      "step": 95
    },
    {
      "entropy": 10.2221510887146,
      "epoch": 0.009924573243350537,
      "grad_norm": 2.171875,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.7406,
      "mean_token_accuracy": 0.06187366805970669,
      "num_tokens": 218915.0,
      "step": 100
    },
    {
      "entropy": 10.15353012084961,
      "epoch": 0.010420801905518062,
      "grad_norm": 2.203125,
      "learning_rate": 5.2e-05,
      "loss": 8.6056,
      "mean_token_accuracy": 0.06669228337705135,
      "num_tokens": 230211.0,
      "step": 105
    },
    {
      "entropy": 10.054761028289795,
      "epoch": 0.010917030567685589,
      "grad_norm": 1.9296875,
      "learning_rate": 5.45e-05,
      "loss": 8.5184,
      "mean_token_accuracy": 0.06720395609736443,
      "num_tokens": 241174.0,
      "step": 110
    },
    {
      "entropy": 10.002773189544678,
      "epoch": 0.011413259229853116,
      "grad_norm": 2.40625,
      "learning_rate": 5.7e-05,
      "loss": 8.4513,
      "mean_token_accuracy": 0.06137055568397045,
      "num_tokens": 251708.0,
      "step": 115
    },
    {
      "entropy": 9.951133441925048,
      "epoch": 0.011909487892020643,
      "grad_norm": 1.5703125,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.5122,
      "mean_token_accuracy": 0.05963393449783325,
      "num_tokens": 263287.0,
      "step": 120
    },
    {
      "entropy": 9.863128280639648,
      "epoch": 0.01240571655418817,
      "grad_norm": 1.7421875,
      "learning_rate": 6.2e-05,
      "loss": 8.3063,
      "mean_token_accuracy": 0.05798467844724655,
      "num_tokens": 275179.0,
      "step": 125
    },
    {
      "entropy": 9.68651876449585,
      "epoch": 0.012901945216355697,
      "grad_norm": 1.625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.08,
      "mean_token_accuracy": 0.06860841549932957,
      "num_tokens": 287223.0,
      "step": 130
    },
    {
      "entropy": 9.503121948242187,
      "epoch": 0.013398173878523224,
      "grad_norm": 1.875,
      "learning_rate": 6.7e-05,
      "loss": 7.9829,
      "mean_token_accuracy": 0.06533094197511673,
      "num_tokens": 297868.0,
      "step": 135
    },
    {
      "entropy": 9.286877250671386,
      "epoch": 0.013894402540690751,
      "grad_norm": 1.2109375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.9209,
      "mean_token_accuracy": 0.06693189255893231,
      "num_tokens": 309702.0,
      "step": 140
    },
    {
      "entropy": 9.17711067199707,
      "epoch": 0.014390631202858276,
      "grad_norm": 1.2734375,
      "learning_rate": 7.2e-05,
      "loss": 7.9296,
      "mean_token_accuracy": 0.06745850630104541,
      "num_tokens": 321613.0,
      "step": 145
    },
    {
      "entropy": 8.981420612335205,
      "epoch": 0.014886859865025803,
      "grad_norm": 1.3125,
      "learning_rate": 7.45e-05,
      "loss": 7.7664,
      "mean_token_accuracy": 0.06896167993545532,
      "num_tokens": 332668.0,
      "step": 150
    },
    {
      "entropy": 8.791222381591798,
      "epoch": 0.01538308852719333,
      "grad_norm": 1.21875,
      "learning_rate": 7.7e-05,
      "loss": 7.7693,
      "mean_token_accuracy": 0.06417617723345756,
      "num_tokens": 342579.0,
      "step": 155
    },
    {
      "entropy": 8.651444339752198,
      "epoch": 0.015879317189360857,
      "grad_norm": 1.6015625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.6161,
      "mean_token_accuracy": 0.06922528184950352,
      "num_tokens": 352271.0,
      "step": 160
    },
    {
      "entropy": 8.477163314819336,
      "epoch": 0.016375545851528384,
      "grad_norm": 1.34375,
      "learning_rate": 8.2e-05,
      "loss": 7.6478,
      "mean_token_accuracy": 0.0777987513691187,
      "num_tokens": 364033.0,
      "step": 165
    },
    {
      "entropy": 8.385774326324462,
      "epoch": 0.01687177451369591,
      "grad_norm": 1.59375,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.5986,
      "mean_token_accuracy": 0.06512850522994995,
      "num_tokens": 373788.0,
      "step": 170
    },
    {
      "entropy": 8.3142596244812,
      "epoch": 0.01736800317586344,
      "grad_norm": 1.34375,
      "learning_rate": 8.7e-05,
      "loss": 7.5733,
      "mean_token_accuracy": 0.06904393061995506,
      "num_tokens": 384883.0,
      "step": 175
    },
    {
      "entropy": 8.242665863037109,
      "epoch": 0.017864231838030965,
      "grad_norm": 1.5859375,
      "learning_rate": 8.95e-05,
      "loss": 7.4685,
      "mean_token_accuracy": 0.07145911902189254,
      "num_tokens": 396044.0,
      "step": 180
    },
    {
      "entropy": 8.165802764892579,
      "epoch": 0.018360460500198492,
      "grad_norm": 1.4453125,
      "learning_rate": 9.2e-05,
      "loss": 7.5903,
      "mean_token_accuracy": 0.07685633525252342,
      "num_tokens": 406995.0,
      "step": 185
    },
    {
      "entropy": 8.069672966003418,
      "epoch": 0.01885668916236602,
      "grad_norm": 1.2734375,
      "learning_rate": 9.45e-05,
      "loss": 7.4217,
      "mean_token_accuracy": 0.07526631690561772,
      "num_tokens": 417238.0,
      "step": 190
    },
    {
      "entropy": 8.172780990600586,
      "epoch": 0.019352917824533546,
      "grad_norm": 1.859375,
      "learning_rate": 9.7e-05,
      "loss": 7.4965,
      "mean_token_accuracy": 0.07288087829947472,
      "num_tokens": 427263.0,
      "step": 195
    },
    {
      "entropy": 8.033447551727296,
      "epoch": 0.019849146486701073,
      "grad_norm": 1.2734375,
      "learning_rate": 9.95e-05,
      "loss": 7.3851,
      "mean_token_accuracy": 0.07536251321434975,
      "num_tokens": 438357.0,
      "step": 200
    },
    {
      "entropy": 7.94714241027832,
      "epoch": 0.0203453751488686,
      "grad_norm": 1.296875,
      "learning_rate": 0.000102,
      "loss": 7.4067,
      "mean_token_accuracy": 0.0729083027690649,
      "num_tokens": 448738.0,
      "step": 205
    },
    {
      "entropy": 7.937696313858032,
      "epoch": 0.020841603811036124,
      "grad_norm": 1.390625,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.4298,
      "mean_token_accuracy": 0.0836001418530941,
      "num_tokens": 457873.0,
      "step": 210
    },
    {
      "entropy": 7.968968772888184,
      "epoch": 0.02133783247320365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000107,
      "loss": 7.3665,
      "mean_token_accuracy": 0.07748601958155632,
      "num_tokens": 468640.0,
      "step": 215
    },
    {
      "entropy": 7.902806615829467,
      "epoch": 0.021834061135371178,
      "grad_norm": 1.4375,
      "learning_rate": 0.0001095,
      "loss": 7.3699,
      "mean_token_accuracy": 0.07832009121775627,
      "num_tokens": 479922.0,
      "step": 220
    },
    {
      "entropy": 7.789996480941772,
      "epoch": 0.022330289797538705,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000112,
      "loss": 7.3231,
      "mean_token_accuracy": 0.07841933965682983,
      "num_tokens": 491056.0,
      "step": 225
    },
    {
      "entropy": 7.859673023223877,
      "epoch": 0.022826518459706232,
      "grad_norm": 1.25,
      "learning_rate": 0.0001145,
      "loss": 7.4182,
      "mean_token_accuracy": 0.07705076970160007,
      "num_tokens": 504365.0,
      "step": 230
    },
    {
      "entropy": 7.791559553146362,
      "epoch": 0.02332274712187376,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.2999,
      "mean_token_accuracy": 0.07440029308199883,
      "num_tokens": 514347.0,
      "step": 235
    },
    {
      "entropy": 7.851759672164917,
      "epoch": 0.023818975784041286,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.325,
      "mean_token_accuracy": 0.08229230791330337,
      "num_tokens": 525912.0,
      "step": 240
    },
    {
      "entropy": 7.765729904174805,
      "epoch": 0.024315204446208813,
      "grad_norm": 1.296875,
      "learning_rate": 0.000122,
      "loss": 7.2004,
      "mean_token_accuracy": 0.07634501904249191,
      "num_tokens": 535566.0,
      "step": 245
    },
    {
      "entropy": 7.780554628372192,
      "epoch": 0.02481143310837634,
      "grad_norm": 1.203125,
      "learning_rate": 0.0001245,
      "loss": 7.2501,
      "mean_token_accuracy": 0.08192761614918709,
      "num_tokens": 546474.0,
      "step": 250
    },
    {
      "entropy": 7.747992897033692,
      "epoch": 0.025307661770543867,
      "grad_norm": 1.03125,
      "learning_rate": 0.000127,
      "loss": 7.3489,
      "mean_token_accuracy": 0.07340065240859986,
      "num_tokens": 558132.0,
      "step": 255
    },
    {
      "entropy": 7.8211634159088135,
      "epoch": 0.025803890432711394,
      "grad_norm": 1.375,
      "learning_rate": 0.0001295,
      "loss": 7.3332,
      "mean_token_accuracy": 0.07336415946483613,
      "num_tokens": 568309.0,
      "step": 260
    },
    {
      "entropy": 7.631407022476196,
      "epoch": 0.02630011909487892,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000132,
      "loss": 7.2426,
      "mean_token_accuracy": 0.07597642093896866,
      "num_tokens": 579701.0,
      "step": 265
    },
    {
      "entropy": 7.738198757171631,
      "epoch": 0.026796347757046448,
      "grad_norm": 0.984375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.1884,
      "mean_token_accuracy": 0.08395353108644485,
      "num_tokens": 590832.0,
      "step": 270
    },
    {
      "entropy": 7.638070821762085,
      "epoch": 0.027292576419213975,
      "grad_norm": 1.375,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1208,
      "mean_token_accuracy": 0.08120987564325333,
      "num_tokens": 601094.0,
      "step": 275
    },
    {
      "entropy": 7.569035053253174,
      "epoch": 0.027788805081381502,
      "grad_norm": 1.28125,
      "learning_rate": 0.0001395,
      "loss": 7.1377,
      "mean_token_accuracy": 0.08517721891403199,
      "num_tokens": 612027.0,
      "step": 280
    },
    {
      "entropy": 7.633553075790405,
      "epoch": 0.02828503374354903,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1841,
      "mean_token_accuracy": 0.08428357169032097,
      "num_tokens": 623698.0,
      "step": 285
    },
    {
      "entropy": 7.666120433807373,
      "epoch": 0.028781262405716553,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001445,
      "loss": 7.1991,
      "mean_token_accuracy": 0.08183786422014236,
      "num_tokens": 635631.0,
      "step": 290
    },
    {
      "entropy": 7.63528003692627,
      "epoch": 0.02927749106788408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000147,
      "loss": 7.1819,
      "mean_token_accuracy": 0.08330717980861664,
      "num_tokens": 647663.0,
      "step": 295
    },
    {
      "entropy": 7.453550720214844,
      "epoch": 0.029773719730051607,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0001495,
      "loss": 7.1489,
      "mean_token_accuracy": 0.07976513430476188,
      "num_tokens": 659352.0,
      "step": 300
    },
    {
      "entropy": 7.641481256484985,
      "epoch": 0.030269948392219134,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000152,
      "loss": 7.1201,
      "mean_token_accuracy": 0.08412392735481262,
      "num_tokens": 669379.0,
      "step": 305
    },
    {
      "entropy": 7.542712736129761,
      "epoch": 0.03076617705438666,
      "grad_norm": 1.40625,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.0948,
      "mean_token_accuracy": 0.08485196307301521,
      "num_tokens": 680144.0,
      "step": 310
    },
    {
      "entropy": 7.560588741302491,
      "epoch": 0.03126240571655419,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000157,
      "loss": 7.0416,
      "mean_token_accuracy": 0.0779119621962309,
      "num_tokens": 691165.0,
      "step": 315
    },
    {
      "entropy": 7.518430519104004,
      "epoch": 0.031758634378721715,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001595,
      "loss": 7.1409,
      "mean_token_accuracy": 0.07992475032806397,
      "num_tokens": 703767.0,
      "step": 320
    },
    {
      "entropy": 7.489030599594116,
      "epoch": 0.032254863040889245,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000162,
      "loss": 7.1289,
      "mean_token_accuracy": 0.08132177069783211,
      "num_tokens": 714298.0,
      "step": 325
    },
    {
      "entropy": 7.477584648132324,
      "epoch": 0.03275109170305677,
      "grad_norm": 1.234375,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.1088,
      "mean_token_accuracy": 0.08670444339513779,
      "num_tokens": 724135.0,
      "step": 330
    },
    {
      "entropy": 7.487472200393677,
      "epoch": 0.03324732036522429,
      "grad_norm": 1.09375,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.1704,
      "mean_token_accuracy": 0.07990842834115028,
      "num_tokens": 735050.0,
      "step": 335
    },
    {
      "entropy": 7.561232852935791,
      "epoch": 0.03374354902739182,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.0652,
      "mean_token_accuracy": 0.08222858682274818,
      "num_tokens": 746073.0,
      "step": 340
    },
    {
      "entropy": 7.5026990413665775,
      "epoch": 0.034239777689559346,
      "grad_norm": 1.203125,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.0587,
      "mean_token_accuracy": 0.07908397540450096,
      "num_tokens": 758998.0,
      "step": 345
    },
    {
      "entropy": 7.490911149978638,
      "epoch": 0.03473600635172688,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.9922,
      "mean_token_accuracy": 0.08149008676409722,
      "num_tokens": 771608.0,
      "step": 350
    },
    {
      "entropy": 7.411250829696655,
      "epoch": 0.0352322350138944,
      "grad_norm": 0.96875,
      "learning_rate": 0.000177,
      "loss": 7.0564,
      "mean_token_accuracy": 0.0840062066912651,
      "num_tokens": 782394.0,
      "step": 355
    },
    {
      "entropy": 7.412871408462524,
      "epoch": 0.03572846367606193,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0001795,
      "loss": 6.9639,
      "mean_token_accuracy": 0.08742633014917374,
      "num_tokens": 792389.0,
      "step": 360
    },
    {
      "entropy": 7.482435989379883,
      "epoch": 0.036224692338229454,
      "grad_norm": 1.3125,
      "learning_rate": 0.000182,
      "loss": 7.0042,
      "mean_token_accuracy": 0.08739039152860642,
      "num_tokens": 803219.0,
      "step": 365
    },
    {
      "entropy": 7.400604963302612,
      "epoch": 0.036720921000396985,
      "grad_norm": 1.21875,
      "learning_rate": 0.0001845,
      "loss": 7.0724,
      "mean_token_accuracy": 0.08268633931875229,
      "num_tokens": 814249.0,
      "step": 370
    },
    {
      "entropy": 7.4228143215179445,
      "epoch": 0.03721714966256451,
      "grad_norm": 1.109375,
      "learning_rate": 0.000187,
      "loss": 6.9675,
      "mean_token_accuracy": 0.09332375228404999,
      "num_tokens": 825138.0,
      "step": 375
    },
    {
      "entropy": 7.332874011993408,
      "epoch": 0.03771337832473204,
      "grad_norm": 1.234375,
      "learning_rate": 0.0001895,
      "loss": 7.0155,
      "mean_token_accuracy": 0.08916119858622551,
      "num_tokens": 836024.0,
      "step": 380
    },
    {
      "entropy": 7.422525310516358,
      "epoch": 0.03820960698689956,
      "grad_norm": 1.09375,
      "learning_rate": 0.000192,
      "loss": 7.0106,
      "mean_token_accuracy": 0.08618483543395997,
      "num_tokens": 847765.0,
      "step": 385
    },
    {
      "entropy": 7.309659004211426,
      "epoch": 0.03870583564906709,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001945,
      "loss": 6.9978,
      "mean_token_accuracy": 0.08317906334996224,
      "num_tokens": 859980.0,
      "step": 390
    },
    {
      "entropy": 7.398729848861694,
      "epoch": 0.039202064311234616,
      "grad_norm": 1.171875,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.9852,
      "mean_token_accuracy": 0.08303623273968697,
      "num_tokens": 871514.0,
      "step": 395
    },
    {
      "entropy": 7.358026456832886,
      "epoch": 0.03969829297340215,
      "grad_norm": 1.046875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.9883,
      "mean_token_accuracy": 0.08641744777560234,
      "num_tokens": 881984.0,
      "step": 400
    },
    {
      "entropy": 7.339498233795166,
      "epoch": 0.04019452163556967,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000202,
      "loss": 6.9422,
      "mean_token_accuracy": 0.08860462978482246,
      "num_tokens": 894198.0,
      "step": 405
    },
    {
      "entropy": 7.329129791259765,
      "epoch": 0.0406907502977372,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.9528,
      "mean_token_accuracy": 0.08584670275449753,
      "num_tokens": 905394.0,
      "step": 410
    },
    {
      "entropy": 7.4144409656524655,
      "epoch": 0.041186978959904724,
      "grad_norm": 1.25,
      "learning_rate": 0.000207,
      "loss": 7.0093,
      "mean_token_accuracy": 0.08373580649495124,
      "num_tokens": 917010.0,
      "step": 415
    },
    {
      "entropy": 7.2484229564666744,
      "epoch": 0.04168320762207225,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002095,
      "loss": 6.8328,
      "mean_token_accuracy": 0.09602309390902519,
      "num_tokens": 926761.0,
      "step": 420
    },
    {
      "entropy": 7.169170570373535,
      "epoch": 0.04217943628423978,
      "grad_norm": 1.125,
      "learning_rate": 0.000212,
      "loss": 6.8508,
      "mean_token_accuracy": 0.09352190122008323,
      "num_tokens": 938385.0,
      "step": 425
    },
    {
      "entropy": 7.252145004272461,
      "epoch": 0.0426756649464073,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002145,
      "loss": 6.8835,
      "mean_token_accuracy": 0.0913587599992752,
      "num_tokens": 949709.0,
      "step": 430
    },
    {
      "entropy": 7.300234842300415,
      "epoch": 0.04317189360857483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.9009,
      "mean_token_accuracy": 0.08750406205654145,
      "num_tokens": 960837.0,
      "step": 435
    },
    {
      "entropy": 7.217148399353027,
      "epoch": 0.043668122270742356,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002195,
      "loss": 6.7532,
      "mean_token_accuracy": 0.096002509444952,
      "num_tokens": 971036.0,
      "step": 440
    },
    {
      "entropy": 7.26617751121521,
      "epoch": 0.044164350932909886,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000222,
      "loss": 6.8966,
      "mean_token_accuracy": 0.08577612191438674,
      "num_tokens": 983057.0,
      "step": 445
    },
    {
      "entropy": 7.207610130310059,
      "epoch": 0.04466057959507741,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002245,
      "loss": 6.8685,
      "mean_token_accuracy": 0.09361535310745239,
      "num_tokens": 994303.0,
      "step": 450
    },
    {
      "entropy": 7.1220794200897215,
      "epoch": 0.04515680825724494,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.8559,
      "mean_token_accuracy": 0.09114001989364624,
      "num_tokens": 1005033.0,
      "step": 455
    },
    {
      "entropy": 7.20084171295166,
      "epoch": 0.045653036919412464,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.9188,
      "mean_token_accuracy": 0.09030571505427361,
      "num_tokens": 1015940.0,
      "step": 460
    },
    {
      "entropy": 7.239992237091064,
      "epoch": 0.046149265581579994,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.8753,
      "mean_token_accuracy": 0.09003765881061554,
      "num_tokens": 1027582.0,
      "step": 465
    },
    {
      "entropy": 7.178811550140381,
      "epoch": 0.04664549424374752,
      "grad_norm": 0.953125,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.8353,
      "mean_token_accuracy": 0.08800699189305305,
      "num_tokens": 1040046.0,
      "step": 470
    },
    {
      "entropy": 7.159101486206055,
      "epoch": 0.04714172290591505,
      "grad_norm": 1.609375,
      "learning_rate": 0.000237,
      "loss": 6.9207,
      "mean_token_accuracy": 0.07997897453606129,
      "num_tokens": 1050636.0,
      "step": 475
    },
    {
      "entropy": 7.22628493309021,
      "epoch": 0.04763795156808257,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002395,
      "loss": 6.7957,
      "mean_token_accuracy": 0.09439989402890206,
      "num_tokens": 1060748.0,
      "step": 480
    },
    {
      "entropy": 7.260278987884521,
      "epoch": 0.0481341802302501,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000242,
      "loss": 6.8359,
      "mean_token_accuracy": 0.08930574655532837,
      "num_tokens": 1071940.0,
      "step": 485
    },
    {
      "entropy": 7.081393527984619,
      "epoch": 0.048630408892417626,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002445,
      "loss": 6.8633,
      "mean_token_accuracy": 0.09053891003131867,
      "num_tokens": 1082962.0,
      "step": 490
    },
    {
      "entropy": 7.151085042953492,
      "epoch": 0.04912663755458515,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000247,
      "loss": 6.8503,
      "mean_token_accuracy": 0.093645728379488,
      "num_tokens": 1093082.0,
      "step": 495
    },
    {
      "entropy": 7.2744776725769045,
      "epoch": 0.04962286621675268,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002495,
      "loss": 6.9163,
      "mean_token_accuracy": 0.0890045516192913,
      "num_tokens": 1103528.0,
      "step": 500
    },
    {
      "entropy": 7.072138071060181,
      "epoch": 0.0501190948789202,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000252,
      "loss": 6.7991,
      "mean_token_accuracy": 0.08873311057686806,
      "num_tokens": 1114813.0,
      "step": 505
    },
    {
      "entropy": 7.072871494293213,
      "epoch": 0.050615323541087734,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002545,
      "loss": 6.7897,
      "mean_token_accuracy": 0.08742236867547035,
      "num_tokens": 1125345.0,
      "step": 510
    },
    {
      "entropy": 7.088561058044434,
      "epoch": 0.05111155220325526,
      "grad_norm": 1.125,
      "learning_rate": 0.000257,
      "loss": 6.8045,
      "mean_token_accuracy": 0.09052546843886375,
      "num_tokens": 1134985.0,
      "step": 515
    },
    {
      "entropy": 7.123575067520141,
      "epoch": 0.05160778086542279,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002595,
      "loss": 6.8147,
      "mean_token_accuracy": 0.09349623173475266,
      "num_tokens": 1145796.0,
      "step": 520
    },
    {
      "entropy": 7.224857807159424,
      "epoch": 0.05210400952759031,
      "grad_norm": 1.21875,
      "learning_rate": 0.000262,
      "loss": 6.9443,
      "mean_token_accuracy": 0.08439943566918373,
      "num_tokens": 1156936.0,
      "step": 525
    },
    {
      "entropy": 7.091599845886231,
      "epoch": 0.05260023818975784,
      "grad_norm": 1.125,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.8415,
      "mean_token_accuracy": 0.09559056684374809,
      "num_tokens": 1167526.0,
      "step": 530
    },
    {
      "entropy": 7.065865850448608,
      "epoch": 0.053096466851925365,
      "grad_norm": 1.265625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.8723,
      "mean_token_accuracy": 0.08711705654859543,
      "num_tokens": 1177647.0,
      "step": 535
    },
    {
      "entropy": 7.144536447525025,
      "epoch": 0.053592695514092896,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.8338,
      "mean_token_accuracy": 0.0977127991616726,
      "num_tokens": 1188255.0,
      "step": 540
    },
    {
      "entropy": 6.989999294281006,
      "epoch": 0.05408892417626042,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.7547,
      "mean_token_accuracy": 0.09971059560775757,
      "num_tokens": 1199433.0,
      "step": 545
    },
    {
      "entropy": 7.007773780822754,
      "epoch": 0.05458515283842795,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002745,
      "loss": 6.8445,
      "mean_token_accuracy": 0.09036741033196449,
      "num_tokens": 1210067.0,
      "step": 550
    },
    {
      "entropy": 7.000052261352539,
      "epoch": 0.055081381500595473,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000277,
      "loss": 6.7252,
      "mean_token_accuracy": 0.09621484205126762,
      "num_tokens": 1220639.0,
      "step": 555
    },
    {
      "entropy": 7.0871340274810795,
      "epoch": 0.055577610162763004,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002795,
      "loss": 6.769,
      "mean_token_accuracy": 0.09496552497148514,
      "num_tokens": 1233158.0,
      "step": 560
    },
    {
      "entropy": 6.9964536190032955,
      "epoch": 0.05607383882493053,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.7508,
      "mean_token_accuracy": 0.08838386237621307,
      "num_tokens": 1244552.0,
      "step": 565
    },
    {
      "entropy": 7.103402709960937,
      "epoch": 0.05657006748709806,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002845,
      "loss": 6.8534,
      "mean_token_accuracy": 0.08600240871310234,
      "num_tokens": 1255478.0,
      "step": 570
    },
    {
      "entropy": 6.939698266983032,
      "epoch": 0.05706629614926558,
      "grad_norm": 1.046875,
      "learning_rate": 0.000287,
      "loss": 6.6773,
      "mean_token_accuracy": 0.10140636935830116,
      "num_tokens": 1267507.0,
      "step": 575
    },
    {
      "entropy": 6.916869688034057,
      "epoch": 0.057562524811433105,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002895,
      "loss": 6.695,
      "mean_token_accuracy": 0.08891933038830757,
      "num_tokens": 1279377.0,
      "step": 580
    },
    {
      "entropy": 7.038671827316284,
      "epoch": 0.058058753473600636,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000292,
      "loss": 6.7852,
      "mean_token_accuracy": 0.09897390529513359,
      "num_tokens": 1290671.0,
      "step": 585
    },
    {
      "entropy": 6.851378393173218,
      "epoch": 0.05855498213576816,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002945,
      "loss": 6.7456,
      "mean_token_accuracy": 0.09530175179243087,
      "num_tokens": 1302792.0,
      "step": 590
    },
    {
      "entropy": 7.058105230331421,
      "epoch": 0.05905121079793569,
      "grad_norm": 1.359375,
      "learning_rate": 0.000297,
      "loss": 6.6992,
      "mean_token_accuracy": 0.09145391955971718,
      "num_tokens": 1313315.0,
      "step": 595
    },
    {
      "entropy": 6.8826432704925535,
      "epoch": 0.05954743946010321,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002995,
      "loss": 6.7687,
      "mean_token_accuracy": 0.08865362480282783,
      "num_tokens": 1326358.0,
      "step": 600
    },
    {
      "entropy": 6.999434471130371,
      "epoch": 0.060043668122270744,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000302,
      "loss": 6.6327,
      "mean_token_accuracy": 0.0945927508175373,
      "num_tokens": 1336266.0,
      "step": 605
    },
    {
      "entropy": 6.812005853652954,
      "epoch": 0.06053989678443827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003045,
      "loss": 6.583,
      "mean_token_accuracy": 0.09936992004513741,
      "num_tokens": 1347610.0,
      "step": 610
    },
    {
      "entropy": 6.810373449325562,
      "epoch": 0.0610361254466058,
      "grad_norm": 1.203125,
      "learning_rate": 0.000307,
      "loss": 6.5545,
      "mean_token_accuracy": 0.1030762791633606,
      "num_tokens": 1358317.0,
      "step": 615
    },
    {
      "entropy": 6.903623342514038,
      "epoch": 0.06153235410877332,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003095,
      "loss": 6.6804,
      "mean_token_accuracy": 0.09128398001194,
      "num_tokens": 1370126.0,
      "step": 620
    },
    {
      "entropy": 6.882534837722778,
      "epoch": 0.06202858277094085,
      "grad_norm": 1.09375,
      "learning_rate": 0.000312,
      "loss": 6.6464,
      "mean_token_accuracy": 0.0963453158736229,
      "num_tokens": 1381571.0,
      "step": 625
    },
    {
      "entropy": 6.7741142272949215,
      "epoch": 0.06252481143310838,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003145,
      "loss": 6.6708,
      "mean_token_accuracy": 0.0964869923889637,
      "num_tokens": 1392486.0,
      "step": 630
    },
    {
      "entropy": 6.902181005477905,
      "epoch": 0.0630210400952759,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000317,
      "loss": 6.68,
      "mean_token_accuracy": 0.09968246370553971,
      "num_tokens": 1403086.0,
      "step": 635
    },
    {
      "entropy": 6.83144588470459,
      "epoch": 0.06351726875744343,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003195,
      "loss": 6.6507,
      "mean_token_accuracy": 0.09976455271244049,
      "num_tokens": 1414537.0,
      "step": 640
    },
    {
      "entropy": 6.840048837661743,
      "epoch": 0.06401349741961096,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000322,
      "loss": 6.6542,
      "mean_token_accuracy": 0.09846014603972435,
      "num_tokens": 1426998.0,
      "step": 645
    },
    {
      "entropy": 6.847141170501709,
      "epoch": 0.06450972608177849,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6781,
      "mean_token_accuracy": 0.09744567275047303,
      "num_tokens": 1437667.0,
      "step": 650
    },
    {
      "entropy": 6.8108092784881595,
      "epoch": 0.065005954743946,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.547,
      "mean_token_accuracy": 0.10085381418466569,
      "num_tokens": 1448280.0,
      "step": 655
    },
    {
      "entropy": 6.831343030929565,
      "epoch": 0.06550218340611354,
      "grad_norm": 1.296875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.6163,
      "mean_token_accuracy": 0.09790344089269638,
      "num_tokens": 1458608.0,
      "step": 660
    },
    {
      "entropy": 6.936217546463013,
      "epoch": 0.06599841206828107,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.8059,
      "mean_token_accuracy": 0.08812156543135644,
      "num_tokens": 1470435.0,
      "step": 665
    },
    {
      "entropy": 6.771727180480957,
      "epoch": 0.06649464073044858,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.6608,
      "mean_token_accuracy": 0.09632581174373626,
      "num_tokens": 1481959.0,
      "step": 670
    },
    {
      "entropy": 6.825786304473877,
      "epoch": 0.06699086939261611,
      "grad_norm": 1.078125,
      "learning_rate": 0.000337,
      "loss": 6.6976,
      "mean_token_accuracy": 0.0911058507859707,
      "num_tokens": 1493798.0,
      "step": 675
    },
    {
      "entropy": 6.816014671325684,
      "epoch": 0.06748709805478365,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003395,
      "loss": 6.503,
      "mean_token_accuracy": 0.10050572380423546,
      "num_tokens": 1502969.0,
      "step": 680
    },
    {
      "entropy": 6.787315845489502,
      "epoch": 0.06798332671695118,
      "grad_norm": 1.109375,
      "learning_rate": 0.000342,
      "loss": 6.67,
      "mean_token_accuracy": 0.09576347395777703,
      "num_tokens": 1513825.0,
      "step": 685
    },
    {
      "entropy": 6.755472135543823,
      "epoch": 0.06847955537911869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.5492,
      "mean_token_accuracy": 0.10139472037553787,
      "num_tokens": 1525328.0,
      "step": 690
    },
    {
      "entropy": 6.805799102783203,
      "epoch": 0.06897578404128622,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000347,
      "loss": 6.5664,
      "mean_token_accuracy": 0.10111717209219932,
      "num_tokens": 1537383.0,
      "step": 695
    },
    {
      "entropy": 6.773342514038086,
      "epoch": 0.06947201270345375,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003495,
      "loss": 6.6756,
      "mean_token_accuracy": 0.09535346329212188,
      "num_tokens": 1549671.0,
      "step": 700
    },
    {
      "entropy": 6.707566404342652,
      "epoch": 0.06996824136562128,
      "grad_norm": 1.15625,
      "learning_rate": 0.000352,
      "loss": 6.55,
      "mean_token_accuracy": 0.0958769753575325,
      "num_tokens": 1559974.0,
      "step": 705
    },
    {
      "entropy": 6.82678599357605,
      "epoch": 0.0704644700277888,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003545,
      "loss": 6.6602,
      "mean_token_accuracy": 0.09039726406335831,
      "num_tokens": 1570890.0,
      "step": 710
    },
    {
      "entropy": 6.738782691955566,
      "epoch": 0.07096069868995633,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000357,
      "loss": 6.5641,
      "mean_token_accuracy": 0.1056017056107521,
      "num_tokens": 1581434.0,
      "step": 715
    },
    {
      "entropy": 6.804084634780883,
      "epoch": 0.07145692735212386,
      "grad_norm": 1.125,
      "learning_rate": 0.0003595,
      "loss": 6.6318,
      "mean_token_accuracy": 0.09419183582067489,
      "num_tokens": 1593710.0,
      "step": 720
    },
    {
      "entropy": 6.76332368850708,
      "epoch": 0.07195315601429139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000362,
      "loss": 6.5919,
      "mean_token_accuracy": 0.09819654747843742,
      "num_tokens": 1604635.0,
      "step": 725
    },
    {
      "entropy": 6.889756727218628,
      "epoch": 0.07244938467645891,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003645,
      "loss": 6.6595,
      "mean_token_accuracy": 0.09840102568268776,
      "num_tokens": 1615258.0,
      "step": 730
    },
    {
      "entropy": 6.677707576751709,
      "epoch": 0.07294561333862644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000367,
      "loss": 6.5582,
      "mean_token_accuracy": 0.09775017052888871,
      "num_tokens": 1625942.0,
      "step": 735
    },
    {
      "entropy": 6.687536382675171,
      "epoch": 0.07344184200079397,
      "grad_norm": 1.125,
      "learning_rate": 0.0003695,
      "loss": 6.5459,
      "mean_token_accuracy": 0.10621693730354309,
      "num_tokens": 1636270.0,
      "step": 740
    },
    {
      "entropy": 6.677185583114624,
      "epoch": 0.07393807066296149,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000372,
      "loss": 6.5294,
      "mean_token_accuracy": 0.10231522172689438,
      "num_tokens": 1647029.0,
      "step": 745
    },
    {
      "entropy": 6.756840419769287,
      "epoch": 0.07443429932512902,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003745,
      "loss": 6.7135,
      "mean_token_accuracy": 0.09764664024114608,
      "num_tokens": 1659483.0,
      "step": 750
    },
    {
      "entropy": 6.648179912567139,
      "epoch": 0.07493052798729655,
      "grad_norm": 1.0,
      "learning_rate": 0.000377,
      "loss": 6.5518,
      "mean_token_accuracy": 0.0972537100315094,
      "num_tokens": 1669425.0,
      "step": 755
    },
    {
      "entropy": 6.727893924713134,
      "epoch": 0.07542675664946408,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003795,
      "loss": 6.568,
      "mean_token_accuracy": 0.09414743408560752,
      "num_tokens": 1680426.0,
      "step": 760
    },
    {
      "entropy": 6.665664148330689,
      "epoch": 0.0759229853116316,
      "grad_norm": 0.953125,
      "learning_rate": 0.000382,
      "loss": 6.5744,
      "mean_token_accuracy": 0.0945310615003109,
      "num_tokens": 1692803.0,
      "step": 765
    },
    {
      "entropy": 6.669447755813598,
      "epoch": 0.07641921397379912,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003845,
      "loss": 6.5984,
      "mean_token_accuracy": 0.0979096457362175,
      "num_tokens": 1703919.0,
      "step": 770
    },
    {
      "entropy": 6.73317174911499,
      "epoch": 0.07691544263596665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.5787,
      "mean_token_accuracy": 0.1017878644168377,
      "num_tokens": 1714487.0,
      "step": 775
    },
    {
      "entropy": 6.738369035720825,
      "epoch": 0.07741167129813419,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.6143,
      "mean_token_accuracy": 0.09932834580540657,
      "num_tokens": 1726601.0,
      "step": 780
    },
    {
      "entropy": 6.703241872787475,
      "epoch": 0.0779078999603017,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.5226,
      "mean_token_accuracy": 0.09681319668889046,
      "num_tokens": 1738743.0,
      "step": 785
    },
    {
      "entropy": 6.577711486816407,
      "epoch": 0.07840412862246923,
      "grad_norm": 1.125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.6013,
      "mean_token_accuracy": 0.09930569976568222,
      "num_tokens": 1748286.0,
      "step": 790
    },
    {
      "entropy": 6.648367071151734,
      "epoch": 0.07890035728463676,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.6383,
      "mean_token_accuracy": 0.09991247206926346,
      "num_tokens": 1758917.0,
      "step": 795
    },
    {
      "entropy": 6.686027669906617,
      "epoch": 0.0793965859468043,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003995,
      "loss": 6.4599,
      "mean_token_accuracy": 0.10187165811657906,
      "num_tokens": 1769179.0,
      "step": 800
    },
    {
      "entropy": 6.624785614013672,
      "epoch": 0.07989281460897181,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000402,
      "loss": 6.6059,
      "mean_token_accuracy": 0.09910577759146691,
      "num_tokens": 1782369.0,
      "step": 805
    },
    {
      "entropy": 6.665830755233765,
      "epoch": 0.08038904327113934,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004045,
      "loss": 6.5888,
      "mean_token_accuracy": 0.103836190700531,
      "num_tokens": 1795992.0,
      "step": 810
    },
    {
      "entropy": 6.596944665908813,
      "epoch": 0.08088527193330687,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.491,
      "mean_token_accuracy": 0.10671056434512138,
      "num_tokens": 1806938.0,
      "step": 815
    },
    {
      "entropy": 6.592557191848755,
      "epoch": 0.0813815005954744,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004095,
      "loss": 6.4242,
      "mean_token_accuracy": 0.1038404531776905,
      "num_tokens": 1817663.0,
      "step": 820
    },
    {
      "entropy": 6.645978593826294,
      "epoch": 0.08187772925764192,
      "grad_norm": 1.078125,
      "learning_rate": 0.000412,
      "loss": 6.577,
      "mean_token_accuracy": 0.10006028190255165,
      "num_tokens": 1828254.0,
      "step": 825
    },
    {
      "entropy": 6.615597295761108,
      "epoch": 0.08237395791980945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004145,
      "loss": 6.5017,
      "mean_token_accuracy": 0.10172841027379036,
      "num_tokens": 1839001.0,
      "step": 830
    },
    {
      "entropy": 6.63342056274414,
      "epoch": 0.08287018658197698,
      "grad_norm": 1.03125,
      "learning_rate": 0.000417,
      "loss": 6.6031,
      "mean_token_accuracy": 0.09923635870218277,
      "num_tokens": 1849906.0,
      "step": 835
    },
    {
      "entropy": 6.54316554069519,
      "epoch": 0.0833664152441445,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004195,
      "loss": 6.4851,
      "mean_token_accuracy": 0.09840435534715652,
      "num_tokens": 1860982.0,
      "step": 840
    },
    {
      "entropy": 6.650843954086303,
      "epoch": 0.08386264390631203,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000422,
      "loss": 6.4543,
      "mean_token_accuracy": 0.10228981822729111,
      "num_tokens": 1872177.0,
      "step": 845
    },
    {
      "entropy": 6.506167984008789,
      "epoch": 0.08435887256847956,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004245,
      "loss": 6.4298,
      "mean_token_accuracy": 0.10004795640707016,
      "num_tokens": 1882783.0,
      "step": 850
    },
    {
      "entropy": 6.652998065948486,
      "epoch": 0.08485510123064709,
      "grad_norm": 1.140625,
      "learning_rate": 0.000427,
      "loss": 6.6083,
      "mean_token_accuracy": 0.09435755014419556,
      "num_tokens": 1893706.0,
      "step": 855
    },
    {
      "entropy": 6.550818347930909,
      "epoch": 0.0853513298928146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004295,
      "loss": 6.4819,
      "mean_token_accuracy": 0.10135309919714927,
      "num_tokens": 1903852.0,
      "step": 860
    },
    {
      "entropy": 6.591495037078857,
      "epoch": 0.08584755855498213,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000432,
      "loss": 6.5065,
      "mean_token_accuracy": 0.10633492544293403,
      "num_tokens": 1914157.0,
      "step": 865
    },
    {
      "entropy": 6.572027492523193,
      "epoch": 0.08634378721714966,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004345,
      "loss": 6.4494,
      "mean_token_accuracy": 0.10061108618974686,
      "num_tokens": 1924684.0,
      "step": 870
    },
    {
      "entropy": 6.659038972854614,
      "epoch": 0.0868400158793172,
      "grad_norm": 0.953125,
      "learning_rate": 0.000437,
      "loss": 6.6039,
      "mean_token_accuracy": 0.09810293242335319,
      "num_tokens": 1935657.0,
      "step": 875
    },
    {
      "entropy": 6.516001558303833,
      "epoch": 0.08733624454148471,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004395,
      "loss": 6.4699,
      "mean_token_accuracy": 0.10223818272352218,
      "num_tokens": 1946735.0,
      "step": 880
    },
    {
      "entropy": 6.609535169601441,
      "epoch": 0.08783247320365224,
      "grad_norm": 1.171875,
      "learning_rate": 0.000442,
      "loss": 6.6563,
      "mean_token_accuracy": 0.09723427742719651,
      "num_tokens": 1957871.0,
      "step": 885
    },
    {
      "entropy": 6.502126550674438,
      "epoch": 0.08832870186581977,
      "grad_norm": 1.25,
      "learning_rate": 0.0004445,
      "loss": 6.4901,
      "mean_token_accuracy": 0.10245168954133987,
      "num_tokens": 1969032.0,
      "step": 890
    },
    {
      "entropy": 6.63283109664917,
      "epoch": 0.0888249305279873,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000447,
      "loss": 6.4947,
      "mean_token_accuracy": 0.10377224013209344,
      "num_tokens": 1980212.0,
      "step": 895
    },
    {
      "entropy": 6.4811985969543455,
      "epoch": 0.08932115919015482,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.4302,
      "mean_token_accuracy": 0.10036835372447968,
      "num_tokens": 1990333.0,
      "step": 900
    },
    {
      "entropy": 6.624222469329834,
      "epoch": 0.08981738785232235,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.5093,
      "mean_token_accuracy": 0.10546329617500305,
      "num_tokens": 2001486.0,
      "step": 905
    },
    {
      "entropy": 6.498252010345459,
      "epoch": 0.09031361651448988,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.4836,
      "mean_token_accuracy": 0.10137376263737678,
      "num_tokens": 2012535.0,
      "step": 910
    },
    {
      "entropy": 6.448726797103882,
      "epoch": 0.0908098451766574,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.4479,
      "mean_token_accuracy": 0.10513061881065369,
      "num_tokens": 2024116.0,
      "step": 915
    },
    {
      "entropy": 6.52455415725708,
      "epoch": 0.09130607383882493,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.4937,
      "mean_token_accuracy": 0.10239877849817276,
      "num_tokens": 2036341.0,
      "step": 920
    },
    {
      "entropy": 6.582788896560669,
      "epoch": 0.09180230250099246,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000462,
      "loss": 6.4589,
      "mean_token_accuracy": 0.10736544132232666,
      "num_tokens": 2046768.0,
      "step": 925
    },
    {
      "entropy": 6.458439350128174,
      "epoch": 0.09229853116315999,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004645,
      "loss": 6.434,
      "mean_token_accuracy": 0.10159429535269737,
      "num_tokens": 2057315.0,
      "step": 930
    },
    {
      "entropy": 6.482909297943115,
      "epoch": 0.0927947598253275,
      "grad_norm": 0.953125,
      "learning_rate": 0.000467,
      "loss": 6.4105,
      "mean_token_accuracy": 0.10489368736743927,
      "num_tokens": 2068267.0,
      "step": 935
    },
    {
      "entropy": 6.438783884048462,
      "epoch": 0.09329098848749504,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004695,
      "loss": 6.4727,
      "mean_token_accuracy": 0.1056045152246952,
      "num_tokens": 2078271.0,
      "step": 940
    },
    {
      "entropy": 6.471507453918457,
      "epoch": 0.09378721714966257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000472,
      "loss": 6.4045,
      "mean_token_accuracy": 0.09843646511435508,
      "num_tokens": 2088381.0,
      "step": 945
    },
    {
      "entropy": 6.484247779846191,
      "epoch": 0.0942834458118301,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004745,
      "loss": 6.3704,
      "mean_token_accuracy": 0.10096605494618416,
      "num_tokens": 2098899.0,
      "step": 950
    },
    {
      "entropy": 6.458307695388794,
      "epoch": 0.09477967447399761,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477,
      "loss": 6.5349,
      "mean_token_accuracy": 0.09965546950697898,
      "num_tokens": 2110057.0,
      "step": 955
    },
    {
      "entropy": 6.403871393203735,
      "epoch": 0.09527590313616514,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004795,
      "loss": 6.439,
      "mean_token_accuracy": 0.10709322020411491,
      "num_tokens": 2121758.0,
      "step": 960
    },
    {
      "entropy": 6.51858696937561,
      "epoch": 0.09577213179833267,
      "grad_norm": 1.015625,
      "learning_rate": 0.000482,
      "loss": 6.4436,
      "mean_token_accuracy": 0.10227768048644066,
      "num_tokens": 2131840.0,
      "step": 965
    },
    {
      "entropy": 6.40182580947876,
      "epoch": 0.0962683604605002,
      "grad_norm": 1.0,
      "learning_rate": 0.0004845,
      "loss": 6.4111,
      "mean_token_accuracy": 0.10386593788862228,
      "num_tokens": 2142084.0,
      "step": 970
    },
    {
      "entropy": 6.511473226547241,
      "epoch": 0.09676458912266772,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000487,
      "loss": 6.4699,
      "mean_token_accuracy": 0.10299977660179138,
      "num_tokens": 2152785.0,
      "step": 975
    },
    {
      "entropy": 6.500741481781006,
      "epoch": 0.09726081778483525,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004895,
      "loss": 6.4753,
      "mean_token_accuracy": 0.10271376669406891,
      "num_tokens": 2163774.0,
      "step": 980
    },
    {
      "entropy": 6.387545013427735,
      "epoch": 0.09775704644700278,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000492,
      "loss": 6.3898,
      "mean_token_accuracy": 0.10328807160258294,
      "num_tokens": 2175378.0,
      "step": 985
    },
    {
      "entropy": 6.624388551712036,
      "epoch": 0.0982532751091703,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004945,
      "loss": 6.5951,
      "mean_token_accuracy": 0.09878992438316345,
      "num_tokens": 2186510.0,
      "step": 990
    },
    {
      "entropy": 6.389967775344848,
      "epoch": 0.09874950377133783,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000497,
      "loss": 6.3943,
      "mean_token_accuracy": 0.10771489515900612,
      "num_tokens": 2197482.0,
      "step": 995
    },
    {
      "entropy": 6.380904293060302,
      "epoch": 0.09924573243350536,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995,
      "loss": 6.3707,
      "mean_token_accuracy": 0.10641616210341454,
      "num_tokens": 2208329.0,
      "step": 1000
    },
    {
      "entropy": 6.359027290344239,
      "epoch": 0.09974196109567289,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999999982145551,
      "loss": 6.2907,
      "mean_token_accuracy": 0.10751227289438248,
      "num_tokens": 2218632.0,
      "step": 1005
    },
    {
      "entropy": 6.3495697498321535,
      "epoch": 0.1002381897578404,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999999909611855,
      "loss": 6.4102,
      "mean_token_accuracy": 0.09977167919278145,
      "num_tokens": 2231335.0,
      "step": 1010
    },
    {
      "entropy": 6.492461919784546,
      "epoch": 0.10073441842000794,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999999781283008,
      "loss": 6.3835,
      "mean_token_accuracy": 0.10530572384595871,
      "num_tokens": 2241704.0,
      "step": 1015
    },
    {
      "entropy": 6.349406576156616,
      "epoch": 0.10123064708217547,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999999597159017,
      "loss": 6.4696,
      "mean_token_accuracy": 0.09799574688076973,
      "num_tokens": 2252310.0,
      "step": 1020
    },
    {
      "entropy": 6.513053894042969,
      "epoch": 0.101726875744343,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999999357239882,
      "loss": 6.3749,
      "mean_token_accuracy": 0.10209976211190223,
      "num_tokens": 2262270.0,
      "step": 1025
    },
    {
      "entropy": 6.457979393005371,
      "epoch": 0.10222310440651051,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999999061525612,
      "loss": 6.5111,
      "mean_token_accuracy": 0.10106326639652252,
      "num_tokens": 2273986.0,
      "step": 1030
    },
    {
      "entropy": 6.485824108123779,
      "epoch": 0.10271933306867805,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999998710016214,
      "loss": 6.3925,
      "mean_token_accuracy": 0.10402849987149239,
      "num_tokens": 2283430.0,
      "step": 1035
    },
    {
      "entropy": 6.439836835861206,
      "epoch": 0.10321556173084558,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999998302711696,
      "loss": 6.4099,
      "mean_token_accuracy": 0.1028740145266056,
      "num_tokens": 2294201.0,
      "step": 1040
    },
    {
      "entropy": 6.307459545135498,
      "epoch": 0.1037117903930131,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999997839612069,
      "loss": 6.3057,
      "mean_token_accuracy": 0.10684936791658402,
      "num_tokens": 2303876.0,
      "step": 1045
    },
    {
      "entropy": 6.308732223510742,
      "epoch": 0.10420801905518062,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999997320717342,
      "loss": 6.3598,
      "mean_token_accuracy": 0.10919448658823967,
      "num_tokens": 2313713.0,
      "step": 1050
    },
    {
      "entropy": 6.328538656234741,
      "epoch": 0.10470424771734815,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999996746027531,
      "loss": 6.3731,
      "mean_token_accuracy": 0.10021564215421677,
      "num_tokens": 2324862.0,
      "step": 1055
    },
    {
      "entropy": 6.477358341217041,
      "epoch": 0.10520047637951568,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499999611554265,
      "loss": 6.4335,
      "mean_token_accuracy": 0.10697967633605003,
      "num_tokens": 2334974.0,
      "step": 1060
    },
    {
      "entropy": 6.4058030128479,
      "epoch": 0.10569670504168321,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999995429262712,
      "loss": 6.478,
      "mean_token_accuracy": 0.10283638387918473,
      "num_tokens": 2345945.0,
      "step": 1065
    },
    {
      "entropy": 6.446000862121582,
      "epoch": 0.10619293370385073,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999994687187737,
      "loss": 6.401,
      "mean_token_accuracy": 0.10349917486310005,
      "num_tokens": 2356749.0,
      "step": 1070
    },
    {
      "entropy": 6.440154552459717,
      "epoch": 0.10668916236601826,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999993889317742,
      "loss": 6.4816,
      "mean_token_accuracy": 0.09764729216694831,
      "num_tokens": 2368181.0,
      "step": 1075
    },
    {
      "entropy": 6.341366386413574,
      "epoch": 0.10718539102818579,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999993035652746,
      "loss": 6.3788,
      "mean_token_accuracy": 0.10620132908225059,
      "num_tokens": 2377737.0,
      "step": 1080
    },
    {
      "entropy": 6.3908861637115475,
      "epoch": 0.10768161969035331,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999992126192771,
      "loss": 6.292,
      "mean_token_accuracy": 0.1073671467602253,
      "num_tokens": 2388944.0,
      "step": 1085
    },
    {
      "entropy": 6.353183174133301,
      "epoch": 0.10817784835252084,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499999116093784,
      "loss": 6.3801,
      "mean_token_accuracy": 0.10383658036589623,
      "num_tokens": 2400758.0,
      "step": 1090
    },
    {
      "entropy": 6.32312593460083,
      "epoch": 0.10867407701468837,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999990139887976,
      "loss": 6.375,
      "mean_token_accuracy": 0.1099216289818287,
      "num_tokens": 2411916.0,
      "step": 1095
    },
    {
      "entropy": 6.5799376487731935,
      "epoch": 0.1091703056768559,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999989063043206,
      "loss": 6.3937,
      "mean_token_accuracy": 0.10351028591394425,
      "num_tokens": 2423310.0,
      "step": 1100
    },
    {
      "entropy": 6.210668468475342,
      "epoch": 0.10966653433902342,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999987930403555,
      "loss": 6.331,
      "mean_token_accuracy": 0.10480535998940468,
      "num_tokens": 2432100.0,
      "step": 1105
    },
    {
      "entropy": 6.410353755950927,
      "epoch": 0.11016276300119095,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999986741969051,
      "loss": 6.4061,
      "mean_token_accuracy": 0.10518641620874405,
      "num_tokens": 2443563.0,
      "step": 1110
    },
    {
      "entropy": 6.427771806716919,
      "epoch": 0.11065899166335848,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999985497739724,
      "loss": 6.4161,
      "mean_token_accuracy": 0.10631361752748489,
      "num_tokens": 2455435.0,
      "step": 1115
    },
    {
      "entropy": 6.324798250198365,
      "epoch": 0.11115522032552601,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999984197715606,
      "loss": 6.3474,
      "mean_token_accuracy": 0.10725902542471885,
      "num_tokens": 2466173.0,
      "step": 1120
    },
    {
      "entropy": 6.336659860610962,
      "epoch": 0.11165144898769352,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999982841896728,
      "loss": 6.2889,
      "mean_token_accuracy": 0.10596159473061562,
      "num_tokens": 2476349.0,
      "step": 1125
    },
    {
      "entropy": 6.389477157592774,
      "epoch": 0.11214767764986106,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999981430283123,
      "loss": 6.4296,
      "mean_token_accuracy": 0.11070242971181869,
      "num_tokens": 2487864.0,
      "step": 1130
    },
    {
      "entropy": 6.343891382217407,
      "epoch": 0.11264390631202859,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999979962874827,
      "loss": 6.3521,
      "mean_token_accuracy": 0.1044541135430336,
      "num_tokens": 2499188.0,
      "step": 1135
    },
    {
      "entropy": 6.371397876739502,
      "epoch": 0.11314013497419612,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999978439671877,
      "loss": 6.3489,
      "mean_token_accuracy": 0.10524334833025932,
      "num_tokens": 2509813.0,
      "step": 1140
    },
    {
      "entropy": 6.323387098312378,
      "epoch": 0.11363636363636363,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999976860674309,
      "loss": 6.3801,
      "mean_token_accuracy": 0.10332749933004379,
      "num_tokens": 2520813.0,
      "step": 1145
    },
    {
      "entropy": 6.365607404708863,
      "epoch": 0.11413259229853116,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999975225882164,
      "loss": 6.364,
      "mean_token_accuracy": 0.10385186225175858,
      "num_tokens": 2531802.0,
      "step": 1150
    },
    {
      "entropy": 6.272516870498658,
      "epoch": 0.1146288209606987,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499997353529548,
      "loss": 6.1818,
      "mean_token_accuracy": 0.11805973723530769,
      "num_tokens": 2541935.0,
      "step": 1155
    },
    {
      "entropy": 6.2223155975341795,
      "epoch": 0.11512504962286621,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999971788914303,
      "loss": 6.2789,
      "mean_token_accuracy": 0.11048080250620843,
      "num_tokens": 2553224.0,
      "step": 1160
    },
    {
      "entropy": 6.273540782928467,
      "epoch": 0.11562127828503374,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999969986738672,
      "loss": 6.3094,
      "mean_token_accuracy": 0.1041909970343113,
      "num_tokens": 2565355.0,
      "step": 1165
    },
    {
      "entropy": 6.392901992797851,
      "epoch": 0.11611750694720127,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999968128768634,
      "loss": 6.4025,
      "mean_token_accuracy": 0.10415177568793296,
      "num_tokens": 2575962.0,
      "step": 1170
    },
    {
      "entropy": 6.349782133102417,
      "epoch": 0.1166137356093688,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999966215004236,
      "loss": 6.4036,
      "mean_token_accuracy": 0.11040559411048889,
      "num_tokens": 2588136.0,
      "step": 1175
    },
    {
      "entropy": 6.387732315063476,
      "epoch": 0.11710996427153632,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999964245445522,
      "loss": 6.2871,
      "mean_token_accuracy": 0.10886711925268173,
      "num_tokens": 2597785.0,
      "step": 1180
    },
    {
      "entropy": 6.262753820419311,
      "epoch": 0.11760619293370385,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999962220092545,
      "loss": 6.3365,
      "mean_token_accuracy": 0.11655003651976585,
      "num_tokens": 2609738.0,
      "step": 1185
    },
    {
      "entropy": 6.280587577819825,
      "epoch": 0.11810242159587138,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999960138945351,
      "loss": 6.2248,
      "mean_token_accuracy": 0.11575354188680649,
      "num_tokens": 2620179.0,
      "step": 1190
    },
    {
      "entropy": 6.265934133529663,
      "epoch": 0.11859865025803891,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999958002003996,
      "loss": 6.2575,
      "mean_token_accuracy": 0.1127023346722126,
      "num_tokens": 2631550.0,
      "step": 1195
    },
    {
      "entropy": 6.305636072158814,
      "epoch": 0.11909487892020643,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999955809268529,
      "loss": 6.3107,
      "mean_token_accuracy": 0.10845914930105209,
      "num_tokens": 2642467.0,
      "step": 1200
    },
    {
      "entropy": 6.269169902801513,
      "epoch": 0.11959110758237396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999953560739008,
      "loss": 6.2755,
      "mean_token_accuracy": 0.1076017014682293,
      "num_tokens": 2652132.0,
      "step": 1205
    },
    {
      "entropy": 6.31046495437622,
      "epoch": 0.12008733624454149,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999951256415485,
      "loss": 6.3359,
      "mean_token_accuracy": 0.10693469643592834,
      "num_tokens": 2663336.0,
      "step": 1210
    },
    {
      "entropy": 6.4014545440673825,
      "epoch": 0.12058356490670902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499994889629802,
      "loss": 6.4135,
      "mean_token_accuracy": 0.10004161298274994,
      "num_tokens": 2674201.0,
      "step": 1215
    },
    {
      "entropy": 6.318805313110351,
      "epoch": 0.12107979356887653,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499994648038667,
      "loss": 6.3365,
      "mean_token_accuracy": 0.11189582347869872,
      "num_tokens": 2685073.0,
      "step": 1220
    },
    {
      "entropy": 6.3630119323730465,
      "epoch": 0.12157602223104406,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999944008681496,
      "loss": 6.2728,
      "mean_token_accuracy": 0.10747573971748352,
      "num_tokens": 2695024.0,
      "step": 1225
    },
    {
      "entropy": 6.2960004806518555,
      "epoch": 0.1220722508932116,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999941481182559,
      "loss": 6.3667,
      "mean_token_accuracy": 0.10157103613018989,
      "num_tokens": 2706232.0,
      "step": 1230
    },
    {
      "entropy": 6.2944573879241945,
      "epoch": 0.12256847955537911,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999938897889921,
      "loss": 6.3495,
      "mean_token_accuracy": 0.11251293942332267,
      "num_tokens": 2715874.0,
      "step": 1235
    },
    {
      "entropy": 6.293622159957886,
      "epoch": 0.12306470821754664,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999936258803647,
      "loss": 6.2806,
      "mean_token_accuracy": 0.11366166844964028,
      "num_tokens": 2727409.0,
      "step": 1240
    },
    {
      "entropy": 6.216445827484131,
      "epoch": 0.12356093687971417,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999933563923802,
      "loss": 6.1902,
      "mean_token_accuracy": 0.11650204434990882,
      "num_tokens": 2737259.0,
      "step": 1245
    },
    {
      "entropy": 6.291175603866577,
      "epoch": 0.1240571655418817,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999930813250451,
      "loss": 6.3362,
      "mean_token_accuracy": 0.10701057985424996,
      "num_tokens": 2748416.0,
      "step": 1250
    },
    {
      "entropy": 6.2936169624328615,
      "epoch": 0.12455339420404922,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999928006783667,
      "loss": 6.2966,
      "mean_token_accuracy": 0.11243170276284217,
      "num_tokens": 2760470.0,
      "step": 1255
    },
    {
      "entropy": 6.355133771896362,
      "epoch": 0.12504962286621676,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999925144523515,
      "loss": 6.3282,
      "mean_token_accuracy": 0.10313374847173691,
      "num_tokens": 2770222.0,
      "step": 1260
    },
    {
      "entropy": 6.230967378616333,
      "epoch": 0.12554585152838427,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999922226470066,
      "loss": 6.2494,
      "mean_token_accuracy": 0.11002011597156525,
      "num_tokens": 2780212.0,
      "step": 1265
    },
    {
      "entropy": 6.241526460647583,
      "epoch": 0.1260420801905518,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999919252623397,
      "loss": 6.207,
      "mean_token_accuracy": 0.1117776483297348,
      "num_tokens": 2790239.0,
      "step": 1270
    },
    {
      "entropy": 6.312730836868286,
      "epoch": 0.12653830885271933,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999916222983577,
      "loss": 6.2607,
      "mean_token_accuracy": 0.11156586781144143,
      "num_tokens": 2801902.0,
      "step": 1275
    },
    {
      "entropy": 6.192100143432617,
      "epoch": 0.12703453751488686,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999913137550682,
      "loss": 6.2456,
      "mean_token_accuracy": 0.11060267984867096,
      "num_tokens": 2813268.0,
      "step": 1280
    },
    {
      "entropy": 6.3050133228302006,
      "epoch": 0.1275307661770544,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999909996324791,
      "loss": 6.3387,
      "mean_token_accuracy": 0.1083206556737423,
      "num_tokens": 2824506.0,
      "step": 1285
    },
    {
      "entropy": 6.242365169525146,
      "epoch": 0.12802699483922192,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499990679930598,
      "loss": 6.2258,
      "mean_token_accuracy": 0.1135059081017971,
      "num_tokens": 2835144.0,
      "step": 1290
    },
    {
      "entropy": 6.199550008773803,
      "epoch": 0.12852322350138945,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999903546494327,
      "loss": 6.2496,
      "mean_token_accuracy": 0.10745349675416946,
      "num_tokens": 2845382.0,
      "step": 1295
    },
    {
      "entropy": 6.277227210998535,
      "epoch": 0.12901945216355698,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999900237889916,
      "loss": 6.2721,
      "mean_token_accuracy": 0.11390693038702011,
      "num_tokens": 2854697.0,
      "step": 1300
    },
    {
      "entropy": 6.250763368606568,
      "epoch": 0.12951568082572448,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999896873492827,
      "loss": 6.2708,
      "mean_token_accuracy": 0.10581153780221939,
      "num_tokens": 2866216.0,
      "step": 1305
    },
    {
      "entropy": 6.232112693786621,
      "epoch": 0.130011909487892,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999893453303143,
      "loss": 6.2357,
      "mean_token_accuracy": 0.10801640972495079,
      "num_tokens": 2877734.0,
      "step": 1310
    },
    {
      "entropy": 6.268355464935302,
      "epoch": 0.13050813815005954,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999889977320949,
      "loss": 6.2818,
      "mean_token_accuracy": 0.11378113329410552,
      "num_tokens": 2889445.0,
      "step": 1315
    },
    {
      "entropy": 6.1624878406524655,
      "epoch": 0.13100436681222707,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999886445546332,
      "loss": 6.2205,
      "mean_token_accuracy": 0.11409926936030387,
      "num_tokens": 2900485.0,
      "step": 1320
    },
    {
      "entropy": 6.275825309753418,
      "epoch": 0.1315005954743946,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499988285797938,
      "loss": 6.2114,
      "mean_token_accuracy": 0.11664246246218682,
      "num_tokens": 2909795.0,
      "step": 1325
    },
    {
      "entropy": 6.141288185119629,
      "epoch": 0.13199682413656214,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499987921462018,
      "loss": 6.2406,
      "mean_token_accuracy": 0.12180062383413315,
      "num_tokens": 2920153.0,
      "step": 1330
    },
    {
      "entropy": 6.287366437911987,
      "epoch": 0.13249305279872967,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999875515468825,
      "loss": 6.2265,
      "mean_token_accuracy": 0.11113621741533279,
      "num_tokens": 2930457.0,
      "step": 1335
    },
    {
      "entropy": 6.224073600769043,
      "epoch": 0.13298928146089717,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999871760525403,
      "loss": 6.2823,
      "mean_token_accuracy": 0.11152325570583344,
      "num_tokens": 2941305.0,
      "step": 1340
    },
    {
      "entropy": 6.289729070663452,
      "epoch": 0.1334855101230647,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999867949790011,
      "loss": 6.2335,
      "mean_token_accuracy": 0.11408303305506706,
      "num_tokens": 2951798.0,
      "step": 1345
    },
    {
      "entropy": 6.166990613937378,
      "epoch": 0.13398173878523223,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999864083262743,
      "loss": 6.1428,
      "mean_token_accuracy": 0.11790908426046372,
      "num_tokens": 2962486.0,
      "step": 1350
    },
    {
      "entropy": 6.169702672958374,
      "epoch": 0.13447796744739976,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999860160943691,
      "loss": 6.2522,
      "mean_token_accuracy": 0.11216171905398369,
      "num_tokens": 2974317.0,
      "step": 1355
    },
    {
      "entropy": 6.180521392822266,
      "epoch": 0.1349741961095673,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999856182832958,
      "loss": 6.1615,
      "mean_token_accuracy": 0.1149752452969551,
      "num_tokens": 2985440.0,
      "step": 1360
    },
    {
      "entropy": 6.191335153579712,
      "epoch": 0.13547042477173482,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999852148930637,
      "loss": 6.2346,
      "mean_token_accuracy": 0.11244972348213196,
      "num_tokens": 2997044.0,
      "step": 1365
    },
    {
      "entropy": 6.225315809249878,
      "epoch": 0.13596665343390235,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999848059236832,
      "loss": 6.2361,
      "mean_token_accuracy": 0.11013163551688195,
      "num_tokens": 3005965.0,
      "step": 1370
    },
    {
      "entropy": 6.126869249343872,
      "epoch": 0.13646288209606988,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999843913751643,
      "loss": 6.1093,
      "mean_token_accuracy": 0.1182452142238617,
      "num_tokens": 3015031.0,
      "step": 1375
    },
    {
      "entropy": 6.22548999786377,
      "epoch": 0.13695911075823738,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999839712475174,
      "loss": 6.1845,
      "mean_token_accuracy": 0.11249931305646896,
      "num_tokens": 3026275.0,
      "step": 1380
    },
    {
      "entropy": 6.159559965133667,
      "epoch": 0.13745533942040492,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999835455407526,
      "loss": 6.2224,
      "mean_token_accuracy": 0.10894647613167763,
      "num_tokens": 3036778.0,
      "step": 1385
    },
    {
      "entropy": 6.311957597732544,
      "epoch": 0.13795156808257245,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999831142548808,
      "loss": 6.3275,
      "mean_token_accuracy": 0.11298574656248092,
      "num_tokens": 3048149.0,
      "step": 1390
    },
    {
      "entropy": 6.224467945098877,
      "epoch": 0.13844779674473998,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999826773899124,
      "loss": 6.1851,
      "mean_token_accuracy": 0.10839088261127472,
      "num_tokens": 3059149.0,
      "step": 1395
    },
    {
      "entropy": 6.21338529586792,
      "epoch": 0.1389440254069075,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999822349458586,
      "loss": 6.1999,
      "mean_token_accuracy": 0.11408259272575379,
      "num_tokens": 3071613.0,
      "step": 1400
    },
    {
      "entropy": 6.2782035827636715,
      "epoch": 0.13944025406907504,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999817869227301,
      "loss": 6.3425,
      "mean_token_accuracy": 0.11115401610732079,
      "num_tokens": 3084742.0,
      "step": 1405
    },
    {
      "entropy": 6.234432363510132,
      "epoch": 0.13993648273124257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999813333205381,
      "loss": 6.2621,
      "mean_token_accuracy": 0.11374682188034058,
      "num_tokens": 3095356.0,
      "step": 1410
    },
    {
      "entropy": 6.204420900344848,
      "epoch": 0.14043271139341007,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999808741392937,
      "loss": 6.2371,
      "mean_token_accuracy": 0.11067028492689132,
      "num_tokens": 3107717.0,
      "step": 1415
    },
    {
      "entropy": 6.203734874725342,
      "epoch": 0.1409289400555776,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999804093790086,
      "loss": 6.2391,
      "mean_token_accuracy": 0.11028466001152992,
      "num_tokens": 3119205.0,
      "step": 1420
    },
    {
      "entropy": 6.267031717300415,
      "epoch": 0.14142516871774513,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499979939039694,
      "loss": 6.2803,
      "mean_token_accuracy": 0.11363445743918418,
      "num_tokens": 3130564.0,
      "step": 1425
    },
    {
      "entropy": 6.160692358016968,
      "epoch": 0.14192139737991266,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999794631213618,
      "loss": 6.2052,
      "mean_token_accuracy": 0.10955202728509902,
      "num_tokens": 3141883.0,
      "step": 1430
    },
    {
      "entropy": 6.265772008895874,
      "epoch": 0.1424176260420802,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999789816240236,
      "loss": 6.2202,
      "mean_token_accuracy": 0.10992807522416115,
      "num_tokens": 3152817.0,
      "step": 1435
    },
    {
      "entropy": 6.103646850585937,
      "epoch": 0.14291385470424772,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999784945476915,
      "loss": 6.1302,
      "mean_token_accuracy": 0.11586950048804283,
      "num_tokens": 3163546.0,
      "step": 1440
    },
    {
      "entropy": 6.2015903949737545,
      "epoch": 0.14341008336641525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999780018923775,
      "loss": 6.1753,
      "mean_token_accuracy": 0.10785740315914154,
      "num_tokens": 3173899.0,
      "step": 1445
    },
    {
      "entropy": 6.085457897186279,
      "epoch": 0.14390631202858278,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999775036580939,
      "loss": 6.0816,
      "mean_token_accuracy": 0.12055836096405984,
      "num_tokens": 3184433.0,
      "step": 1450
    },
    {
      "entropy": 6.201753950119018,
      "epoch": 0.1444025406907503,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999769998448529,
      "loss": 6.2147,
      "mean_token_accuracy": 0.1092314139008522,
      "num_tokens": 3196077.0,
      "step": 1455
    },
    {
      "entropy": 6.093402910232544,
      "epoch": 0.14489876935291782,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999764904526672,
      "loss": 6.1219,
      "mean_token_accuracy": 0.11739727333188058,
      "num_tokens": 3206645.0,
      "step": 1460
    },
    {
      "entropy": 6.325100088119507,
      "epoch": 0.14539499801508535,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999759754815492,
      "loss": 6.2504,
      "mean_token_accuracy": 0.11532395109534263,
      "num_tokens": 3216786.0,
      "step": 1465
    },
    {
      "entropy": 6.08738226890564,
      "epoch": 0.14589122667725288,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999754549315118,
      "loss": 6.1654,
      "mean_token_accuracy": 0.11355153918266296,
      "num_tokens": 3227384.0,
      "step": 1470
    },
    {
      "entropy": 6.096103143692017,
      "epoch": 0.1463874553394204,
      "grad_norm": 0.875,
      "learning_rate": 0.000499974928802568,
      "loss": 6.0187,
      "mean_token_accuracy": 0.12060984298586845,
      "num_tokens": 3238800.0,
      "step": 1475
    },
    {
      "entropy": 6.12294373512268,
      "epoch": 0.14688368400158794,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999743970947306,
      "loss": 6.1841,
      "mean_token_accuracy": 0.11260100603103637,
      "num_tokens": 3248627.0,
      "step": 1480
    },
    {
      "entropy": 6.18682632446289,
      "epoch": 0.14737991266375547,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499973859808013,
      "loss": 6.2071,
      "mean_token_accuracy": 0.11222802922129631,
      "num_tokens": 3258989.0,
      "step": 1485
    },
    {
      "entropy": 6.105042552947998,
      "epoch": 0.14787614132592297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999733169424284,
      "loss": 6.1819,
      "mean_token_accuracy": 0.11515006572008132,
      "num_tokens": 3269474.0,
      "step": 1490
    },
    {
      "entropy": 6.185284376144409,
      "epoch": 0.1483723699880905,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999727684979904,
      "loss": 6.1665,
      "mean_token_accuracy": 0.11946634948253632,
      "num_tokens": 3280453.0,
      "step": 1495
    },
    {
      "entropy": 6.12920880317688,
      "epoch": 0.14886859865025803,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999722144747125,
      "loss": 6.102,
      "mean_token_accuracy": 0.11274281144142151,
      "num_tokens": 3290234.0,
      "step": 1500
    },
    {
      "entropy": 6.056111621856689,
      "epoch": 0.14936482731242556,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999716548726085,
      "loss": 6.0829,
      "mean_token_accuracy": 0.12121317386627198,
      "num_tokens": 3300943.0,
      "step": 1505
    },
    {
      "entropy": 6.071580934524536,
      "epoch": 0.1498610559745931,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999710896916921,
      "loss": 6.0445,
      "mean_token_accuracy": 0.12006105631589889,
      "num_tokens": 3311296.0,
      "step": 1510
    },
    {
      "entropy": 6.158681344985962,
      "epoch": 0.15035728463676062,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999705189319774,
      "loss": 6.1927,
      "mean_token_accuracy": 0.11136121302843094,
      "num_tokens": 3322310.0,
      "step": 1515
    },
    {
      "entropy": 6.132280540466309,
      "epoch": 0.15085351329892815,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999699425934788,
      "loss": 6.199,
      "mean_token_accuracy": 0.11201063767075539,
      "num_tokens": 3334176.0,
      "step": 1520
    },
    {
      "entropy": 6.151670312881469,
      "epoch": 0.15134974196109569,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999693606762103,
      "loss": 6.191,
      "mean_token_accuracy": 0.1106182612478733,
      "num_tokens": 3344839.0,
      "step": 1525
    },
    {
      "entropy": 6.192277622222901,
      "epoch": 0.1518459706232632,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999687731801865,
      "loss": 6.1643,
      "mean_token_accuracy": 0.10991204902529716,
      "num_tokens": 3357046.0,
      "step": 1530
    },
    {
      "entropy": 6.178088712692261,
      "epoch": 0.15234219928543072,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999681801054217,
      "loss": 6.2417,
      "mean_token_accuracy": 0.11275468468666076,
      "num_tokens": 3367783.0,
      "step": 1535
    },
    {
      "entropy": 6.12720513343811,
      "epoch": 0.15283842794759825,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499967581451931,
      "loss": 6.0922,
      "mean_token_accuracy": 0.11282803192734718,
      "num_tokens": 3378529.0,
      "step": 1540
    },
    {
      "entropy": 6.151979255676269,
      "epoch": 0.15333465660976578,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499966977219729,
      "loss": 6.3069,
      "mean_token_accuracy": 0.11132591888308525,
      "num_tokens": 3389318.0,
      "step": 1545
    },
    {
      "entropy": 6.220948696136475,
      "epoch": 0.1538308852719333,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999663674088306,
      "loss": 6.2832,
      "mean_token_accuracy": 0.10738798677921295,
      "num_tokens": 3400603.0,
      "step": 1550
    },
    {
      "entropy": 6.218557453155517,
      "epoch": 0.15432711393410084,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999657520192511,
      "loss": 6.1828,
      "mean_token_accuracy": 0.11249059960246086,
      "num_tokens": 3410316.0,
      "step": 1555
    },
    {
      "entropy": 6.169131898880005,
      "epoch": 0.15482334259626837,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999651310510058,
      "loss": 6.1291,
      "mean_token_accuracy": 0.11865686550736428,
      "num_tokens": 3422155.0,
      "step": 1560
    },
    {
      "entropy": 6.150794935226441,
      "epoch": 0.15531957125843587,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999645045041099,
      "loss": 6.0594,
      "mean_token_accuracy": 0.11627922058105469,
      "num_tokens": 3432892.0,
      "step": 1565
    },
    {
      "entropy": 6.084249019622803,
      "epoch": 0.1558157999206034,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999638723785791,
      "loss": 6.1906,
      "mean_token_accuracy": 0.11545449271798133,
      "num_tokens": 3443967.0,
      "step": 1570
    },
    {
      "entropy": 6.1172302722930905,
      "epoch": 0.15631202858277093,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499963234674429,
      "loss": 6.0819,
      "mean_token_accuracy": 0.11921455487608909,
      "num_tokens": 3454354.0,
      "step": 1575
    },
    {
      "entropy": 6.014755821228027,
      "epoch": 0.15680825724493846,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999625913916754,
      "loss": 6.1041,
      "mean_token_accuracy": 0.11667904406785964,
      "num_tokens": 3465232.0,
      "step": 1580
    },
    {
      "entropy": 6.060571384429932,
      "epoch": 0.157304485907106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999619425303343,
      "loss": 6.0632,
      "mean_token_accuracy": 0.1209244854748249,
      "num_tokens": 3475443.0,
      "step": 1585
    },
    {
      "entropy": 6.218507719039917,
      "epoch": 0.15780071456927353,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999612880904219,
      "loss": 6.1082,
      "mean_token_accuracy": 0.11613842472434044,
      "num_tokens": 3486643.0,
      "step": 1590
    },
    {
      "entropy": 6.101606178283691,
      "epoch": 0.15829694323144106,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999606280719542,
      "loss": 6.1576,
      "mean_token_accuracy": 0.11357915550470352,
      "num_tokens": 3497557.0,
      "step": 1595
    },
    {
      "entropy": 6.210038471221924,
      "epoch": 0.1587931718936086,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999599624749475,
      "loss": 6.1679,
      "mean_token_accuracy": 0.11308549344539642,
      "num_tokens": 3509790.0,
      "step": 1600
    },
    {
      "entropy": 6.055289936065674,
      "epoch": 0.1592894005557761,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999592912994187,
      "loss": 6.0736,
      "mean_token_accuracy": 0.11441607624292374,
      "num_tokens": 3521577.0,
      "step": 1605
    },
    {
      "entropy": 6.174909019470215,
      "epoch": 0.15978562921794362,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999586145453842,
      "loss": 6.1138,
      "mean_token_accuracy": 0.11273565515875816,
      "num_tokens": 3532126.0,
      "step": 1610
    },
    {
      "entropy": 5.93880524635315,
      "epoch": 0.16028185788011115,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999579322128608,
      "loss": 5.9598,
      "mean_token_accuracy": 0.11700184866786004,
      "num_tokens": 3542560.0,
      "step": 1615
    },
    {
      "entropy": 6.074759674072266,
      "epoch": 0.16077808654227868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999572443018654,
      "loss": 6.1226,
      "mean_token_accuracy": 0.11485312059521675,
      "num_tokens": 3553863.0,
      "step": 1620
    },
    {
      "entropy": 6.075677394866943,
      "epoch": 0.1612743152044462,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999565508124151,
      "loss": 6.1502,
      "mean_token_accuracy": 0.11331931576132774,
      "num_tokens": 3564953.0,
      "step": 1625
    },
    {
      "entropy": 6.152407836914063,
      "epoch": 0.16177054386661374,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499955851744527,
      "loss": 6.1374,
      "mean_token_accuracy": 0.11724982708692551,
      "num_tokens": 3576247.0,
      "step": 1630
    },
    {
      "entropy": 6.17358717918396,
      "epoch": 0.16226677252878127,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999551470982185,
      "loss": 6.1659,
      "mean_token_accuracy": 0.11510409265756608,
      "num_tokens": 3587600.0,
      "step": 1635
    },
    {
      "entropy": 6.192458963394165,
      "epoch": 0.1627630011909488,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999544368735073,
      "loss": 6.1574,
      "mean_token_accuracy": 0.11837180629372597,
      "num_tokens": 3597727.0,
      "step": 1640
    },
    {
      "entropy": 6.070228481292725,
      "epoch": 0.1632592298531163,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999537210704106,
      "loss": 6.1294,
      "mean_token_accuracy": 0.11915965527296066,
      "num_tokens": 3608066.0,
      "step": 1645
    },
    {
      "entropy": 6.137539434432983,
      "epoch": 0.16375545851528384,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999529996889463,
      "loss": 6.1911,
      "mean_token_accuracy": 0.11308414563536644,
      "num_tokens": 3618263.0,
      "step": 1650
    },
    {
      "entropy": 6.156264162063598,
      "epoch": 0.16425168717745137,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999522727291325,
      "loss": 6.1235,
      "mean_token_accuracy": 0.12028616815805435,
      "num_tokens": 3629465.0,
      "step": 1655
    },
    {
      "entropy": 6.114594221115112,
      "epoch": 0.1647479158396189,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499951540190987,
      "loss": 6.1439,
      "mean_token_accuracy": 0.11190734803676605,
      "num_tokens": 3639657.0,
      "step": 1660
    },
    {
      "entropy": 6.203351068496704,
      "epoch": 0.16524414450178643,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999508020745281,
      "loss": 6.1155,
      "mean_token_accuracy": 0.11490047574043274,
      "num_tokens": 3652374.0,
      "step": 1665
    },
    {
      "entropy": 6.061782884597778,
      "epoch": 0.16574037316395396,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999500583797739,
      "loss": 6.0897,
      "mean_token_accuracy": 0.1153049759566784,
      "num_tokens": 3664083.0,
      "step": 1670
    },
    {
      "entropy": 6.046120309829712,
      "epoch": 0.1662366018261215,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999493091067431,
      "loss": 6.1311,
      "mean_token_accuracy": 0.12204677686095237,
      "num_tokens": 3675314.0,
      "step": 1675
    },
    {
      "entropy": 6.110049533843994,
      "epoch": 0.166732830488289,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999485542554541,
      "loss": 6.0837,
      "mean_token_accuracy": 0.10843433886766433,
      "num_tokens": 3686019.0,
      "step": 1680
    },
    {
      "entropy": 6.155540800094604,
      "epoch": 0.16722905915045652,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999477938259256,
      "loss": 6.1827,
      "mean_token_accuracy": 0.11403719708323479,
      "num_tokens": 3698170.0,
      "step": 1685
    },
    {
      "entropy": 6.084233188629151,
      "epoch": 0.16772528781262405,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999470278181766,
      "loss": 6.1775,
      "mean_token_accuracy": 0.11221502125263214,
      "num_tokens": 3709243.0,
      "step": 1690
    },
    {
      "entropy": 6.14380407333374,
      "epoch": 0.16822151647479158,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999462562322261,
      "loss": 6.1202,
      "mean_token_accuracy": 0.11952908337116241,
      "num_tokens": 3720196.0,
      "step": 1695
    },
    {
      "entropy": 6.120664167404175,
      "epoch": 0.1687177451369591,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499945479068093,
      "loss": 6.1849,
      "mean_token_accuracy": 0.1177287794649601,
      "num_tokens": 3732714.0,
      "step": 1700
    },
    {
      "entropy": 6.192463541030884,
      "epoch": 0.16921397379912664,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004999446963257969,
      "loss": 6.164,
      "mean_token_accuracy": 0.1093910776078701,
      "num_tokens": 3746474.0,
      "step": 1705
    },
    {
      "entropy": 6.031261873245239,
      "epoch": 0.16971020246129417,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999439080053571,
      "loss": 6.0267,
      "mean_token_accuracy": 0.12488168627023696,
      "num_tokens": 3757528.0,
      "step": 1710
    },
    {
      "entropy": 6.242232370376587,
      "epoch": 0.1702064311234617,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999431141067928,
      "loss": 6.1591,
      "mean_token_accuracy": 0.11146312952041626,
      "num_tokens": 3768360.0,
      "step": 1715
    },
    {
      "entropy": 6.0257631778717045,
      "epoch": 0.1707026597856292,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999423146301242,
      "loss": 6.0367,
      "mean_token_accuracy": 0.1159180223941803,
      "num_tokens": 3779809.0,
      "step": 1720
    },
    {
      "entropy": 5.9951259136199955,
      "epoch": 0.17119888844779674,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999415095753709,
      "loss": 6.0272,
      "mean_token_accuracy": 0.11418144479393959,
      "num_tokens": 3791455.0,
      "step": 1725
    },
    {
      "entropy": 6.083953142166138,
      "epoch": 0.17169511710996427,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999406989425527,
      "loss": 6.0646,
      "mean_token_accuracy": 0.11992244198918342,
      "num_tokens": 3803209.0,
      "step": 1730
    },
    {
      "entropy": 6.023218822479248,
      "epoch": 0.1721913457721318,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999398827316901,
      "loss": 6.0844,
      "mean_token_accuracy": 0.11768920570611954,
      "num_tokens": 3813553.0,
      "step": 1735
    },
    {
      "entropy": 6.128506755828857,
      "epoch": 0.17268757443429933,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499939060942803,
      "loss": 6.1466,
      "mean_token_accuracy": 0.11251909062266349,
      "num_tokens": 3824856.0,
      "step": 1740
    },
    {
      "entropy": 6.2042152881622314,
      "epoch": 0.17318380309646686,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999382335759119,
      "loss": 6.2169,
      "mean_token_accuracy": 0.11170056387782097,
      "num_tokens": 3835489.0,
      "step": 1745
    },
    {
      "entropy": 6.1185472965240475,
      "epoch": 0.1736800317586344,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999374006310372,
      "loss": 6.0584,
      "mean_token_accuracy": 0.11355099081993103,
      "num_tokens": 3844838.0,
      "step": 1750
    },
    {
      "entropy": 5.984942770004272,
      "epoch": 0.1741762604208019,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999365621081999,
      "loss": 6.02,
      "mean_token_accuracy": 0.11664588525891303,
      "num_tokens": 3855288.0,
      "step": 1755
    },
    {
      "entropy": 6.108725595474243,
      "epoch": 0.17467248908296942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999357180074204,
      "loss": 6.065,
      "mean_token_accuracy": 0.12154223769903183,
      "num_tokens": 3865585.0,
      "step": 1760
    },
    {
      "entropy": 6.077434921264649,
      "epoch": 0.17516871774513695,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999348683287199,
      "loss": 6.0586,
      "mean_token_accuracy": 0.11305517330765724,
      "num_tokens": 3877694.0,
      "step": 1765
    },
    {
      "entropy": 6.053066635131836,
      "epoch": 0.17566494640730448,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999340130721192,
      "loss": 6.0273,
      "mean_token_accuracy": 0.12292965203523636,
      "num_tokens": 3888802.0,
      "step": 1770
    },
    {
      "entropy": 5.994657707214356,
      "epoch": 0.17616117506947201,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999331522376396,
      "loss": 5.9963,
      "mean_token_accuracy": 0.11429235264658928,
      "num_tokens": 3900294.0,
      "step": 1775
    },
    {
      "entropy": 6.0335695266723635,
      "epoch": 0.17665740373163955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999322858253027,
      "loss": 6.0253,
      "mean_token_accuracy": 0.11919704973697662,
      "num_tokens": 3910865.0,
      "step": 1780
    },
    {
      "entropy": 6.024723052978516,
      "epoch": 0.17715363239380708,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999314138351297,
      "loss": 6.0896,
      "mean_token_accuracy": 0.11677447333931923,
      "num_tokens": 3921960.0,
      "step": 1785
    },
    {
      "entropy": 6.130808353424072,
      "epoch": 0.1776498610559746,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999305362671424,
      "loss": 6.1007,
      "mean_token_accuracy": 0.11497491449117661,
      "num_tokens": 3934085.0,
      "step": 1790
    },
    {
      "entropy": 6.012696743011475,
      "epoch": 0.1781460897181421,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999296531213624,
      "loss": 5.9399,
      "mean_token_accuracy": 0.12995768636465072,
      "num_tokens": 3944532.0,
      "step": 1795
    },
    {
      "entropy": 5.986821126937866,
      "epoch": 0.17864231838030964,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999287643978117,
      "loss": 5.9797,
      "mean_token_accuracy": 0.1237669236958027,
      "num_tokens": 3954995.0,
      "step": 1800
    },
    {
      "entropy": 5.993027114868164,
      "epoch": 0.17913854704247717,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999278700965124,
      "loss": 6.0469,
      "mean_token_accuracy": 0.1215439960360527,
      "num_tokens": 3966621.0,
      "step": 1805
    },
    {
      "entropy": 6.133916234970092,
      "epoch": 0.1796347757046447,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999269702174866,
      "loss": 6.1153,
      "mean_token_accuracy": 0.11664739549160004,
      "num_tokens": 3977568.0,
      "step": 1810
    },
    {
      "entropy": 6.023658990859985,
      "epoch": 0.18013100436681223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999260647607565,
      "loss": 6.1244,
      "mean_token_accuracy": 0.11225999146699905,
      "num_tokens": 3988571.0,
      "step": 1815
    },
    {
      "entropy": 6.082444190979004,
      "epoch": 0.18062723302897976,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999251537263448,
      "loss": 5.9453,
      "mean_token_accuracy": 0.12557341828942298,
      "num_tokens": 3998380.0,
      "step": 1820
    },
    {
      "entropy": 5.99331431388855,
      "epoch": 0.1811234616911473,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999242371142739,
      "loss": 6.1229,
      "mean_token_accuracy": 0.11956584751605988,
      "num_tokens": 4010017.0,
      "step": 1825
    },
    {
      "entropy": 6.134469366073608,
      "epoch": 0.1816196903533148,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999233149245665,
      "loss": 6.0949,
      "mean_token_accuracy": 0.11493808627128602,
      "num_tokens": 4020922.0,
      "step": 1830
    },
    {
      "entropy": 6.083890724182129,
      "epoch": 0.18211591901548232,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999223871572456,
      "loss": 6.0705,
      "mean_token_accuracy": 0.11875133067369462,
      "num_tokens": 4030927.0,
      "step": 1835
    },
    {
      "entropy": 6.0421332836151125,
      "epoch": 0.18261214767764986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999214538123342,
      "loss": 6.1447,
      "mean_token_accuracy": 0.11264287009835243,
      "num_tokens": 4041375.0,
      "step": 1840
    },
    {
      "entropy": 6.1017091274261475,
      "epoch": 0.18310837633981739,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999205148898554,
      "loss": 6.137,
      "mean_token_accuracy": 0.11522981524467468,
      "num_tokens": 4053213.0,
      "step": 1845
    },
    {
      "entropy": 6.06529393196106,
      "epoch": 0.18360460500198492,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999195703898325,
      "loss": 5.9823,
      "mean_token_accuracy": 0.12415849640965462,
      "num_tokens": 4064536.0,
      "step": 1850
    },
    {
      "entropy": 6.056224679946899,
      "epoch": 0.18410083366415245,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999186203122889,
      "loss": 6.0555,
      "mean_token_accuracy": 0.12177475914359093,
      "num_tokens": 4075282.0,
      "step": 1855
    },
    {
      "entropy": 5.956262111663818,
      "epoch": 0.18459706232631998,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999176646572482,
      "loss": 6.0447,
      "mean_token_accuracy": 0.12330468744039536,
      "num_tokens": 4085501.0,
      "step": 1860
    },
    {
      "entropy": 6.081436061859131,
      "epoch": 0.1850932909884875,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499916703424734,
      "loss": 6.0013,
      "mean_token_accuracy": 0.11329571008682252,
      "num_tokens": 4095697.0,
      "step": 1865
    },
    {
      "entropy": 6.003993368148803,
      "epoch": 0.185589519650655,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999157366147702,
      "loss": 6.0672,
      "mean_token_accuracy": 0.11597673147916794,
      "num_tokens": 4106459.0,
      "step": 1870
    },
    {
      "entropy": 6.079233932495117,
      "epoch": 0.18608574831282254,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999147642273809,
      "loss": 5.991,
      "mean_token_accuracy": 0.12156878933310508,
      "num_tokens": 4118320.0,
      "step": 1875
    },
    {
      "entropy": 5.973481559753418,
      "epoch": 0.18658197697499007,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00049991378626259,
      "loss": 6.0224,
      "mean_token_accuracy": 0.12918237894773482,
      "num_tokens": 4131249.0,
      "step": 1880
    },
    {
      "entropy": 6.072328758239746,
      "epoch": 0.1870782056371576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999128027204219,
      "loss": 6.0921,
      "mean_token_accuracy": 0.11764958873391151,
      "num_tokens": 4141480.0,
      "step": 1885
    },
    {
      "entropy": 6.080035591125489,
      "epoch": 0.18757443429932513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499911813600901,
      "loss": 6.1433,
      "mean_token_accuracy": 0.11640222296118737,
      "num_tokens": 4152344.0,
      "step": 1890
    },
    {
      "entropy": 6.137289190292359,
      "epoch": 0.18807066296149266,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999108189040516,
      "loss": 6.1126,
      "mean_token_accuracy": 0.11466791629791259,
      "num_tokens": 4164772.0,
      "step": 1895
    },
    {
      "entropy": 6.027769136428833,
      "epoch": 0.1885668916236602,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999098186298986,
      "loss": 6.0443,
      "mean_token_accuracy": 0.11440513208508492,
      "num_tokens": 4175511.0,
      "step": 1900
    },
    {
      "entropy": 6.023436784744263,
      "epoch": 0.1890631202858277,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999088127784667,
      "loss": 6.0164,
      "mean_token_accuracy": 0.11896950975060464,
      "num_tokens": 4187445.0,
      "step": 1905
    },
    {
      "entropy": 6.051607608795166,
      "epoch": 0.18955934894799523,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499907801349781,
      "loss": 5.9987,
      "mean_token_accuracy": 0.11835644766688347,
      "num_tokens": 4198124.0,
      "step": 1910
    },
    {
      "entropy": 5.978174209594727,
      "epoch": 0.19005557761016276,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999067843438664,
      "loss": 6.047,
      "mean_token_accuracy": 0.11988429352641106,
      "num_tokens": 4208954.0,
      "step": 1915
    },
    {
      "entropy": 6.129126739501953,
      "epoch": 0.1905518062723303,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999057617607482,
      "loss": 6.1052,
      "mean_token_accuracy": 0.11723098531365395,
      "num_tokens": 4218896.0,
      "step": 1920
    },
    {
      "entropy": 5.979246139526367,
      "epoch": 0.19104803493449782,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999047336004517,
      "loss": 6.0138,
      "mean_token_accuracy": 0.11616118773818015,
      "num_tokens": 4228851.0,
      "step": 1925
    },
    {
      "entropy": 5.985688304901123,
      "epoch": 0.19154426359666535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999036998630024,
      "loss": 5.977,
      "mean_token_accuracy": 0.11649016365408897,
      "num_tokens": 4240377.0,
      "step": 1930
    },
    {
      "entropy": 5.950987052917481,
      "epoch": 0.19204049225883288,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499902660548426,
      "loss": 5.9542,
      "mean_token_accuracy": 0.1181262455880642,
      "num_tokens": 4251341.0,
      "step": 1935
    },
    {
      "entropy": 6.056128025054932,
      "epoch": 0.1925367209210004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999016156567482,
      "loss": 6.0579,
      "mean_token_accuracy": 0.11999004855751991,
      "num_tokens": 4261206.0,
      "step": 1940
    },
    {
      "entropy": 6.091788625717163,
      "epoch": 0.1930329495831679,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499900565187995,
      "loss": 6.0659,
      "mean_token_accuracy": 0.11197969540953637,
      "num_tokens": 4272662.0,
      "step": 1945
    },
    {
      "entropy": 6.060900783538818,
      "epoch": 0.19352917824533544,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998995091421924,
      "loss": 6.0496,
      "mean_token_accuracy": 0.11381187662482262,
      "num_tokens": 4283302.0,
      "step": 1950
    },
    {
      "entropy": 5.977718019485474,
      "epoch": 0.19402540690750297,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998984475193665,
      "loss": 5.9944,
      "mean_token_accuracy": 0.11904016137123108,
      "num_tokens": 4296087.0,
      "step": 1955
    },
    {
      "entropy": 5.980944347381592,
      "epoch": 0.1945216355696705,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998973803195439,
      "loss": 6.0088,
      "mean_token_accuracy": 0.12316826283931732,
      "num_tokens": 4305918.0,
      "step": 1960
    },
    {
      "entropy": 6.02653398513794,
      "epoch": 0.19501786423183803,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998963075427508,
      "loss": 6.073,
      "mean_token_accuracy": 0.11951372250914574,
      "num_tokens": 4316507.0,
      "step": 1965
    },
    {
      "entropy": 6.028897380828857,
      "epoch": 0.19551409289400556,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998952291890138,
      "loss": 6.004,
      "mean_token_accuracy": 0.12352709099650383,
      "num_tokens": 4329091.0,
      "step": 1970
    },
    {
      "entropy": 6.0670671463012695,
      "epoch": 0.1960103215561731,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998941452583599,
      "loss": 6.0668,
      "mean_token_accuracy": 0.1214105449616909,
      "num_tokens": 4339266.0,
      "step": 1975
    },
    {
      "entropy": 5.959065103530884,
      "epoch": 0.1965065502183406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998930557508156,
      "loss": 5.9759,
      "mean_token_accuracy": 0.1258009724318981,
      "num_tokens": 4349300.0,
      "step": 1980
    },
    {
      "entropy": 5.941859292984009,
      "epoch": 0.19700277888050813,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998919606664082,
      "loss": 5.9055,
      "mean_token_accuracy": 0.12807131931185722,
      "num_tokens": 4360314.0,
      "step": 1985
    },
    {
      "entropy": 6.043554306030273,
      "epoch": 0.19749900754267566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998908600051648,
      "loss": 5.9557,
      "mean_token_accuracy": 0.12293901592493058,
      "num_tokens": 4370077.0,
      "step": 1990
    },
    {
      "entropy": 5.936801719665527,
      "epoch": 0.1979952362048432,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998897537671126,
      "loss": 5.9691,
      "mean_token_accuracy": 0.11986607536673546,
      "num_tokens": 4381000.0,
      "step": 1995
    },
    {
      "entropy": 6.05441517829895,
      "epoch": 0.19849146486701072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998886419522792,
      "loss": 6.183,
      "mean_token_accuracy": 0.11007200106978417,
      "num_tokens": 4393485.0,
      "step": 2000
    },
    {
      "entropy": 6.172304725646972,
      "epoch": 0.19898769352917825,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998875245606919,
      "loss": 6.0252,
      "mean_token_accuracy": 0.12070057019591332,
      "num_tokens": 4404354.0,
      "step": 2005
    },
    {
      "entropy": 5.957334184646607,
      "epoch": 0.19948392219134578,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998864015923788,
      "loss": 6.1616,
      "mean_token_accuracy": 0.11959527060389519,
      "num_tokens": 4414818.0,
      "step": 2010
    },
    {
      "entropy": 6.080641794204712,
      "epoch": 0.1999801508535133,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998852730473674,
      "loss": 6.0419,
      "mean_token_accuracy": 0.11496741399168968,
      "num_tokens": 4426271.0,
      "step": 2015
    },
    {
      "entropy": 5.992073965072632,
      "epoch": 0.2004763795156808,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998841389256859,
      "loss": 6.0961,
      "mean_token_accuracy": 0.11873921528458595,
      "num_tokens": 4436981.0,
      "step": 2020
    },
    {
      "entropy": 5.966401863098144,
      "epoch": 0.20097260817784834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998829992273622,
      "loss": 5.8953,
      "mean_token_accuracy": 0.12690763026475907,
      "num_tokens": 4446494.0,
      "step": 2025
    },
    {
      "entropy": 5.971911811828614,
      "epoch": 0.20146883684001587,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998818539524248,
      "loss": 6.0038,
      "mean_token_accuracy": 0.12495953813195229,
      "num_tokens": 4458535.0,
      "step": 2030
    },
    {
      "entropy": 6.010282945632935,
      "epoch": 0.2019650655021834,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499880703100902,
      "loss": 5.9931,
      "mean_token_accuracy": 0.12273974418640136,
      "num_tokens": 4469025.0,
      "step": 2035
    },
    {
      "entropy": 6.0635312557220455,
      "epoch": 0.20246129416435094,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998795466728223,
      "loss": 5.9801,
      "mean_token_accuracy": 0.12161237373948097,
      "num_tokens": 4478431.0,
      "step": 2040
    },
    {
      "entropy": 5.947754192352295,
      "epoch": 0.20295752282651847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998783846682144,
      "loss": 6.1296,
      "mean_token_accuracy": 0.11572149395942688,
      "num_tokens": 4490215.0,
      "step": 2045
    },
    {
      "entropy": 6.041604042053223,
      "epoch": 0.203453751488686,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004998772170871072,
      "loss": 5.9895,
      "mean_token_accuracy": 0.12096533477306366,
      "num_tokens": 4502895.0,
      "step": 2050
    },
    {
      "entropy": 5.9692402362823485,
      "epoch": 0.20394998015085353,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998760439295295,
      "loss": 6.0113,
      "mean_token_accuracy": 0.1204931080341339,
      "num_tokens": 4514468.0,
      "step": 2055
    },
    {
      "entropy": 6.084046506881714,
      "epoch": 0.20444620881302103,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998748651955105,
      "loss": 6.0317,
      "mean_token_accuracy": 0.12071766927838326,
      "num_tokens": 4525787.0,
      "step": 2060
    },
    {
      "entropy": 5.992368078231811,
      "epoch": 0.20494243747518856,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998736808850795,
      "loss": 6.1181,
      "mean_token_accuracy": 0.11842440962791442,
      "num_tokens": 4537536.0,
      "step": 2065
    },
    {
      "entropy": 6.07518572807312,
      "epoch": 0.2054386661373561,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998724909982656,
      "loss": 6.0545,
      "mean_token_accuracy": 0.12233483120799064,
      "num_tokens": 4549619.0,
      "step": 2070
    },
    {
      "entropy": 5.984462881088257,
      "epoch": 0.20593489479952362,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998712955350986,
      "loss": 5.9774,
      "mean_token_accuracy": 0.12310493364930153,
      "num_tokens": 4560248.0,
      "step": 2075
    },
    {
      "entropy": 6.027503871917725,
      "epoch": 0.20643112346169115,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998700944956081,
      "loss": 6.0827,
      "mean_token_accuracy": 0.1216159887611866,
      "num_tokens": 4571888.0,
      "step": 2080
    },
    {
      "entropy": 6.05411696434021,
      "epoch": 0.20692735212385868,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998688878798237,
      "loss": 5.9395,
      "mean_token_accuracy": 0.12731964141130447,
      "num_tokens": 4583234.0,
      "step": 2085
    },
    {
      "entropy": 6.018948078155518,
      "epoch": 0.2074235807860262,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998676756877753,
      "loss": 6.1471,
      "mean_token_accuracy": 0.11580592915415763,
      "num_tokens": 4595252.0,
      "step": 2090
    },
    {
      "entropy": 5.987044811248779,
      "epoch": 0.20791980944819372,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998664579194934,
      "loss": 5.9279,
      "mean_token_accuracy": 0.12194859310984611,
      "num_tokens": 4607091.0,
      "step": 2095
    },
    {
      "entropy": 6.006628370285034,
      "epoch": 0.20841603811036125,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998652345750077,
      "loss": 5.8925,
      "mean_token_accuracy": 0.12444933280348777,
      "num_tokens": 4618003.0,
      "step": 2100
    },
    {
      "entropy": 5.890622520446778,
      "epoch": 0.20891226677252878,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998640056543487,
      "loss": 5.9063,
      "mean_token_accuracy": 0.12009405642747879,
      "num_tokens": 4628429.0,
      "step": 2105
    },
    {
      "entropy": 6.005721569061279,
      "epoch": 0.2094084954346963,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998627711575469,
      "loss": 5.9697,
      "mean_token_accuracy": 0.1252060741186142,
      "num_tokens": 4639414.0,
      "step": 2110
    },
    {
      "entropy": 5.956946802139282,
      "epoch": 0.20990472409686384,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000499861531084633,
      "loss": 6.0677,
      "mean_token_accuracy": 0.11779648587107658,
      "num_tokens": 4652594.0,
      "step": 2115
    },
    {
      "entropy": 5.9904883861541744,
      "epoch": 0.21040095275903137,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998602854356376,
      "loss": 5.8914,
      "mean_token_accuracy": 0.1240698166191578,
      "num_tokens": 4662981.0,
      "step": 2120
    },
    {
      "entropy": 5.9469305038452145,
      "epoch": 0.2108971814211989,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998590342105917,
      "loss": 5.9822,
      "mean_token_accuracy": 0.1197172798216343,
      "num_tokens": 4672469.0,
      "step": 2125
    },
    {
      "entropy": 6.015959167480469,
      "epoch": 0.21139341008336643,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998577774095262,
      "loss": 5.9782,
      "mean_token_accuracy": 0.12201139032840728,
      "num_tokens": 4683640.0,
      "step": 2130
    },
    {
      "entropy": 5.936978530883789,
      "epoch": 0.21188963874553393,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998565150324724,
      "loss": 5.9337,
      "mean_token_accuracy": 0.1305203281342983,
      "num_tokens": 4694032.0,
      "step": 2135
    },
    {
      "entropy": 5.874231004714966,
      "epoch": 0.21238586740770146,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998552470794616,
      "loss": 5.8023,
      "mean_token_accuracy": 0.13009038046002389,
      "num_tokens": 4704253.0,
      "step": 2140
    },
    {
      "entropy": 5.9205982208251955,
      "epoch": 0.212882096069869,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998539735505251,
      "loss": 5.9636,
      "mean_token_accuracy": 0.11889011636376381,
      "num_tokens": 4715327.0,
      "step": 2145
    },
    {
      "entropy": 6.089485359191895,
      "epoch": 0.21337832473203652,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998526944456946,
      "loss": 5.983,
      "mean_token_accuracy": 0.11755810230970383,
      "num_tokens": 4726450.0,
      "step": 2150
    },
    {
      "entropy": 5.928031587600708,
      "epoch": 0.21387455339420405,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998514097650018,
      "loss": 6.0091,
      "mean_token_accuracy": 0.11683928668498993,
      "num_tokens": 4736630.0,
      "step": 2155
    },
    {
      "entropy": 6.01918659210205,
      "epoch": 0.21437078205637158,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998501195084786,
      "loss": 5.8827,
      "mean_token_accuracy": 0.1218169391155243,
      "num_tokens": 4746592.0,
      "step": 2160
    },
    {
      "entropy": 5.958590841293335,
      "epoch": 0.21486701071853911,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998488236761569,
      "loss": 6.0581,
      "mean_token_accuracy": 0.12610213384032248,
      "num_tokens": 4757824.0,
      "step": 2165
    },
    {
      "entropy": 6.0240240573883055,
      "epoch": 0.21536323938070662,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998475222680688,
      "loss": 6.0016,
      "mean_token_accuracy": 0.12206201553344727,
      "num_tokens": 4768453.0,
      "step": 2170
    },
    {
      "entropy": 5.921792268753052,
      "epoch": 0.21585946804287415,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998462152842468,
      "loss": 5.9871,
      "mean_token_accuracy": 0.11648527160286903,
      "num_tokens": 4779664.0,
      "step": 2175
    },
    {
      "entropy": 6.0779510021209715,
      "epoch": 0.21635569670504168,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499844902724723,
      "loss": 6.0067,
      "mean_token_accuracy": 0.12150592878460884,
      "num_tokens": 4791012.0,
      "step": 2180
    },
    {
      "entropy": 5.887479829788208,
      "epoch": 0.2168519253672092,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998435845895302,
      "loss": 5.8636,
      "mean_token_accuracy": 0.12329262644052505,
      "num_tokens": 4800578.0,
      "step": 2185
    },
    {
      "entropy": 5.983295154571533,
      "epoch": 0.21734815402937674,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499842260878701,
      "loss": 5.8861,
      "mean_token_accuracy": 0.12191555723547935,
      "num_tokens": 4810940.0,
      "step": 2190
    },
    {
      "entropy": 6.002044200897217,
      "epoch": 0.21784438269154427,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004998409315922683,
      "loss": 6.0268,
      "mean_token_accuracy": 0.1215096190571785,
      "num_tokens": 4822696.0,
      "step": 2195
    },
    {
      "entropy": 5.937045526504517,
      "epoch": 0.2183406113537118,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998395967302649,
      "loss": 5.8976,
      "mean_token_accuracy": 0.12619130909442902,
      "num_tokens": 4833669.0,
      "step": 2200
    },
    {
      "entropy": 6.046645832061768,
      "epoch": 0.21883684001587933,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499838256292724,
      "loss": 6.0795,
      "mean_token_accuracy": 0.12037609070539475,
      "num_tokens": 4844021.0,
      "step": 2205
    },
    {
      "entropy": 5.948421573638916,
      "epoch": 0.21933306867804683,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998369102796787,
      "loss": 5.9532,
      "mean_token_accuracy": 0.12263412401080132,
      "num_tokens": 4855336.0,
      "step": 2210
    },
    {
      "entropy": 5.974256610870361,
      "epoch": 0.21982929734021436,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004998355586911627,
      "loss": 5.9506,
      "mean_token_accuracy": 0.12540801763534545,
      "num_tokens": 4867713.0,
      "step": 2215
    },
    {
      "entropy": 5.975857448577881,
      "epoch": 0.2203255260023819,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998342015272093,
      "loss": 5.9332,
      "mean_token_accuracy": 0.12420732304453849,
      "num_tokens": 4879134.0,
      "step": 2220
    },
    {
      "entropy": 6.003203201293945,
      "epoch": 0.22082175466454942,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998328387878521,
      "loss": 6.09,
      "mean_token_accuracy": 0.11976692378520966,
      "num_tokens": 4891755.0,
      "step": 2225
    },
    {
      "entropy": 6.1218338966369625,
      "epoch": 0.22131798332671695,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998314704731249,
      "loss": 6.0927,
      "mean_token_accuracy": 0.1174064464867115,
      "num_tokens": 4903264.0,
      "step": 2230
    },
    {
      "entropy": 5.947536039352417,
      "epoch": 0.22181421198888449,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998300965830618,
      "loss": 5.9105,
      "mean_token_accuracy": 0.12291811779141426,
      "num_tokens": 4912533.0,
      "step": 2235
    },
    {
      "entropy": 5.84995813369751,
      "epoch": 0.22231044065105202,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998287171176968,
      "loss": 5.9169,
      "mean_token_accuracy": 0.12699113339185714,
      "num_tokens": 4923288.0,
      "step": 2240
    },
    {
      "entropy": 5.968135643005371,
      "epoch": 0.22280666931321952,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998273320770641,
      "loss": 5.9355,
      "mean_token_accuracy": 0.12369446530938148,
      "num_tokens": 4934759.0,
      "step": 2245
    },
    {
      "entropy": 5.942650651931762,
      "epoch": 0.22330289797538705,
      "grad_norm": 1.0,
      "learning_rate": 0.000499825941461198,
      "loss": 5.9819,
      "mean_token_accuracy": 0.12287837192416191,
      "num_tokens": 4947024.0,
      "step": 2250
    },
    {
      "entropy": 5.953667020797729,
      "epoch": 0.22379912663755458,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998245452701329,
      "loss": 5.9185,
      "mean_token_accuracy": 0.12497033029794694,
      "num_tokens": 4959281.0,
      "step": 2255
    },
    {
      "entropy": 6.01146445274353,
      "epoch": 0.2242953552997221,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998231435039036,
      "loss": 5.9817,
      "mean_token_accuracy": 0.11844518184661865,
      "num_tokens": 4970466.0,
      "step": 2260
    },
    {
      "entropy": 5.9487926959991455,
      "epoch": 0.22479158396188964,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998217361625449,
      "loss": 5.9227,
      "mean_token_accuracy": 0.12402417287230491,
      "num_tokens": 4981135.0,
      "step": 2265
    },
    {
      "entropy": 5.973719263076783,
      "epoch": 0.22528781262405717,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998203232460915,
      "loss": 5.9895,
      "mean_token_accuracy": 0.1295408844947815,
      "num_tokens": 4991292.0,
      "step": 2270
    },
    {
      "entropy": 5.957491159439087,
      "epoch": 0.2257840412862247,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998189047545785,
      "loss": 5.9457,
      "mean_token_accuracy": 0.12670203521847725,
      "num_tokens": 5002806.0,
      "step": 2275
    },
    {
      "entropy": 5.917588376998902,
      "epoch": 0.22628026994839223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998174806880411,
      "loss": 5.8916,
      "mean_token_accuracy": 0.13011438325047492,
      "num_tokens": 5013474.0,
      "step": 2280
    },
    {
      "entropy": 5.979103899002075,
      "epoch": 0.22677649861055973,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998160510465146,
      "loss": 5.934,
      "mean_token_accuracy": 0.12663510590791702,
      "num_tokens": 5025151.0,
      "step": 2285
    },
    {
      "entropy": 5.858231687545777,
      "epoch": 0.22727272727272727,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998146158300345,
      "loss": 5.9403,
      "mean_token_accuracy": 0.12440521046519279,
      "num_tokens": 5037166.0,
      "step": 2290
    },
    {
      "entropy": 6.016402339935302,
      "epoch": 0.2277689559348948,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998131750386364,
      "loss": 5.972,
      "mean_token_accuracy": 0.12612202540040016,
      "num_tokens": 5047452.0,
      "step": 2295
    },
    {
      "entropy": 5.914637041091919,
      "epoch": 0.22826518459706233,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998117286723559,
      "loss": 5.9791,
      "mean_token_accuracy": 0.11908696666359901,
      "num_tokens": 5058501.0,
      "step": 2300
    },
    {
      "entropy": 5.990808296203613,
      "epoch": 0.22876141325922986,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499810276731229,
      "loss": 5.9825,
      "mean_token_accuracy": 0.12161915823817253,
      "num_tokens": 5069716.0,
      "step": 2305
    },
    {
      "entropy": 6.073873853683471,
      "epoch": 0.2292576419213974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998088192152918,
      "loss": 5.9592,
      "mean_token_accuracy": 0.1198742963373661,
      "num_tokens": 5078509.0,
      "step": 2310
    },
    {
      "entropy": 5.910210466384887,
      "epoch": 0.22975387058356492,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998073561245801,
      "loss": 6.011,
      "mean_token_accuracy": 0.12116109505295754,
      "num_tokens": 5090355.0,
      "step": 2315
    },
    {
      "entropy": 6.0268512725830075,
      "epoch": 0.23025009924573242,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998058874591305,
      "loss": 5.8952,
      "mean_token_accuracy": 0.1269051119685173,
      "num_tokens": 5100582.0,
      "step": 2320
    },
    {
      "entropy": 5.97053108215332,
      "epoch": 0.23074632790789995,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998044132189793,
      "loss": 5.9553,
      "mean_token_accuracy": 0.12131153792142868,
      "num_tokens": 5112176.0,
      "step": 2325
    },
    {
      "entropy": 5.976939296722412,
      "epoch": 0.23124255657006748,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499802933404163,
      "loss": 5.8825,
      "mean_token_accuracy": 0.12292413413524628,
      "num_tokens": 5122840.0,
      "step": 2330
    },
    {
      "entropy": 5.9486147403717045,
      "epoch": 0.231738785232235,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004998014480147183,
      "loss": 5.88,
      "mean_token_accuracy": 0.1311016447842121,
      "num_tokens": 5132926.0,
      "step": 2335
    },
    {
      "entropy": 5.891398763656616,
      "epoch": 0.23223501389440254,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997999570506823,
      "loss": 5.8679,
      "mean_token_accuracy": 0.12510914131999015,
      "num_tokens": 5143394.0,
      "step": 2340
    },
    {
      "entropy": 5.880256605148316,
      "epoch": 0.23273124255657007,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004997984605120917,
      "loss": 5.8867,
      "mean_token_accuracy": 0.1238784983754158,
      "num_tokens": 5155710.0,
      "step": 2345
    },
    {
      "entropy": 5.813053321838379,
      "epoch": 0.2332274712187376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997969583989835,
      "loss": 5.8256,
      "mean_token_accuracy": 0.12075290903449058,
      "num_tokens": 5167065.0,
      "step": 2350
    },
    {
      "entropy": 5.918558645248413,
      "epoch": 0.23372369988090513,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997954507113952,
      "loss": 5.9403,
      "mean_token_accuracy": 0.1209035113453865,
      "num_tokens": 5177535.0,
      "step": 2355
    },
    {
      "entropy": 5.928993511199951,
      "epoch": 0.23421992854307264,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997939374493642,
      "loss": 5.8424,
      "mean_token_accuracy": 0.12632499486207963,
      "num_tokens": 5188324.0,
      "step": 2360
    },
    {
      "entropy": 5.843468523025512,
      "epoch": 0.23471615720524017,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499792418612928,
      "loss": 5.9191,
      "mean_token_accuracy": 0.12788572460412978,
      "num_tokens": 5199223.0,
      "step": 2365
    },
    {
      "entropy": 5.968633937835693,
      "epoch": 0.2352123858674077,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997908942021241,
      "loss": 5.9036,
      "mean_token_accuracy": 0.12390781864523888,
      "num_tokens": 5210080.0,
      "step": 2370
    },
    {
      "entropy": 5.974946546554565,
      "epoch": 0.23570861452957523,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997893642169904,
      "loss": 6.0327,
      "mean_token_accuracy": 0.12202257290482521,
      "num_tokens": 5220867.0,
      "step": 2375
    },
    {
      "entropy": 5.952085304260254,
      "epoch": 0.23620484319174276,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997878286575648,
      "loss": 5.8665,
      "mean_token_accuracy": 0.1263822741806507,
      "num_tokens": 5231451.0,
      "step": 2380
    },
    {
      "entropy": 5.933335781097412,
      "epoch": 0.2367010718539103,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997862875238855,
      "loss": 5.9826,
      "mean_token_accuracy": 0.11929206326603889,
      "num_tokens": 5244643.0,
      "step": 2385
    },
    {
      "entropy": 5.931532573699951,
      "epoch": 0.23719730051607782,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997847408159907,
      "loss": 5.9823,
      "mean_token_accuracy": 0.11998239532113075,
      "num_tokens": 5256279.0,
      "step": 2390
    },
    {
      "entropy": 6.0057477951049805,
      "epoch": 0.23769352917824532,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997831885339186,
      "loss": 5.9682,
      "mean_token_accuracy": 0.12426605671644211,
      "num_tokens": 5268498.0,
      "step": 2395
    },
    {
      "entropy": 5.878379583358765,
      "epoch": 0.23818975784041285,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997816306777077,
      "loss": 5.9211,
      "mean_token_accuracy": 0.12448853626847267,
      "num_tokens": 5280869.0,
      "step": 2400
    },
    {
      "entropy": 6.116397285461426,
      "epoch": 0.23868598650258038,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997800672473969,
      "loss": 6.0303,
      "mean_token_accuracy": 0.12171175703406334,
      "num_tokens": 5293082.0,
      "step": 2405
    },
    {
      "entropy": 5.848374509811402,
      "epoch": 0.2391822151647479,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997784982430247,
      "loss": 5.8142,
      "mean_token_accuracy": 0.13605227544903756,
      "num_tokens": 5304039.0,
      "step": 2410
    },
    {
      "entropy": 5.858838510513306,
      "epoch": 0.23967844382691544,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997769236646301,
      "loss": 5.8998,
      "mean_token_accuracy": 0.12478934675455093,
      "num_tokens": 5315016.0,
      "step": 2415
    },
    {
      "entropy": 6.031330347061157,
      "epoch": 0.24017467248908297,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997753435122522,
      "loss": 6.0241,
      "mean_token_accuracy": 0.12272524014115334,
      "num_tokens": 5326090.0,
      "step": 2420
    },
    {
      "entropy": 5.902796745300293,
      "epoch": 0.2406709011512505,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049977375778593,
      "loss": 5.807,
      "mean_token_accuracy": 0.12468553930521012,
      "num_tokens": 5336423.0,
      "step": 2425
    },
    {
      "entropy": 5.924130249023437,
      "epoch": 0.24116712981341804,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997721664857032,
      "loss": 5.9727,
      "mean_token_accuracy": 0.12364738434553146,
      "num_tokens": 5346982.0,
      "step": 2430
    },
    {
      "entropy": 5.934656763076783,
      "epoch": 0.24166335847558554,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997705696116108,
      "loss": 5.885,
      "mean_token_accuracy": 0.1343015268445015,
      "num_tokens": 5359210.0,
      "step": 2435
    },
    {
      "entropy": 5.892851829528809,
      "epoch": 0.24215958713775307,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997689671636926,
      "loss": 5.8994,
      "mean_token_accuracy": 0.11987031921744347,
      "num_tokens": 5369891.0,
      "step": 2440
    },
    {
      "entropy": 5.88413896560669,
      "epoch": 0.2426558157999206,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997673591419884,
      "loss": 5.8828,
      "mean_token_accuracy": 0.1243226870894432,
      "num_tokens": 5380850.0,
      "step": 2445
    },
    {
      "entropy": 5.868440723419189,
      "epoch": 0.24315204446208813,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997657455465381,
      "loss": 5.9584,
      "mean_token_accuracy": 0.1275911472737789,
      "num_tokens": 5391903.0,
      "step": 2450
    },
    {
      "entropy": 5.989337539672851,
      "epoch": 0.24364827312425566,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997641263773815,
      "loss": 5.9994,
      "mean_token_accuracy": 0.12448158636689186,
      "num_tokens": 5402517.0,
      "step": 2455
    },
    {
      "entropy": 5.891087675094605,
      "epoch": 0.2441445017864232,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499762501634559,
      "loss": 5.9118,
      "mean_token_accuracy": 0.12482894137501717,
      "num_tokens": 5414110.0,
      "step": 2460
    },
    {
      "entropy": 5.962598514556885,
      "epoch": 0.24464073044859072,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997608713181108,
      "loss": 5.9243,
      "mean_token_accuracy": 0.1277947887778282,
      "num_tokens": 5424466.0,
      "step": 2465
    },
    {
      "entropy": 5.984759044647217,
      "epoch": 0.24513695911075822,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997592354280772,
      "loss": 5.9161,
      "mean_token_accuracy": 0.13129449263215065,
      "num_tokens": 5434461.0,
      "step": 2470
    },
    {
      "entropy": 5.84287428855896,
      "epoch": 0.24563318777292575,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499757593964499,
      "loss": 5.8991,
      "mean_token_accuracy": 0.1189004898071289,
      "num_tokens": 5445921.0,
      "step": 2475
    },
    {
      "entropy": 5.9045604228973385,
      "epoch": 0.24612941643509328,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997559469274166,
      "loss": 5.8835,
      "mean_token_accuracy": 0.12377197816967964,
      "num_tokens": 5456283.0,
      "step": 2480
    },
    {
      "entropy": 5.9440374851226805,
      "epoch": 0.24662564509726081,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997542943168712,
      "loss": 5.9296,
      "mean_token_accuracy": 0.12799091935157775,
      "num_tokens": 5467616.0,
      "step": 2485
    },
    {
      "entropy": 5.940678691864013,
      "epoch": 0.24712187375942835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997526361329035,
      "loss": 5.879,
      "mean_token_accuracy": 0.13008256927132605,
      "num_tokens": 5478050.0,
      "step": 2490
    },
    {
      "entropy": 5.970621490478516,
      "epoch": 0.24761810242159588,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997509723755547,
      "loss": 5.8866,
      "mean_token_accuracy": 0.12871679589152335,
      "num_tokens": 5487185.0,
      "step": 2495
    },
    {
      "entropy": 5.921407842636109,
      "epoch": 0.2481143310837634,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997493030448662,
      "loss": 5.9162,
      "mean_token_accuracy": 0.12141515985131264,
      "num_tokens": 5498802.0,
      "step": 2500
    },
    {
      "entropy": 5.850944328308105,
      "epoch": 0.24861055974593094,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997476281408793,
      "loss": 5.9169,
      "mean_token_accuracy": 0.11781499832868576,
      "num_tokens": 5510285.0,
      "step": 2505
    },
    {
      "entropy": 5.905643796920776,
      "epoch": 0.24910678840809844,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997459476636354,
      "loss": 5.8152,
      "mean_token_accuracy": 0.13310370296239854,
      "num_tokens": 5520028.0,
      "step": 2510
    },
    {
      "entropy": 5.869578409194946,
      "epoch": 0.24960301707026597,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997442616131763,
      "loss": 5.9188,
      "mean_token_accuracy": 0.12631641551852227,
      "num_tokens": 5529271.0,
      "step": 2515
    },
    {
      "entropy": 5.923573017120361,
      "epoch": 0.25009924573243353,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997425699895438,
      "loss": 5.8464,
      "mean_token_accuracy": 0.12032231464982032,
      "num_tokens": 5539396.0,
      "step": 2520
    },
    {
      "entropy": 5.949633979797364,
      "epoch": 0.25059547439460106,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997408727927798,
      "loss": 5.998,
      "mean_token_accuracy": 0.12382508367300034,
      "num_tokens": 5551674.0,
      "step": 2525
    },
    {
      "entropy": 5.928415966033936,
      "epoch": 0.25109170305676853,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997391700229264,
      "loss": 5.9155,
      "mean_token_accuracy": 0.12471647784113885,
      "num_tokens": 5562749.0,
      "step": 2530
    },
    {
      "entropy": 5.931261014938355,
      "epoch": 0.25158793171893606,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499737461680026,
      "loss": 5.9329,
      "mean_token_accuracy": 0.12794146463274955,
      "num_tokens": 5573478.0,
      "step": 2535
    },
    {
      "entropy": 5.888597440719605,
      "epoch": 0.2520841603811036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997357477641207,
      "loss": 5.9059,
      "mean_token_accuracy": 0.1306145206093788,
      "num_tokens": 5584546.0,
      "step": 2540
    },
    {
      "entropy": 5.862578821182251,
      "epoch": 0.2525803890432711,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997340282752532,
      "loss": 5.7961,
      "mean_token_accuracy": 0.12940113618969917,
      "num_tokens": 5594004.0,
      "step": 2545
    },
    {
      "entropy": 5.944779109954834,
      "epoch": 0.25307661770543866,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499732303213466,
      "loss": 5.8567,
      "mean_token_accuracy": 0.1197636641561985,
      "num_tokens": 5606095.0,
      "step": 2550
    },
    {
      "entropy": 5.916587829589844,
      "epoch": 0.2535728463676062,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004997305725788019,
      "loss": 5.8908,
      "mean_token_accuracy": 0.12375114262104034,
      "num_tokens": 5615596.0,
      "step": 2555
    },
    {
      "entropy": 5.87527642250061,
      "epoch": 0.2540690750297737,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499728836371304,
      "loss": 5.8512,
      "mean_token_accuracy": 0.12574743628501892,
      "num_tokens": 5626753.0,
      "step": 2560
    },
    {
      "entropy": 5.84861912727356,
      "epoch": 0.25456530369194125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997270945910151,
      "loss": 5.8587,
      "mean_token_accuracy": 0.12392849400639534,
      "num_tokens": 5636135.0,
      "step": 2565
    },
    {
      "entropy": 5.922877359390259,
      "epoch": 0.2550615323541088,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997253472379785,
      "loss": 5.9171,
      "mean_token_accuracy": 0.11965455114841461,
      "num_tokens": 5647074.0,
      "step": 2570
    },
    {
      "entropy": 5.952490043640137,
      "epoch": 0.2555577610162763,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997235943122375,
      "loss": 5.9575,
      "mean_token_accuracy": 0.11915252357721329,
      "num_tokens": 5658600.0,
      "step": 2575
    },
    {
      "entropy": 5.838838052749634,
      "epoch": 0.25605398967844384,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997218358138356,
      "loss": 5.8292,
      "mean_token_accuracy": 0.12637119889259338,
      "num_tokens": 5668453.0,
      "step": 2580
    },
    {
      "entropy": 5.9623559474945065,
      "epoch": 0.25655021834061137,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997200717428165,
      "loss": 5.9026,
      "mean_token_accuracy": 0.12589198648929595,
      "num_tokens": 5678855.0,
      "step": 2585
    },
    {
      "entropy": 5.920707082748413,
      "epoch": 0.2570464470027789,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997183020992237,
      "loss": 5.9657,
      "mean_token_accuracy": 0.11761240959167481,
      "num_tokens": 5690063.0,
      "step": 2590
    },
    {
      "entropy": 5.885443162918091,
      "epoch": 0.25754267566494643,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997165268831014,
      "loss": 5.8649,
      "mean_token_accuracy": 0.12652077302336692,
      "num_tokens": 5701487.0,
      "step": 2595
    },
    {
      "entropy": 5.859537076950073,
      "epoch": 0.25803890432711396,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997147460944933,
      "loss": 5.8464,
      "mean_token_accuracy": 0.12707249894738198,
      "num_tokens": 5713034.0,
      "step": 2600
    },
    {
      "entropy": 5.949060773849487,
      "epoch": 0.25853513298928144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997129597334438,
      "loss": 5.8933,
      "mean_token_accuracy": 0.12450260668992996,
      "num_tokens": 5725469.0,
      "step": 2605
    },
    {
      "entropy": 5.900791788101197,
      "epoch": 0.25903136165144897,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997111677999971,
      "loss": 5.9001,
      "mean_token_accuracy": 0.12473684847354889,
      "num_tokens": 5737099.0,
      "step": 2610
    },
    {
      "entropy": 5.959123134613037,
      "epoch": 0.2595275903136165,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997093702941976,
      "loss": 5.9027,
      "mean_token_accuracy": 0.12286011129617691,
      "num_tokens": 5748756.0,
      "step": 2615
    },
    {
      "entropy": 5.9113915920257565,
      "epoch": 0.260023818975784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049970756721609,
      "loss": 5.8852,
      "mean_token_accuracy": 0.129507152736187,
      "num_tokens": 5759598.0,
      "step": 2620
    },
    {
      "entropy": 5.901254463195801,
      "epoch": 0.26052004763795156,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499705758565719,
      "loss": 5.9362,
      "mean_token_accuracy": 0.12127219662070274,
      "num_tokens": 5770397.0,
      "step": 2625
    },
    {
      "entropy": 5.990640354156494,
      "epoch": 0.2610162763001191,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997039443431292,
      "loss": 6.0079,
      "mean_token_accuracy": 0.11671988666057587,
      "num_tokens": 5780935.0,
      "step": 2630
    },
    {
      "entropy": 5.943032073974609,
      "epoch": 0.2615125049622866,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997021245483659,
      "loss": 5.8759,
      "mean_token_accuracy": 0.12639233022928237,
      "num_tokens": 5791557.0,
      "step": 2635
    },
    {
      "entropy": 5.829169988632202,
      "epoch": 0.26200873362445415,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499700299181474,
      "loss": 5.8935,
      "mean_token_accuracy": 0.1263813316822052,
      "num_tokens": 5802233.0,
      "step": 2640
    },
    {
      "entropy": 5.879264640808105,
      "epoch": 0.2625049622866217,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996984682424989,
      "loss": 5.8335,
      "mean_token_accuracy": 0.13040376976132392,
      "num_tokens": 5813424.0,
      "step": 2645
    },
    {
      "entropy": 5.876601123809815,
      "epoch": 0.2630011909487892,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499696631731486,
      "loss": 5.8084,
      "mean_token_accuracy": 0.12445717677474022,
      "num_tokens": 5825652.0,
      "step": 2650
    },
    {
      "entropy": 5.923290538787842,
      "epoch": 0.26349741961095674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996947896484809,
      "loss": 5.9033,
      "mean_token_accuracy": 0.12721100226044654,
      "num_tokens": 5836603.0,
      "step": 2655
    },
    {
      "entropy": 5.893552494049072,
      "epoch": 0.26399364827312427,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499692941993529,
      "loss": 5.9528,
      "mean_token_accuracy": 0.12557620778679848,
      "num_tokens": 5849365.0,
      "step": 2660
    },
    {
      "entropy": 5.9200966358184814,
      "epoch": 0.2644898769352918,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996910887666765,
      "loss": 5.8669,
      "mean_token_accuracy": 0.12856241166591645,
      "num_tokens": 5860305.0,
      "step": 2665
    },
    {
      "entropy": 5.83866138458252,
      "epoch": 0.26498610559745933,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499689229967969,
      "loss": 5.8055,
      "mean_token_accuracy": 0.12650213465094567,
      "num_tokens": 5872272.0,
      "step": 2670
    },
    {
      "entropy": 5.881111478805542,
      "epoch": 0.26548233425962686,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996873655974529,
      "loss": 5.8348,
      "mean_token_accuracy": 0.12477733865380287,
      "num_tokens": 5884107.0,
      "step": 2675
    },
    {
      "entropy": 5.749620342254639,
      "epoch": 0.26597856292179434,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996854956551743,
      "loss": 5.8186,
      "mean_token_accuracy": 0.1303606040775776,
      "num_tokens": 5894060.0,
      "step": 2680
    },
    {
      "entropy": 5.881046438217163,
      "epoch": 0.26647479158396187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996836201411794,
      "loss": 5.8581,
      "mean_token_accuracy": 0.1279796727001667,
      "num_tokens": 5904804.0,
      "step": 2685
    },
    {
      "entropy": 5.954271507263184,
      "epoch": 0.2669710202461294,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499681739055515,
      "loss": 5.9239,
      "mean_token_accuracy": 0.12369774505496026,
      "num_tokens": 5916496.0,
      "step": 2690
    },
    {
      "entropy": 5.884972763061524,
      "epoch": 0.26746724890829693,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996798523982276,
      "loss": 5.8601,
      "mean_token_accuracy": 0.12348931431770324,
      "num_tokens": 5926920.0,
      "step": 2695
    },
    {
      "entropy": 5.999528169631958,
      "epoch": 0.26796347757046446,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499677960169364,
      "loss": 5.9731,
      "mean_token_accuracy": 0.12307530716061592,
      "num_tokens": 5940098.0,
      "step": 2700
    },
    {
      "entropy": 5.873397493362427,
      "epoch": 0.268459706232632,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996760623689711,
      "loss": 5.8108,
      "mean_token_accuracy": 0.12795935273170472,
      "num_tokens": 5949687.0,
      "step": 2705
    },
    {
      "entropy": 5.904886293411255,
      "epoch": 0.2689559348947995,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499674158997096,
      "loss": 5.9344,
      "mean_token_accuracy": 0.12130291312932968,
      "num_tokens": 5960777.0,
      "step": 2710
    },
    {
      "entropy": 5.976747798919678,
      "epoch": 0.26945216355696705,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996722500537859,
      "loss": 5.8323,
      "mean_token_accuracy": 0.12468361780047417,
      "num_tokens": 5972298.0,
      "step": 2715
    },
    {
      "entropy": 5.831518363952637,
      "epoch": 0.2699483922191346,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996703355390881,
      "loss": 5.9595,
      "mean_token_accuracy": 0.11904360502958297,
      "num_tokens": 5985613.0,
      "step": 2720
    },
    {
      "entropy": 5.997792959213257,
      "epoch": 0.2704446208813021,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049966841545305,
      "loss": 6.0036,
      "mean_token_accuracy": 0.12301708981394768,
      "num_tokens": 5997530.0,
      "step": 2725
    },
    {
      "entropy": 5.925682544708252,
      "epoch": 0.27094084954346964,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996664897957195,
      "loss": 5.8757,
      "mean_token_accuracy": 0.12393735498189926,
      "num_tokens": 6008123.0,
      "step": 2730
    },
    {
      "entropy": 5.934986400604248,
      "epoch": 0.2714370782056372,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996645585671441,
      "loss": 5.8868,
      "mean_token_accuracy": 0.1257225215435028,
      "num_tokens": 6020578.0,
      "step": 2735
    },
    {
      "entropy": 5.876024007797241,
      "epoch": 0.2719333068678047,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996626217673717,
      "loss": 5.816,
      "mean_token_accuracy": 0.13031245917081832,
      "num_tokens": 6031625.0,
      "step": 2740
    },
    {
      "entropy": 5.788500356674194,
      "epoch": 0.27242953552997223,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996606793964504,
      "loss": 5.8589,
      "mean_token_accuracy": 0.12932414785027505,
      "num_tokens": 6041320.0,
      "step": 2745
    },
    {
      "entropy": 5.9028442859649655,
      "epoch": 0.27292576419213976,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996587314544284,
      "loss": 5.8742,
      "mean_token_accuracy": 0.123659198731184,
      "num_tokens": 6052627.0,
      "step": 2750
    },
    {
      "entropy": 5.916861200332642,
      "epoch": 0.27342199285430724,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996567779413539,
      "loss": 5.8286,
      "mean_token_accuracy": 0.1305013693869114,
      "num_tokens": 6063468.0,
      "step": 2755
    },
    {
      "entropy": 5.808893394470215,
      "epoch": 0.27391822151647477,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004996548188572755,
      "loss": 5.8031,
      "mean_token_accuracy": 0.13299782574176788,
      "num_tokens": 6074208.0,
      "step": 2760
    },
    {
      "entropy": 5.939206218719482,
      "epoch": 0.2744144501786423,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996528542022416,
      "loss": 5.8706,
      "mean_token_accuracy": 0.12870003804564475,
      "num_tokens": 6085068.0,
      "step": 2765
    },
    {
      "entropy": 5.880002784729004,
      "epoch": 0.27491067884080983,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499650883976301,
      "loss": 5.8617,
      "mean_token_accuracy": 0.12842236310243607,
      "num_tokens": 6096445.0,
      "step": 2770
    },
    {
      "entropy": 5.792717790603637,
      "epoch": 0.27540690750297736,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996489081795026,
      "loss": 5.77,
      "mean_token_accuracy": 0.1276902064681053,
      "num_tokens": 6106527.0,
      "step": 2775
    },
    {
      "entropy": 5.80667667388916,
      "epoch": 0.2759031361651449,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004996469268118953,
      "loss": 5.7975,
      "mean_token_accuracy": 0.13269607350230217,
      "num_tokens": 6115923.0,
      "step": 2780
    },
    {
      "entropy": 5.800922632217407,
      "epoch": 0.2763993648273124,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004996449398735283,
      "loss": 5.6777,
      "mean_token_accuracy": 0.13338218480348588,
      "num_tokens": 6125861.0,
      "step": 2785
    },
    {
      "entropy": 5.881807422637939,
      "epoch": 0.27689559348947995,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996429473644509,
      "loss": 5.8893,
      "mean_token_accuracy": 0.1284797951579094,
      "num_tokens": 6137189.0,
      "step": 2790
    },
    {
      "entropy": 5.898675632476807,
      "epoch": 0.2773918221516475,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996409492847124,
      "loss": 5.8179,
      "mean_token_accuracy": 0.12720710635185242,
      "num_tokens": 6147920.0,
      "step": 2795
    },
    {
      "entropy": 5.831118535995484,
      "epoch": 0.277888050813815,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996389456343624,
      "loss": 5.8105,
      "mean_token_accuracy": 0.13541241586208344,
      "num_tokens": 6159371.0,
      "step": 2800
    },
    {
      "entropy": 5.961232089996338,
      "epoch": 0.27838427947598254,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996369364134507,
      "loss": 6.0155,
      "mean_token_accuracy": 0.1189655639231205,
      "num_tokens": 6172679.0,
      "step": 2805
    },
    {
      "entropy": 5.847831535339355,
      "epoch": 0.2788805081381501,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499634921622027,
      "loss": 5.8821,
      "mean_token_accuracy": 0.12079857736825943,
      "num_tokens": 6184244.0,
      "step": 2810
    },
    {
      "entropy": 5.881469202041626,
      "epoch": 0.2793767368003176,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996329012601413,
      "loss": 5.7422,
      "mean_token_accuracy": 0.12676719948649406,
      "num_tokens": 6194826.0,
      "step": 2815
    },
    {
      "entropy": 5.890796756744384,
      "epoch": 0.27987296546248513,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996308753278437,
      "loss": 5.8708,
      "mean_token_accuracy": 0.1250182457268238,
      "num_tokens": 6206808.0,
      "step": 2820
    },
    {
      "entropy": 5.92862229347229,
      "epoch": 0.28036919412465267,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996288438251843,
      "loss": 5.8984,
      "mean_token_accuracy": 0.12058991715312004,
      "num_tokens": 6217303.0,
      "step": 2825
    },
    {
      "entropy": 5.912028837203979,
      "epoch": 0.28086542278682014,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996268067522137,
      "loss": 5.8827,
      "mean_token_accuracy": 0.12254092544317245,
      "num_tokens": 6227965.0,
      "step": 2830
    },
    {
      "entropy": 5.92435474395752,
      "epoch": 0.28136165144898767,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004996247641089823,
      "loss": 5.9517,
      "mean_token_accuracy": 0.12450585365295411,
      "num_tokens": 6239011.0,
      "step": 2835
    },
    {
      "entropy": 5.813734006881714,
      "epoch": 0.2818578801111552,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996227158955409,
      "loss": 5.7928,
      "mean_token_accuracy": 0.1279340922832489,
      "num_tokens": 6249084.0,
      "step": 2840
    },
    {
      "entropy": 5.905808782577514,
      "epoch": 0.28235410877332273,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049962066211194,
      "loss": 5.8718,
      "mean_token_accuracy": 0.1256870672106743,
      "num_tokens": 6260706.0,
      "step": 2845
    },
    {
      "entropy": 5.821885061264038,
      "epoch": 0.28285033743549026,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996186027582308,
      "loss": 5.7954,
      "mean_token_accuracy": 0.13122305497527123,
      "num_tokens": 6272848.0,
      "step": 2850
    },
    {
      "entropy": 5.886135005950928,
      "epoch": 0.2833465660976578,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996165378344642,
      "loss": 5.8119,
      "mean_token_accuracy": 0.1294620506465435,
      "num_tokens": 6283820.0,
      "step": 2855
    },
    {
      "entropy": 5.8642865180969235,
      "epoch": 0.2838427947598253,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996144673406914,
      "loss": 5.971,
      "mean_token_accuracy": 0.11893427968025208,
      "num_tokens": 6294193.0,
      "step": 2860
    },
    {
      "entropy": 5.916045618057251,
      "epoch": 0.28433902342199285,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996123912769639,
      "loss": 5.8113,
      "mean_token_accuracy": 0.13239249885082244,
      "num_tokens": 6304407.0,
      "step": 2865
    },
    {
      "entropy": 5.837239408493042,
      "epoch": 0.2848352520841604,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996103096433332,
      "loss": 5.8682,
      "mean_token_accuracy": 0.1274930603802204,
      "num_tokens": 6316311.0,
      "step": 2870
    },
    {
      "entropy": 5.894733238220215,
      "epoch": 0.2853314807463279,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996082224398506,
      "loss": 5.8508,
      "mean_token_accuracy": 0.12729406282305716,
      "num_tokens": 6327248.0,
      "step": 2875
    },
    {
      "entropy": 5.829097318649292,
      "epoch": 0.28582770940849545,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996061296665682,
      "loss": 5.8571,
      "mean_token_accuracy": 0.12545564025640488,
      "num_tokens": 6338494.0,
      "step": 2880
    },
    {
      "entropy": 5.894459867477417,
      "epoch": 0.286323938070663,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004996040313235377,
      "loss": 5.7517,
      "mean_token_accuracy": 0.12818305119872092,
      "num_tokens": 6349038.0,
      "step": 2885
    },
    {
      "entropy": 5.864468955993653,
      "epoch": 0.2868201667328305,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996019274108112,
      "loss": 5.8608,
      "mean_token_accuracy": 0.12657166197896003,
      "num_tokens": 6360180.0,
      "step": 2890
    },
    {
      "entropy": 5.771185398101807,
      "epoch": 0.28731639539499804,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499599817928441,
      "loss": 5.8005,
      "mean_token_accuracy": 0.12896572723984717,
      "num_tokens": 6370931.0,
      "step": 2895
    },
    {
      "entropy": 5.833103847503662,
      "epoch": 0.28781262405716557,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995977028764792,
      "loss": 5.7767,
      "mean_token_accuracy": 0.1355249285697937,
      "num_tokens": 6381810.0,
      "step": 2900
    },
    {
      "entropy": 5.802249383926392,
      "epoch": 0.28830885271933304,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995955822549783,
      "loss": 5.7855,
      "mean_token_accuracy": 0.12975866422057153,
      "num_tokens": 6392268.0,
      "step": 2905
    },
    {
      "entropy": 5.865897560119629,
      "epoch": 0.2888050813815006,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995934560639909,
      "loss": 5.8417,
      "mean_token_accuracy": 0.13064575865864753,
      "num_tokens": 6403650.0,
      "step": 2910
    },
    {
      "entropy": 5.859321403503418,
      "epoch": 0.2893013100436681,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995913243035699,
      "loss": 5.8183,
      "mean_token_accuracy": 0.12550777047872544,
      "num_tokens": 6414672.0,
      "step": 2915
    },
    {
      "entropy": 5.863259506225586,
      "epoch": 0.28979753870583563,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995891869737678,
      "loss": 5.8909,
      "mean_token_accuracy": 0.12621842250227927,
      "num_tokens": 6424916.0,
      "step": 2920
    },
    {
      "entropy": 5.932774877548217,
      "epoch": 0.29029376736800316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995870440746379,
      "loss": 5.8467,
      "mean_token_accuracy": 0.13387326076626777,
      "num_tokens": 6435950.0,
      "step": 2925
    },
    {
      "entropy": 5.828493309020996,
      "epoch": 0.2907899960301707,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995848956062333,
      "loss": 5.8022,
      "mean_token_accuracy": 0.12572760954499246,
      "num_tokens": 6446336.0,
      "step": 2930
    },
    {
      "entropy": 5.844927358627319,
      "epoch": 0.2912862246923382,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995827415686071,
      "loss": 5.8729,
      "mean_token_accuracy": 0.12402887716889381,
      "num_tokens": 6457655.0,
      "step": 2935
    },
    {
      "entropy": 5.882878684997559,
      "epoch": 0.29178245335450576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499580581961813,
      "loss": 5.7676,
      "mean_token_accuracy": 0.1295201987028122,
      "num_tokens": 6469627.0,
      "step": 2940
    },
    {
      "entropy": 5.810547113418579,
      "epoch": 0.2922786820166733,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995784167859043,
      "loss": 5.8584,
      "mean_token_accuracy": 0.12580792158842086,
      "num_tokens": 6480351.0,
      "step": 2945
    },
    {
      "entropy": 5.904936742782593,
      "epoch": 0.2927749106788408,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995762460409349,
      "loss": 5.8915,
      "mean_token_accuracy": 0.1285228781402111,
      "num_tokens": 6491651.0,
      "step": 2950
    },
    {
      "entropy": 5.824595022201538,
      "epoch": 0.29327113934100835,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995740697269584,
      "loss": 5.7323,
      "mean_token_accuracy": 0.1294215962290764,
      "num_tokens": 6502531.0,
      "step": 2955
    },
    {
      "entropy": 5.748151111602783,
      "epoch": 0.2937673680031759,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995718878440289,
      "loss": 5.7433,
      "mean_token_accuracy": 0.13419339433312416,
      "num_tokens": 6511938.0,
      "step": 2960
    },
    {
      "entropy": 5.907798385620117,
      "epoch": 0.2942635966653434,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004995697003922006,
      "loss": 5.9029,
      "mean_token_accuracy": 0.12496517151594162,
      "num_tokens": 6522775.0,
      "step": 2965
    },
    {
      "entropy": 5.857905149459839,
      "epoch": 0.29475982532751094,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995675073715274,
      "loss": 5.731,
      "mean_token_accuracy": 0.12566524893045425,
      "num_tokens": 6534660.0,
      "step": 2970
    },
    {
      "entropy": 5.825377321243286,
      "epoch": 0.29525605398967847,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004995653087820641,
      "loss": 5.8841,
      "mean_token_accuracy": 0.12718252837657928,
      "num_tokens": 6545673.0,
      "step": 2975
    },
    {
      "entropy": 5.906738996505737,
      "epoch": 0.29575228265184594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499563104623865,
      "loss": 5.8274,
      "mean_token_accuracy": 0.12779038175940513,
      "num_tokens": 6556912.0,
      "step": 2980
    },
    {
      "entropy": 5.861005258560181,
      "epoch": 0.2962485113140135,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995608948969849,
      "loss": 5.9059,
      "mean_token_accuracy": 0.12463748157024383,
      "num_tokens": 6567532.0,
      "step": 2985
    },
    {
      "entropy": 5.91699070930481,
      "epoch": 0.296744739976181,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004995586796014784,
      "loss": 5.8914,
      "mean_token_accuracy": 0.12917492017149926,
      "num_tokens": 6577830.0,
      "step": 2990
    },
    {
      "entropy": 5.8236791610717775,
      "epoch": 0.29724096863834853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995564587374005,
      "loss": 5.768,
      "mean_token_accuracy": 0.12926611825823783,
      "num_tokens": 6587254.0,
      "step": 2995
    },
    {
      "entropy": 5.8685643672943115,
      "epoch": 0.29773719730051607,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004995542323048063,
      "loss": 5.9674,
      "mean_token_accuracy": 0.12186386585235595,
      "num_tokens": 6598294.0,
      "step": 3000
    },
    {
      "epoch": 0.29773719730051607,
      "eval_entropy": 5.765697889244168,
      "eval_loss": 5.914527416229248,
      "eval_mean_token_accuracy": 0.12812341096967877,
      "eval_num_tokens": 6598294.0,
      "eval_runtime": 28.0358,
      "eval_samples_per_second": 1200.929,
      "eval_steps_per_second": 150.129,
      "step": 3000
    },
    {
      "entropy": 5.939652395248413,
      "epoch": 0.2982334259626836,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499552000303751,
      "loss": 5.8421,
      "mean_token_accuracy": 0.1334158658981323,
      "num_tokens": 6608772.0,
      "step": 3005
    },
    {
      "entropy": 5.87982029914856,
      "epoch": 0.2987296546248511,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049954976273429,
      "loss": 5.8505,
      "mean_token_accuracy": 0.12715167179703712,
      "num_tokens": 6619973.0,
      "step": 3010
    },
    {
      "entropy": 5.854435682296753,
      "epoch": 0.29922588328701866,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004995475195964787,
      "loss": 5.8763,
      "mean_token_accuracy": 0.12607175037264823,
      "num_tokens": 6631172.0,
      "step": 3015
    },
    {
      "entropy": 5.83586630821228,
      "epoch": 0.2997221119491862,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995452708903727,
      "loss": 5.8563,
      "mean_token_accuracy": 0.12598838657140732,
      "num_tokens": 6641343.0,
      "step": 3020
    },
    {
      "entropy": 5.88385534286499,
      "epoch": 0.3002183406113537,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499543016616028,
      "loss": 5.812,
      "mean_token_accuracy": 0.12634400129318238,
      "num_tokens": 6652892.0,
      "step": 3025
    },
    {
      "entropy": 5.875179481506348,
      "epoch": 0.30071456927352125,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995407567735002,
      "loss": 5.8321,
      "mean_token_accuracy": 0.12487765029072762,
      "num_tokens": 6664177.0,
      "step": 3030
    },
    {
      "entropy": 5.828985166549683,
      "epoch": 0.3012107979356888,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995384913628456,
      "loss": 5.7954,
      "mean_token_accuracy": 0.1299532249569893,
      "num_tokens": 6674402.0,
      "step": 3035
    },
    {
      "entropy": 5.9306165218353275,
      "epoch": 0.3017070265978563,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995362203841202,
      "loss": 5.9226,
      "mean_token_accuracy": 0.12396132349967956,
      "num_tokens": 6684886.0,
      "step": 3040
    },
    {
      "entropy": 5.903984403610229,
      "epoch": 0.30220325526002384,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995339438373804,
      "loss": 5.7556,
      "mean_token_accuracy": 0.13111717849969864,
      "num_tokens": 6694774.0,
      "step": 3045
    },
    {
      "entropy": 5.756884288787842,
      "epoch": 0.30269948392219137,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995316617226825,
      "loss": 5.8306,
      "mean_token_accuracy": 0.13095860034227372,
      "num_tokens": 6705346.0,
      "step": 3050
    },
    {
      "entropy": 5.870118427276611,
      "epoch": 0.30319571258435885,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995293740400833,
      "loss": 5.7787,
      "mean_token_accuracy": 0.130900689214468,
      "num_tokens": 6715579.0,
      "step": 3055
    },
    {
      "entropy": 5.776624917984009,
      "epoch": 0.3036919412465264,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995270807896395,
      "loss": 5.7291,
      "mean_token_accuracy": 0.13014699071645736,
      "num_tokens": 6726382.0,
      "step": 3060
    },
    {
      "entropy": 5.851511716842651,
      "epoch": 0.3041881699086939,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995247819714079,
      "loss": 5.8066,
      "mean_token_accuracy": 0.12530786246061326,
      "num_tokens": 6737245.0,
      "step": 3065
    },
    {
      "entropy": 5.8999974727630615,
      "epoch": 0.30468439857086144,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004995224775854454,
      "loss": 6.0163,
      "mean_token_accuracy": 0.11953989639878274,
      "num_tokens": 6749113.0,
      "step": 3070
    },
    {
      "entropy": 5.955473518371582,
      "epoch": 0.30518062723302897,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995201676318093,
      "loss": 5.8645,
      "mean_token_accuracy": 0.12974076643586158,
      "num_tokens": 6761987.0,
      "step": 3075
    },
    {
      "entropy": 5.858785009384155,
      "epoch": 0.3056768558951965,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995178521105568,
      "loss": 5.8002,
      "mean_token_accuracy": 0.12481995448470115,
      "num_tokens": 6771279.0,
      "step": 3080
    },
    {
      "entropy": 5.815781879425049,
      "epoch": 0.30617308455736403,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995155310217454,
      "loss": 5.74,
      "mean_token_accuracy": 0.13142770677804946,
      "num_tokens": 6781038.0,
      "step": 3085
    },
    {
      "entropy": 5.793752861022949,
      "epoch": 0.30666931321953156,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995132043654327,
      "loss": 5.891,
      "mean_token_accuracy": 0.12459158003330231,
      "num_tokens": 6792686.0,
      "step": 3090
    },
    {
      "entropy": 5.881446695327758,
      "epoch": 0.3071655418816991,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995108721416763,
      "loss": 5.8433,
      "mean_token_accuracy": 0.1327853061258793,
      "num_tokens": 6804414.0,
      "step": 3095
    },
    {
      "entropy": 5.88617844581604,
      "epoch": 0.3076617705438666,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995085343505339,
      "loss": 5.8145,
      "mean_token_accuracy": 0.13131516203284263,
      "num_tokens": 6816180.0,
      "step": 3100
    },
    {
      "entropy": 5.8470758438110355,
      "epoch": 0.30815799920603415,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995061909920638,
      "loss": 5.8143,
      "mean_token_accuracy": 0.12712864801287652,
      "num_tokens": 6827114.0,
      "step": 3105
    },
    {
      "entropy": 5.876535511016845,
      "epoch": 0.3086542278682017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995038420663238,
      "loss": 5.8691,
      "mean_token_accuracy": 0.12930315136909484,
      "num_tokens": 6836953.0,
      "step": 3110
    },
    {
      "entropy": 5.864369058609009,
      "epoch": 0.3091504565303692,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995014875733724,
      "loss": 5.9164,
      "mean_token_accuracy": 0.12485591024160385,
      "num_tokens": 6847963.0,
      "step": 3115
    },
    {
      "entropy": 5.961886978149414,
      "epoch": 0.30964668519253674,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994991275132678,
      "loss": 5.9278,
      "mean_token_accuracy": 0.12286274209618568,
      "num_tokens": 6858264.0,
      "step": 3120
    },
    {
      "entropy": 5.879438591003418,
      "epoch": 0.31014291385470427,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994967618860685,
      "loss": 5.811,
      "mean_token_accuracy": 0.13317865282297134,
      "num_tokens": 6868861.0,
      "step": 3125
    },
    {
      "entropy": 5.873682880401612,
      "epoch": 0.31063914251687175,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994943906918333,
      "loss": 5.8327,
      "mean_token_accuracy": 0.13562217727303505,
      "num_tokens": 6879797.0,
      "step": 3130
    },
    {
      "entropy": 5.850006103515625,
      "epoch": 0.3111353711790393,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994920139306211,
      "loss": 5.7891,
      "mean_token_accuracy": 0.12811487391591073,
      "num_tokens": 6891323.0,
      "step": 3135
    },
    {
      "entropy": 5.790000009536743,
      "epoch": 0.3116315998412068,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004994896316024906,
      "loss": 5.8779,
      "mean_token_accuracy": 0.12558741196990014,
      "num_tokens": 6903345.0,
      "step": 3140
    },
    {
      "entropy": 5.798595714569092,
      "epoch": 0.31212782850337434,
      "grad_norm": 1.25,
      "learning_rate": 0.000499487243707501,
      "loss": 5.8054,
      "mean_token_accuracy": 0.1267201617360115,
      "num_tokens": 6914342.0,
      "step": 3145
    },
    {
      "entropy": 5.8805296421051025,
      "epoch": 0.31262405716554187,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994848502457114,
      "loss": 5.8603,
      "mean_token_accuracy": 0.1250415749847889,
      "num_tokens": 6926691.0,
      "step": 3150
    },
    {
      "entropy": 5.84386978149414,
      "epoch": 0.3131202858277094,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994824512171812,
      "loss": 5.7316,
      "mean_token_accuracy": 0.13148539662361144,
      "num_tokens": 6937841.0,
      "step": 3155
    },
    {
      "entropy": 5.782042407989502,
      "epoch": 0.31361651448987693,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994800466219701,
      "loss": 5.8296,
      "mean_token_accuracy": 0.12499077171087265,
      "num_tokens": 6948601.0,
      "step": 3160
    },
    {
      "entropy": 5.89652795791626,
      "epoch": 0.31411274315204446,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994776364601374,
      "loss": 5.7555,
      "mean_token_accuracy": 0.12984561249613763,
      "num_tokens": 6959775.0,
      "step": 3165
    },
    {
      "entropy": 5.777927494049072,
      "epoch": 0.314608971814212,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994752207317431,
      "loss": 5.7795,
      "mean_token_accuracy": 0.13198756724596022,
      "num_tokens": 6969202.0,
      "step": 3170
    },
    {
      "entropy": 5.792110872268677,
      "epoch": 0.3151052004763795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499472799436847,
      "loss": 5.7636,
      "mean_token_accuracy": 0.130733972042799,
      "num_tokens": 6978633.0,
      "step": 3175
    },
    {
      "entropy": 5.91819577217102,
      "epoch": 0.31560142913854705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994703725755092,
      "loss": 5.8334,
      "mean_token_accuracy": 0.12615416571497917,
      "num_tokens": 6989638.0,
      "step": 3180
    },
    {
      "entropy": 5.792121648788452,
      "epoch": 0.3160976578007146,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994679401477899,
      "loss": 5.8248,
      "mean_token_accuracy": 0.1252963811159134,
      "num_tokens": 7001606.0,
      "step": 3185
    },
    {
      "entropy": 5.81984133720398,
      "epoch": 0.3165938864628821,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004994655021537492,
      "loss": 5.7664,
      "mean_token_accuracy": 0.12752437070012093,
      "num_tokens": 7012915.0,
      "step": 3190
    },
    {
      "entropy": 5.818609809875488,
      "epoch": 0.31709011512504964,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994630585934478,
      "loss": 5.8369,
      "mean_token_accuracy": 0.12405973374843597,
      "num_tokens": 7023191.0,
      "step": 3195
    },
    {
      "entropy": 5.827565288543701,
      "epoch": 0.3175863437872172,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994606094669463,
      "loss": 5.7335,
      "mean_token_accuracy": 0.1335385449230671,
      "num_tokens": 7032506.0,
      "step": 3200
    },
    {
      "entropy": 5.860045909881592,
      "epoch": 0.3180825724493847,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994581547743053,
      "loss": 5.8684,
      "mean_token_accuracy": 0.12564630582928657,
      "num_tokens": 7046098.0,
      "step": 3205
    },
    {
      "entropy": 5.919556093215943,
      "epoch": 0.3185788011115522,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994556945155857,
      "loss": 5.8868,
      "mean_token_accuracy": 0.1291634827852249,
      "num_tokens": 7057434.0,
      "step": 3210
    },
    {
      "entropy": 5.795089626312256,
      "epoch": 0.3190750297737197,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994532286908486,
      "loss": 5.8037,
      "mean_token_accuracy": 0.12871215268969535,
      "num_tokens": 7067674.0,
      "step": 3215
    },
    {
      "entropy": 5.929013442993164,
      "epoch": 0.31957125843588724,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499450757300155,
      "loss": 5.8511,
      "mean_token_accuracy": 0.12283276841044426,
      "num_tokens": 7077898.0,
      "step": 3220
    },
    {
      "entropy": 5.9026130676269535,
      "epoch": 0.32006748709805477,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004994482803435664,
      "loss": 5.8347,
      "mean_token_accuracy": 0.1284059502184391,
      "num_tokens": 7088566.0,
      "step": 3225
    },
    {
      "entropy": 5.760651159286499,
      "epoch": 0.3205637157602223,
      "grad_norm": 1.171875,
      "learning_rate": 0.000499445797821144,
      "loss": 5.8289,
      "mean_token_accuracy": 0.12847434878349304,
      "num_tokens": 7099533.0,
      "step": 3230
    },
    {
      "entropy": 5.871236896514892,
      "epoch": 0.32105994442238983,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994433097329494,
      "loss": 5.8282,
      "mean_token_accuracy": 0.1276480548083782,
      "num_tokens": 7112254.0,
      "step": 3235
    },
    {
      "entropy": 5.910565805435181,
      "epoch": 0.32155617308455736,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994408160790444,
      "loss": 5.8538,
      "mean_token_accuracy": 0.12448023408651351,
      "num_tokens": 7124704.0,
      "step": 3240
    },
    {
      "entropy": 5.965430355072021,
      "epoch": 0.3220524017467249,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994383168594907,
      "loss": 6.0342,
      "mean_token_accuracy": 0.11537200659513473,
      "num_tokens": 7137741.0,
      "step": 3245
    },
    {
      "entropy": 5.926316213607788,
      "epoch": 0.3225486304088924,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994358120743506,
      "loss": 5.7955,
      "mean_token_accuracy": 0.12387502938508987,
      "num_tokens": 7147888.0,
      "step": 3250
    },
    {
      "entropy": 5.7006324291229244,
      "epoch": 0.32304485907105995,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994333017236858,
      "loss": 5.7627,
      "mean_token_accuracy": 0.13563922122120858,
      "num_tokens": 7158950.0,
      "step": 3255
    },
    {
      "entropy": 5.894726276397705,
      "epoch": 0.3235410877332275,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994307858075587,
      "loss": 5.8963,
      "mean_token_accuracy": 0.1240608274936676,
      "num_tokens": 7170156.0,
      "step": 3260
    },
    {
      "entropy": 5.8563618659973145,
      "epoch": 0.324037316395395,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994282643260319,
      "loss": 5.7506,
      "mean_token_accuracy": 0.1301948718726635,
      "num_tokens": 7180576.0,
      "step": 3265
    },
    {
      "entropy": 5.751006937026977,
      "epoch": 0.32453354505756254,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994257372791676,
      "loss": 5.8139,
      "mean_token_accuracy": 0.12954910472035408,
      "num_tokens": 7192380.0,
      "step": 3270
    },
    {
      "entropy": 5.945131874084472,
      "epoch": 0.3250297737197301,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994232046670286,
      "loss": 5.873,
      "mean_token_accuracy": 0.12636842131614684,
      "num_tokens": 7203898.0,
      "step": 3275
    },
    {
      "entropy": 5.89789924621582,
      "epoch": 0.3255260023818976,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994206664896778,
      "loss": 5.8348,
      "mean_token_accuracy": 0.12694439068436622,
      "num_tokens": 7215241.0,
      "step": 3280
    },
    {
      "entropy": 5.839521884918213,
      "epoch": 0.3260222310440651,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994181227471779,
      "loss": 5.789,
      "mean_token_accuracy": 0.13032658770680428,
      "num_tokens": 7226039.0,
      "step": 3285
    },
    {
      "entropy": 5.782558107376099,
      "epoch": 0.3265184597062326,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994155734395923,
      "loss": 5.8054,
      "mean_token_accuracy": 0.13057178035378456,
      "num_tokens": 7237148.0,
      "step": 3290
    },
    {
      "entropy": 5.818170642852783,
      "epoch": 0.32701468836840014,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499413018566984,
      "loss": 5.8164,
      "mean_token_accuracy": 0.1280258573591709,
      "num_tokens": 7249284.0,
      "step": 3295
    },
    {
      "entropy": 5.887258100509643,
      "epoch": 0.32751091703056767,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994104581294163,
      "loss": 5.8128,
      "mean_token_accuracy": 0.12699150666594505,
      "num_tokens": 7259967.0,
      "step": 3300
    },
    {
      "entropy": 5.763713073730469,
      "epoch": 0.3280071456927352,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994078921269529,
      "loss": 5.7799,
      "mean_token_accuracy": 0.13115021362900733,
      "num_tokens": 7271145.0,
      "step": 3305
    },
    {
      "entropy": 5.8840430736541744,
      "epoch": 0.32850337435490273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994053205596573,
      "loss": 5.79,
      "mean_token_accuracy": 0.13236218318343163,
      "num_tokens": 7281872.0,
      "step": 3310
    },
    {
      "entropy": 5.802930545806885,
      "epoch": 0.32899960301707026,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994027434275934,
      "loss": 5.8181,
      "mean_token_accuracy": 0.12497121542692184,
      "num_tokens": 7293330.0,
      "step": 3315
    },
    {
      "entropy": 5.883807945251465,
      "epoch": 0.3294958316792378,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994001607308249,
      "loss": 5.8639,
      "mean_token_accuracy": 0.1230167992413044,
      "num_tokens": 7304187.0,
      "step": 3320
    },
    {
      "entropy": 5.88583402633667,
      "epoch": 0.3299920603414053,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499397572469416,
      "loss": 5.7709,
      "mean_token_accuracy": 0.12690152004361152,
      "num_tokens": 7314756.0,
      "step": 3325
    },
    {
      "entropy": 5.791846513748169,
      "epoch": 0.33048828900357285,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993949786434308,
      "loss": 5.8053,
      "mean_token_accuracy": 0.12929374650120734,
      "num_tokens": 7325711.0,
      "step": 3330
    },
    {
      "entropy": 5.783223104476929,
      "epoch": 0.3309845176657404,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993923792529337,
      "loss": 5.7293,
      "mean_token_accuracy": 0.13179031014442444,
      "num_tokens": 7335639.0,
      "step": 3335
    },
    {
      "entropy": 5.904985094070435,
      "epoch": 0.3314807463279079,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993897742979892,
      "loss": 5.8495,
      "mean_token_accuracy": 0.12467518821358681,
      "num_tokens": 7346621.0,
      "step": 3340
    },
    {
      "entropy": 5.883304119110107,
      "epoch": 0.33197697499007545,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993871637786616,
      "loss": 5.8213,
      "mean_token_accuracy": 0.12235530912876129,
      "num_tokens": 7358950.0,
      "step": 3345
    },
    {
      "entropy": 5.771950149536133,
      "epoch": 0.332473203652243,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499384547695016,
      "loss": 5.6827,
      "mean_token_accuracy": 0.1342015653848648,
      "num_tokens": 7369629.0,
      "step": 3350
    },
    {
      "entropy": 5.83166766166687,
      "epoch": 0.3329694323144105,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499381926047117,
      "loss": 5.7706,
      "mean_token_accuracy": 0.13085206598043442,
      "num_tokens": 7379853.0,
      "step": 3355
    },
    {
      "entropy": 5.797023868560791,
      "epoch": 0.333465660976578,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049937929883503,
      "loss": 5.8104,
      "mean_token_accuracy": 0.12401416078209877,
      "num_tokens": 7391323.0,
      "step": 3360
    },
    {
      "entropy": 5.817302465438843,
      "epoch": 0.3339618896387455,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004993766660588196,
      "loss": 5.7897,
      "mean_token_accuracy": 0.12517665177583695,
      "num_tokens": 7402201.0,
      "step": 3365
    },
    {
      "entropy": 5.912862777709961,
      "epoch": 0.33445811830091304,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993740277185515,
      "loss": 6.0276,
      "mean_token_accuracy": 0.11729341372847557,
      "num_tokens": 7413783.0,
      "step": 3370
    },
    {
      "entropy": 5.829028797149658,
      "epoch": 0.3349543469630806,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499371383814291,
      "loss": 5.7823,
      "mean_token_accuracy": 0.13103819414973258,
      "num_tokens": 7425909.0,
      "step": 3375
    },
    {
      "entropy": 5.774939775466919,
      "epoch": 0.3354505756252481,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993687343461036,
      "loss": 5.6859,
      "mean_token_accuracy": 0.13329387232661247,
      "num_tokens": 7436346.0,
      "step": 3380
    },
    {
      "entropy": 5.786150550842285,
      "epoch": 0.33594680428741563,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993660793140551,
      "loss": 5.7329,
      "mean_token_accuracy": 0.1312025487422943,
      "num_tokens": 7446596.0,
      "step": 3385
    },
    {
      "entropy": 5.764281702041626,
      "epoch": 0.33644303294958317,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004993634187182113,
      "loss": 5.8272,
      "mean_token_accuracy": 0.12373630180954934,
      "num_tokens": 7457697.0,
      "step": 3390
    },
    {
      "entropy": 5.917959403991699,
      "epoch": 0.3369392616117507,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004993607525586381,
      "loss": 5.9732,
      "mean_token_accuracy": 0.12327027171850205,
      "num_tokens": 7468541.0,
      "step": 3395
    },
    {
      "entropy": 5.897768545150757,
      "epoch": 0.3374354902739182,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993580808354018,
      "loss": 5.8165,
      "mean_token_accuracy": 0.13104327097535134,
      "num_tokens": 7479781.0,
      "step": 3400
    },
    {
      "entropy": 5.784489107131958,
      "epoch": 0.33793171893608576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993554035485685,
      "loss": 5.7256,
      "mean_token_accuracy": 0.12670466378331185,
      "num_tokens": 7490722.0,
      "step": 3405
    },
    {
      "entropy": 5.846339559555053,
      "epoch": 0.3384279475982533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993527206982047,
      "loss": 5.8409,
      "mean_token_accuracy": 0.12730559408664704,
      "num_tokens": 7502293.0,
      "step": 3410
    },
    {
      "entropy": 5.863930034637451,
      "epoch": 0.3389241762604208,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993500322843767,
      "loss": 5.7785,
      "mean_token_accuracy": 0.1292334884405136,
      "num_tokens": 7513846.0,
      "step": 3415
    },
    {
      "entropy": 5.770083951950073,
      "epoch": 0.33942040492258835,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993473383071514,
      "loss": 5.7935,
      "mean_token_accuracy": 0.1315175212919712,
      "num_tokens": 7524062.0,
      "step": 3420
    },
    {
      "entropy": 5.854935216903686,
      "epoch": 0.3399166335847559,
      "grad_norm": 1.25,
      "learning_rate": 0.0004993446387665956,
      "loss": 5.8413,
      "mean_token_accuracy": 0.13042339831590652,
      "num_tokens": 7535533.0,
      "step": 3425
    },
    {
      "entropy": 5.856246852874756,
      "epoch": 0.3404128622469234,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993419336627761,
      "loss": 5.7325,
      "mean_token_accuracy": 0.13155495524406433,
      "num_tokens": 7545951.0,
      "step": 3430
    },
    {
      "entropy": 5.7223429679870605,
      "epoch": 0.3409090909090909,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993392229957602,
      "loss": 5.6977,
      "mean_token_accuracy": 0.1299275867640972,
      "num_tokens": 7557320.0,
      "step": 3435
    },
    {
      "entropy": 5.792779493331909,
      "epoch": 0.3414053195712584,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993365067656148,
      "loss": 5.7962,
      "mean_token_accuracy": 0.12860119119286537,
      "num_tokens": 7568078.0,
      "step": 3440
    },
    {
      "entropy": 5.849704742431641,
      "epoch": 0.34190154823342594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993337849724076,
      "loss": 5.7972,
      "mean_token_accuracy": 0.1253453992307186,
      "num_tokens": 7578940.0,
      "step": 3445
    },
    {
      "entropy": 5.843108415603638,
      "epoch": 0.3423977768955935,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993310576162059,
      "loss": 5.8008,
      "mean_token_accuracy": 0.12783829718828202,
      "num_tokens": 7589864.0,
      "step": 3450
    },
    {
      "entropy": 5.8652417182922365,
      "epoch": 0.342894005557761,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993283246970772,
      "loss": 5.7527,
      "mean_token_accuracy": 0.12657534927129746,
      "num_tokens": 7599795.0,
      "step": 3455
    },
    {
      "entropy": 5.803637981414795,
      "epoch": 0.34339023421992854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993255862150896,
      "loss": 5.7443,
      "mean_token_accuracy": 0.12936430498957635,
      "num_tokens": 7610056.0,
      "step": 3460
    },
    {
      "entropy": 5.743375253677368,
      "epoch": 0.34388646288209607,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993228421703108,
      "loss": 5.7325,
      "mean_token_accuracy": 0.12720296680927276,
      "num_tokens": 7620218.0,
      "step": 3465
    },
    {
      "entropy": 5.809446525573731,
      "epoch": 0.3443826915442636,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993200925628088,
      "loss": 5.8379,
      "mean_token_accuracy": 0.12768347039818764,
      "num_tokens": 7632781.0,
      "step": 3470
    },
    {
      "entropy": 5.81134819984436,
      "epoch": 0.3448789202064311,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499317337392652,
      "loss": 5.8099,
      "mean_token_accuracy": 0.12796177938580514,
      "num_tokens": 7643845.0,
      "step": 3475
    },
    {
      "entropy": 5.829691648483276,
      "epoch": 0.34537514886859866,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993145766599085,
      "loss": 5.8271,
      "mean_token_accuracy": 0.12467768192291259,
      "num_tokens": 7654583.0,
      "step": 3480
    },
    {
      "entropy": 5.81046953201294,
      "epoch": 0.3458713775307662,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993118103646469,
      "loss": 5.6915,
      "mean_token_accuracy": 0.129618851095438,
      "num_tokens": 7665714.0,
      "step": 3485
    },
    {
      "entropy": 5.820547008514405,
      "epoch": 0.3463676061929337,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004993090385069357,
      "loss": 5.7308,
      "mean_token_accuracy": 0.1329895183444023,
      "num_tokens": 7676011.0,
      "step": 3490
    },
    {
      "entropy": 5.773534440994263,
      "epoch": 0.34686383485510125,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993062610868437,
      "loss": 5.8027,
      "mean_token_accuracy": 0.12820897549390792,
      "num_tokens": 7686927.0,
      "step": 3495
    },
    {
      "entropy": 5.771674966812133,
      "epoch": 0.3473600635172688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993034781044398,
      "loss": 5.7363,
      "mean_token_accuracy": 0.13013842552900315,
      "num_tokens": 7698419.0,
      "step": 3500
    },
    {
      "entropy": 5.8535843849182125,
      "epoch": 0.3478562921794363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499300689559793,
      "loss": 5.7269,
      "mean_token_accuracy": 0.13277798742055893,
      "num_tokens": 7709695.0,
      "step": 3505
    },
    {
      "entropy": 5.764010238647461,
      "epoch": 0.3483525208416038,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004992978954529724,
      "loss": 5.7217,
      "mean_token_accuracy": 0.1324878789484501,
      "num_tokens": 7721309.0,
      "step": 3510
    },
    {
      "entropy": 5.826256799697876,
      "epoch": 0.3488487495037713,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992950957840473,
      "loss": 5.787,
      "mean_token_accuracy": 0.13231385126709938,
      "num_tokens": 7732074.0,
      "step": 3515
    },
    {
      "entropy": 5.832966279983521,
      "epoch": 0.34934497816593885,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499292290553087,
      "loss": 5.8431,
      "mean_token_accuracy": 0.12805222868919372,
      "num_tokens": 7742869.0,
      "step": 3520
    },
    {
      "entropy": 5.725128602981568,
      "epoch": 0.3498412068281064,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992894797601614,
      "loss": 5.7599,
      "mean_token_accuracy": 0.13100404441356658,
      "num_tokens": 7754895.0,
      "step": 3525
    },
    {
      "entropy": 5.865936183929444,
      "epoch": 0.3503374354902739,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049928666340534,
      "loss": 5.7325,
      "mean_token_accuracy": 0.133123978972435,
      "num_tokens": 7765262.0,
      "step": 3530
    },
    {
      "entropy": 5.7653392314910885,
      "epoch": 0.35083366415244144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992838414886925,
      "loss": 5.8065,
      "mean_token_accuracy": 0.12678795903921128,
      "num_tokens": 7777348.0,
      "step": 3535
    },
    {
      "entropy": 5.909204387664795,
      "epoch": 0.35132989281460897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992810140102891,
      "loss": 5.7971,
      "mean_token_accuracy": 0.1277055561542511,
      "num_tokens": 7788161.0,
      "step": 3540
    },
    {
      "entropy": 5.789294052124023,
      "epoch": 0.3518261214767765,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992781809701999,
      "loss": 5.7838,
      "mean_token_accuracy": 0.127443815022707,
      "num_tokens": 7798629.0,
      "step": 3545
    },
    {
      "entropy": 5.752184247970581,
      "epoch": 0.35232235013894403,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992753423684951,
      "loss": 5.7803,
      "mean_token_accuracy": 0.1291307680308819,
      "num_tokens": 7810148.0,
      "step": 3550
    },
    {
      "entropy": 5.854223585128784,
      "epoch": 0.35281857880111156,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992724982052449,
      "loss": 5.772,
      "mean_token_accuracy": 0.12724444046616554,
      "num_tokens": 7821667.0,
      "step": 3555
    },
    {
      "entropy": 5.773481988906861,
      "epoch": 0.3533148074632791,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00049926964848052,
      "loss": 5.7516,
      "mean_token_accuracy": 0.131265527009964,
      "num_tokens": 7832184.0,
      "step": 3560
    },
    {
      "entropy": 5.8223296165466305,
      "epoch": 0.3538110361254466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992667931943913,
      "loss": 5.7725,
      "mean_token_accuracy": 0.1282011777162552,
      "num_tokens": 7842983.0,
      "step": 3565
    },
    {
      "entropy": 5.7912578105926515,
      "epoch": 0.35430726478761415,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992639323469293,
      "loss": 5.7035,
      "mean_token_accuracy": 0.13819223642349243,
      "num_tokens": 7854783.0,
      "step": 3570
    },
    {
      "entropy": 5.731419610977173,
      "epoch": 0.3548034934497817,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499261065938205,
      "loss": 5.7208,
      "mean_token_accuracy": 0.1421904519200325,
      "num_tokens": 7864684.0,
      "step": 3575
    },
    {
      "entropy": 5.806283044815063,
      "epoch": 0.3552997221119492,
      "grad_norm": 1.125,
      "learning_rate": 0.0004992581939682895,
      "loss": 5.8822,
      "mean_token_accuracy": 0.12328537628054619,
      "num_tokens": 7876846.0,
      "step": 3580
    },
    {
      "entropy": 5.9445267677307125,
      "epoch": 0.3557959507741167,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499255316437254,
      "loss": 5.7571,
      "mean_token_accuracy": 0.1253294549882412,
      "num_tokens": 7886826.0,
      "step": 3585
    },
    {
      "entropy": 5.70992693901062,
      "epoch": 0.3562921794362842,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992524333451699,
      "loss": 5.8164,
      "mean_token_accuracy": 0.1269104838371277,
      "num_tokens": 7898941.0,
      "step": 3590
    },
    {
      "entropy": 5.778099727630615,
      "epoch": 0.35678840809845175,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004992495446921087,
      "loss": 5.7158,
      "mean_token_accuracy": 0.13140188977122308,
      "num_tokens": 7909916.0,
      "step": 3595
    },
    {
      "entropy": 5.763299942016602,
      "epoch": 0.3572846367606193,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992466504781419,
      "loss": 5.7521,
      "mean_token_accuracy": 0.13256163597106935,
      "num_tokens": 7922137.0,
      "step": 3600
    },
    {
      "entropy": 5.705967044830322,
      "epoch": 0.3577808654227868,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992437507033415,
      "loss": 5.682,
      "mean_token_accuracy": 0.13475254103541373,
      "num_tokens": 7932940.0,
      "step": 3605
    },
    {
      "entropy": 5.849222087860108,
      "epoch": 0.35827709408495434,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992408453677792,
      "loss": 5.8457,
      "mean_token_accuracy": 0.12133882716298103,
      "num_tokens": 7943735.0,
      "step": 3610
    },
    {
      "entropy": 5.820350694656372,
      "epoch": 0.35877332274712187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992379344715273,
      "loss": 5.7181,
      "mean_token_accuracy": 0.13387718945741653,
      "num_tokens": 7955323.0,
      "step": 3615
    },
    {
      "entropy": 5.682558965682984,
      "epoch": 0.3592695514092894,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992350180146577,
      "loss": 5.6873,
      "mean_token_accuracy": 0.1312878042459488,
      "num_tokens": 7966073.0,
      "step": 3620
    },
    {
      "entropy": 5.83023738861084,
      "epoch": 0.35976578007145693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992320959972428,
      "loss": 5.8631,
      "mean_token_accuracy": 0.12575180530548097,
      "num_tokens": 7977027.0,
      "step": 3625
    },
    {
      "entropy": 5.770997953414917,
      "epoch": 0.36026200873362446,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004992291684193551,
      "loss": 5.7147,
      "mean_token_accuracy": 0.1385928638279438,
      "num_tokens": 7987368.0,
      "step": 3630
    },
    {
      "entropy": 5.824091863632202,
      "epoch": 0.360758237395792,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992262352810672,
      "loss": 5.7941,
      "mean_token_accuracy": 0.12880266457796097,
      "num_tokens": 8000060.0,
      "step": 3635
    },
    {
      "entropy": 5.865066623687744,
      "epoch": 0.3612544660579595,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992232965824518,
      "loss": 5.8425,
      "mean_token_accuracy": 0.1258557267487049,
      "num_tokens": 8011388.0,
      "step": 3640
    },
    {
      "entropy": 5.767081451416016,
      "epoch": 0.36175069472012705,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499220352323582,
      "loss": 5.6936,
      "mean_token_accuracy": 0.13545938059687615,
      "num_tokens": 8021682.0,
      "step": 3645
    },
    {
      "entropy": 5.740088939666748,
      "epoch": 0.3622469233822946,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992174025045304,
      "loss": 5.7226,
      "mean_token_accuracy": 0.13166359663009644,
      "num_tokens": 8032799.0,
      "step": 3650
    },
    {
      "entropy": 5.749204874038696,
      "epoch": 0.3627431520444621,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992144471253704,
      "loss": 5.6042,
      "mean_token_accuracy": 0.13931264877319335,
      "num_tokens": 8042742.0,
      "step": 3655
    },
    {
      "entropy": 5.751514959335327,
      "epoch": 0.3632393807066296,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992114861861753,
      "loss": 5.7875,
      "mean_token_accuracy": 0.1301817037165165,
      "num_tokens": 8054068.0,
      "step": 3660
    },
    {
      "entropy": 5.826439046859742,
      "epoch": 0.3637356093687971,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992085196870184,
      "loss": 5.7834,
      "mean_token_accuracy": 0.12465299963951111,
      "num_tokens": 8065286.0,
      "step": 3665
    },
    {
      "entropy": 5.783612251281738,
      "epoch": 0.36423183803096465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992055476279734,
      "loss": 5.8248,
      "mean_token_accuracy": 0.12121150195598603,
      "num_tokens": 8077122.0,
      "step": 3670
    },
    {
      "entropy": 5.8765967845916744,
      "epoch": 0.3647280666931322,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992025700091139,
      "loss": 5.7674,
      "mean_token_accuracy": 0.13578004464507104,
      "num_tokens": 8087036.0,
      "step": 3675
    },
    {
      "entropy": 5.809439992904663,
      "epoch": 0.3652242953552997,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991995868305138,
      "loss": 5.7908,
      "mean_token_accuracy": 0.1262356698513031,
      "num_tokens": 8098517.0,
      "step": 3680
    },
    {
      "entropy": 5.718502473831177,
      "epoch": 0.36572052401746724,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991965980922471,
      "loss": 5.6595,
      "mean_token_accuracy": 0.14001391604542732,
      "num_tokens": 8107475.0,
      "step": 3685
    },
    {
      "entropy": 5.762603807449341,
      "epoch": 0.36621675267963477,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991936037943878,
      "loss": 5.773,
      "mean_token_accuracy": 0.1286217078566551,
      "num_tokens": 8118529.0,
      "step": 3690
    },
    {
      "entropy": 5.857455635070801,
      "epoch": 0.3667129813418023,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991906039370104,
      "loss": 5.6908,
      "mean_token_accuracy": 0.13512016087770462,
      "num_tokens": 8128951.0,
      "step": 3695
    },
    {
      "entropy": 5.686303615570068,
      "epoch": 0.36720921000396983,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991875985201889,
      "loss": 5.7075,
      "mean_token_accuracy": 0.13315951973199844,
      "num_tokens": 8139385.0,
      "step": 3700
    },
    {
      "entropy": 5.811685514450073,
      "epoch": 0.36770543866613736,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499184587543998,
      "loss": 5.742,
      "mean_token_accuracy": 0.13608823269605635,
      "num_tokens": 8149376.0,
      "step": 3705
    },
    {
      "entropy": 5.820437431335449,
      "epoch": 0.3682016673283049,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991815710085126,
      "loss": 5.7654,
      "mean_token_accuracy": 0.13046534657478331,
      "num_tokens": 8160687.0,
      "step": 3710
    },
    {
      "entropy": 5.7882054328918455,
      "epoch": 0.3686978959904724,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004991785489138072,
      "loss": 5.7638,
      "mean_token_accuracy": 0.1327088348567486,
      "num_tokens": 8171762.0,
      "step": 3715
    },
    {
      "entropy": 5.69438533782959,
      "epoch": 0.36919412465263995,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991755212599569,
      "loss": 5.5439,
      "mean_token_accuracy": 0.13677264451980592,
      "num_tokens": 8181643.0,
      "step": 3720
    },
    {
      "entropy": 5.643247032165528,
      "epoch": 0.3696903533148075,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991724880470366,
      "loss": 5.7955,
      "mean_token_accuracy": 0.13163572698831558,
      "num_tokens": 8192692.0,
      "step": 3725
    },
    {
      "entropy": 5.873797512054443,
      "epoch": 0.370186581976975,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004991694492751217,
      "loss": 5.7475,
      "mean_token_accuracy": 0.128364247828722,
      "num_tokens": 8202941.0,
      "step": 3730
    },
    {
      "entropy": 5.812956142425537,
      "epoch": 0.3706828106391425,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991664049442876,
      "loss": 5.7333,
      "mean_token_accuracy": 0.13337556198239325,
      "num_tokens": 8214206.0,
      "step": 3735
    },
    {
      "entropy": 5.695444488525391,
      "epoch": 0.37117903930131,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004991633550546097,
      "loss": 5.581,
      "mean_token_accuracy": 0.13965674489736557,
      "num_tokens": 8223672.0,
      "step": 3740
    },
    {
      "entropy": 5.721037578582764,
      "epoch": 0.37167526796347755,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991602996061635,
      "loss": 5.715,
      "mean_token_accuracy": 0.13506461381912233,
      "num_tokens": 8234189.0,
      "step": 3745
    },
    {
      "entropy": 5.856908130645752,
      "epoch": 0.3721714966256451,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991572385990249,
      "loss": 5.7461,
      "mean_token_accuracy": 0.13235442489385604,
      "num_tokens": 8245382.0,
      "step": 3750
    },
    {
      "entropy": 5.841987228393554,
      "epoch": 0.3726677252878126,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991541720332698,
      "loss": 5.6943,
      "mean_token_accuracy": 0.13437160477042198,
      "num_tokens": 8254530.0,
      "step": 3755
    },
    {
      "entropy": 5.719356870651245,
      "epoch": 0.37316395394998014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004991510999089743,
      "loss": 5.7434,
      "mean_token_accuracy": 0.13505279794335365,
      "num_tokens": 8264030.0,
      "step": 3760
    },
    {
      "entropy": 5.722121715545654,
      "epoch": 0.3736601826121477,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991480222262145,
      "loss": 5.7103,
      "mean_token_accuracy": 0.13800882697105407,
      "num_tokens": 8273537.0,
      "step": 3765
    },
    {
      "entropy": 5.8228706359863285,
      "epoch": 0.3741564112743152,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991449389850668,
      "loss": 5.7481,
      "mean_token_accuracy": 0.12772027105093003,
      "num_tokens": 8286123.0,
      "step": 3770
    },
    {
      "entropy": 5.774958610534668,
      "epoch": 0.37465263993648273,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004991418501856076,
      "loss": 5.7525,
      "mean_token_accuracy": 0.1304101087152958,
      "num_tokens": 8298115.0,
      "step": 3775
    },
    {
      "entropy": 5.807371997833252,
      "epoch": 0.37514886859865026,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004991387558279134,
      "loss": 5.7969,
      "mean_token_accuracy": 0.1281567707657814,
      "num_tokens": 8309376.0,
      "step": 3780
    },
    {
      "entropy": 5.785305213928223,
      "epoch": 0.3756450972608178,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004991356559120612,
      "loss": 5.7394,
      "mean_token_accuracy": 0.1342696100473404,
      "num_tokens": 8319272.0,
      "step": 3785
    },
    {
      "entropy": 5.77508773803711,
      "epoch": 0.3761413259229853,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004991325504381275,
      "loss": 5.7417,
      "mean_token_accuracy": 0.12871941551566124,
      "num_tokens": 8330042.0,
      "step": 3790
    },
    {
      "entropy": 5.745781087875367,
      "epoch": 0.37663755458515286,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991294394061896,
      "loss": 5.6953,
      "mean_token_accuracy": 0.13092486634850503,
      "num_tokens": 8340394.0,
      "step": 3795
    },
    {
      "entropy": 5.705888795852661,
      "epoch": 0.3771337832473204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991263228163246,
      "loss": 5.7297,
      "mean_token_accuracy": 0.1320924386382103,
      "num_tokens": 8351103.0,
      "step": 3800
    },
    {
      "entropy": 5.807416343688965,
      "epoch": 0.3776300119094879,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991232006686096,
      "loss": 5.7529,
      "mean_token_accuracy": 0.13187019526958466,
      "num_tokens": 8363215.0,
      "step": 3805
    },
    {
      "entropy": 5.846437931060791,
      "epoch": 0.3781262405716554,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991200729631223,
      "loss": 5.806,
      "mean_token_accuracy": 0.12410712242126465,
      "num_tokens": 8374602.0,
      "step": 3810
    },
    {
      "entropy": 5.717173862457275,
      "epoch": 0.3786224692338229,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991169396999401,
      "loss": 5.7259,
      "mean_token_accuracy": 0.13081120625138282,
      "num_tokens": 8386470.0,
      "step": 3815
    },
    {
      "entropy": 5.785051250457764,
      "epoch": 0.37911869789599045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004991138008791408,
      "loss": 5.7534,
      "mean_token_accuracy": 0.128629819303751,
      "num_tokens": 8396284.0,
      "step": 3820
    },
    {
      "entropy": 5.862023115158081,
      "epoch": 0.379614926558158,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991106565008021,
      "loss": 5.7746,
      "mean_token_accuracy": 0.1275525361299515,
      "num_tokens": 8407856.0,
      "step": 3825
    },
    {
      "entropy": 5.771010065078736,
      "epoch": 0.3801111552203255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991075065650019,
      "loss": 5.6827,
      "mean_token_accuracy": 0.13699871823191642,
      "num_tokens": 8419560.0,
      "step": 3830
    },
    {
      "entropy": 5.767512941360474,
      "epoch": 0.38060738388249304,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991043510718186,
      "loss": 5.6836,
      "mean_token_accuracy": 0.1351518914103508,
      "num_tokens": 8430594.0,
      "step": 3835
    },
    {
      "entropy": 5.679202079772949,
      "epoch": 0.3811036125446606,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991011900213303,
      "loss": 5.6467,
      "mean_token_accuracy": 0.1384870320558548,
      "num_tokens": 8440549.0,
      "step": 3840
    },
    {
      "entropy": 5.689004707336426,
      "epoch": 0.3815998412068281,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990980234136152,
      "loss": 5.698,
      "mean_token_accuracy": 0.13232575953006745,
      "num_tokens": 8451553.0,
      "step": 3845
    },
    {
      "entropy": 5.762536334991455,
      "epoch": 0.38209606986899564,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990948512487522,
      "loss": 5.6864,
      "mean_token_accuracy": 0.1370437279343605,
      "num_tokens": 8462899.0,
      "step": 3850
    },
    {
      "entropy": 5.751942157745361,
      "epoch": 0.38259229853116317,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004990916735268198,
      "loss": 5.7507,
      "mean_token_accuracy": 0.13097452595829964,
      "num_tokens": 8474566.0,
      "step": 3855
    },
    {
      "entropy": 5.813113880157471,
      "epoch": 0.3830885271933307,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990884902478966,
      "loss": 5.7593,
      "mean_token_accuracy": 0.13342129066586494,
      "num_tokens": 8484786.0,
      "step": 3860
    },
    {
      "entropy": 5.697579097747803,
      "epoch": 0.3835847558554982,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990853014120619,
      "loss": 5.7002,
      "mean_token_accuracy": 0.13859767690300942,
      "num_tokens": 8496416.0,
      "step": 3865
    },
    {
      "entropy": 5.739745616912842,
      "epoch": 0.38408098451766576,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990821070193944,
      "loss": 5.7227,
      "mean_token_accuracy": 0.13303989171981812,
      "num_tokens": 8507287.0,
      "step": 3870
    },
    {
      "entropy": 5.820405292510986,
      "epoch": 0.3845772131798333,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004990789070699736,
      "loss": 5.7556,
      "mean_token_accuracy": 0.12703920155763626,
      "num_tokens": 8517922.0,
      "step": 3875
    },
    {
      "entropy": 5.721597576141358,
      "epoch": 0.3850734418420008,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004990757015638787,
      "loss": 5.6972,
      "mean_token_accuracy": 0.13949232697486877,
      "num_tokens": 8529532.0,
      "step": 3880
    },
    {
      "entropy": 5.7820713996887205,
      "epoch": 0.3855696705041683,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990724905011894,
      "loss": 5.6571,
      "mean_token_accuracy": 0.1373422287404537,
      "num_tokens": 8540207.0,
      "step": 3885
    },
    {
      "entropy": 5.762848043441773,
      "epoch": 0.3860658991663358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499069273881985,
      "loss": 5.7017,
      "mean_token_accuracy": 0.12694768533110617,
      "num_tokens": 8551634.0,
      "step": 3890
    },
    {
      "entropy": 5.752759838104248,
      "epoch": 0.38656212782850335,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004990660517063455,
      "loss": 5.7688,
      "mean_token_accuracy": 0.12492152824997901,
      "num_tokens": 8563256.0,
      "step": 3895
    },
    {
      "entropy": 5.749443483352661,
      "epoch": 0.3870583564906709,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990628239743508,
      "loss": 5.779,
      "mean_token_accuracy": 0.13167861476540565,
      "num_tokens": 8573833.0,
      "step": 3900
    },
    {
      "entropy": 5.754673337936401,
      "epoch": 0.3875545851528384,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990595906860807,
      "loss": 5.6974,
      "mean_token_accuracy": 0.13239262998104095,
      "num_tokens": 8584283.0,
      "step": 3905
    },
    {
      "entropy": 5.706582689285279,
      "epoch": 0.38805081381500595,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004990563518416157,
      "loss": 5.7623,
      "mean_token_accuracy": 0.13148060888051988,
      "num_tokens": 8595344.0,
      "step": 3910
    },
    {
      "entropy": 5.780977535247803,
      "epoch": 0.3885470424771735,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990531074410358,
      "loss": 5.6837,
      "mean_token_accuracy": 0.13093154430389403,
      "num_tokens": 8605625.0,
      "step": 3915
    },
    {
      "entropy": 5.7571268558502195,
      "epoch": 0.389043271139341,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990498574844218,
      "loss": 5.6458,
      "mean_token_accuracy": 0.1339184783399105,
      "num_tokens": 8616259.0,
      "step": 3920
    },
    {
      "entropy": 5.728362083435059,
      "epoch": 0.38953949980150854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990466019718542,
      "loss": 5.6615,
      "mean_token_accuracy": 0.1356678619980812,
      "num_tokens": 8627179.0,
      "step": 3925
    },
    {
      "entropy": 5.78178448677063,
      "epoch": 0.39003572846367607,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004990433409034134,
      "loss": 5.7846,
      "mean_token_accuracy": 0.1375245787203312,
      "num_tokens": 8637448.0,
      "step": 3930
    },
    {
      "entropy": 5.728059768676758,
      "epoch": 0.3905319571258436,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990400742791806,
      "loss": 5.6886,
      "mean_token_accuracy": 0.13078238666057587,
      "num_tokens": 8649168.0,
      "step": 3935
    },
    {
      "entropy": 5.829911756515503,
      "epoch": 0.39102818578801113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990368020992368,
      "loss": 5.7997,
      "mean_token_accuracy": 0.1307859167456627,
      "num_tokens": 8660653.0,
      "step": 3940
    },
    {
      "entropy": 5.804625225067139,
      "epoch": 0.39152441445017866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499033524363663,
      "loss": 5.7734,
      "mean_token_accuracy": 0.13561775982379914,
      "num_tokens": 8672000.0,
      "step": 3945
    },
    {
      "entropy": 5.6982584476470945,
      "epoch": 0.3920206431123462,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990302410725405,
      "loss": 5.7061,
      "mean_token_accuracy": 0.13653287291526794,
      "num_tokens": 8682455.0,
      "step": 3950
    },
    {
      "entropy": 5.729883527755737,
      "epoch": 0.3925168717745137,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990269522259509,
      "loss": 5.6706,
      "mean_token_accuracy": 0.13183292672038077,
      "num_tokens": 8692975.0,
      "step": 3955
    },
    {
      "entropy": 5.70659441947937,
      "epoch": 0.3930131004366812,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990236578239755,
      "loss": 5.6486,
      "mean_token_accuracy": 0.13803303763270378,
      "num_tokens": 8703569.0,
      "step": 3960
    },
    {
      "entropy": 5.760754776000977,
      "epoch": 0.3935093290988487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990203578666962,
      "loss": 5.7539,
      "mean_token_accuracy": 0.12984618917107582,
      "num_tokens": 8713734.0,
      "step": 3965
    },
    {
      "entropy": 5.768778276443482,
      "epoch": 0.39400555776101626,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990170523541948,
      "loss": 5.6486,
      "mean_token_accuracy": 0.14092251285910606,
      "num_tokens": 8723852.0,
      "step": 3970
    },
    {
      "entropy": 5.759503269195557,
      "epoch": 0.3945017864231838,
      "grad_norm": 1.0,
      "learning_rate": 0.0004990137412865531,
      "loss": 5.7984,
      "mean_token_accuracy": 0.13060248047113418,
      "num_tokens": 8734300.0,
      "step": 3975
    },
    {
      "entropy": 5.8276502132415775,
      "epoch": 0.3949980150853513,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004990104246638533,
      "loss": 5.8104,
      "mean_token_accuracy": 0.1265176624059677,
      "num_tokens": 8745022.0,
      "step": 3980
    },
    {
      "entropy": 5.754931688308716,
      "epoch": 0.39549424374751885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990071024861778,
      "loss": 5.6921,
      "mean_token_accuracy": 0.12658729776740074,
      "num_tokens": 8754870.0,
      "step": 3985
    },
    {
      "entropy": 5.858518075942993,
      "epoch": 0.3959904724096864,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990037747536087,
      "loss": 5.7495,
      "mean_token_accuracy": 0.13611997216939925,
      "num_tokens": 8765536.0,
      "step": 3990
    },
    {
      "entropy": 5.748577499389649,
      "epoch": 0.3964867010718539,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004990004414662288,
      "loss": 5.7216,
      "mean_token_accuracy": 0.12786149084568024,
      "num_tokens": 8776898.0,
      "step": 3995
    },
    {
      "entropy": 5.792099571228027,
      "epoch": 0.39698292973402144,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989971026241205,
      "loss": 5.7755,
      "mean_token_accuracy": 0.12892835810780526,
      "num_tokens": 8788176.0,
      "step": 4000
    },
    {
      "entropy": 5.733926153182983,
      "epoch": 0.39747915839618897,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004989937582273668,
      "loss": 5.7005,
      "mean_token_accuracy": 0.13020156919956208,
      "num_tokens": 8798939.0,
      "step": 4005
    },
    {
      "entropy": 5.747070169448852,
      "epoch": 0.3979753870583565,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989904082760506,
      "loss": 5.7388,
      "mean_token_accuracy": 0.13505999445915223,
      "num_tokens": 8809092.0,
      "step": 4010
    },
    {
      "entropy": 5.787953567504883,
      "epoch": 0.39847161572052403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498987052770255,
      "loss": 5.6508,
      "mean_token_accuracy": 0.13869067206978797,
      "num_tokens": 8819352.0,
      "step": 4015
    },
    {
      "entropy": 5.787040710449219,
      "epoch": 0.39896784438269156,
      "grad_norm": 1.109375,
      "learning_rate": 0.000498983691710063,
      "loss": 5.7749,
      "mean_token_accuracy": 0.12908290550112725,
      "num_tokens": 8829878.0,
      "step": 4020
    },
    {
      "entropy": 5.729852914810181,
      "epoch": 0.3994640730448591,
      "grad_norm": 1.0,
      "learning_rate": 0.000498980325095558,
      "loss": 5.6428,
      "mean_token_accuracy": 0.13541897460818292,
      "num_tokens": 8840364.0,
      "step": 4025
    },
    {
      "entropy": 5.7771155834198,
      "epoch": 0.3999603017070266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989769529268238,
      "loss": 5.7598,
      "mean_token_accuracy": 0.12786714881658554,
      "num_tokens": 8851340.0,
      "step": 4030
    },
    {
      "entropy": 5.7071342945098875,
      "epoch": 0.4004565303691941,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989735752039436,
      "loss": 5.7372,
      "mean_token_accuracy": 0.13070829212665558,
      "num_tokens": 8862571.0,
      "step": 4035
    },
    {
      "entropy": 5.798667860031128,
      "epoch": 0.4009527590313616,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989701919270015,
      "loss": 5.7351,
      "mean_token_accuracy": 0.13419504538178445,
      "num_tokens": 8873927.0,
      "step": 4040
    },
    {
      "entropy": 5.842920541763306,
      "epoch": 0.40144898769352916,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004989668030960812,
      "loss": 5.9317,
      "mean_token_accuracy": 0.12279433906078338,
      "num_tokens": 8886542.0,
      "step": 4045
    },
    {
      "entropy": 5.792601203918457,
      "epoch": 0.4019452163556967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989634087112667,
      "loss": 5.6629,
      "mean_token_accuracy": 0.13334398940205575,
      "num_tokens": 8896114.0,
      "step": 4050
    },
    {
      "entropy": 5.806399488449097,
      "epoch": 0.4024414450178642,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989600087726425,
      "loss": 5.7324,
      "mean_token_accuracy": 0.1301003411412239,
      "num_tokens": 8906303.0,
      "step": 4055
    },
    {
      "entropy": 5.814186906814575,
      "epoch": 0.40293767368003175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989566032802924,
      "loss": 5.788,
      "mean_token_accuracy": 0.12885491251945497,
      "num_tokens": 8916961.0,
      "step": 4060
    },
    {
      "entropy": 5.859943914413452,
      "epoch": 0.4034339023421993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989531922343013,
      "loss": 5.8735,
      "mean_token_accuracy": 0.1225248247385025,
      "num_tokens": 8927700.0,
      "step": 4065
    },
    {
      "entropy": 5.808253908157349,
      "epoch": 0.4039301310043668,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989497756347534,
      "loss": 5.6946,
      "mean_token_accuracy": 0.13116186931729318,
      "num_tokens": 8939655.0,
      "step": 4070
    },
    {
      "entropy": 5.718573713302613,
      "epoch": 0.40442635966653434,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989463534817339,
      "loss": 5.6443,
      "mean_token_accuracy": 0.13748994767665862,
      "num_tokens": 8951238.0,
      "step": 4075
    },
    {
      "entropy": 5.767578506469727,
      "epoch": 0.40492258832870187,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989429257753273,
      "loss": 5.7007,
      "mean_token_accuracy": 0.13767925277352333,
      "num_tokens": 8962019.0,
      "step": 4080
    },
    {
      "entropy": 5.699139547348023,
      "epoch": 0.4054188169908694,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989394925156187,
      "loss": 5.7023,
      "mean_token_accuracy": 0.13865290507674216,
      "num_tokens": 8973861.0,
      "step": 4085
    },
    {
      "entropy": 5.702186584472656,
      "epoch": 0.40591504565303693,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989360537026932,
      "loss": 5.7172,
      "mean_token_accuracy": 0.13503216281533242,
      "num_tokens": 8985981.0,
      "step": 4090
    },
    {
      "entropy": 5.798025703430175,
      "epoch": 0.40641127431520446,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498932609336636,
      "loss": 5.6096,
      "mean_token_accuracy": 0.13680272847414016,
      "num_tokens": 8995318.0,
      "step": 4095
    },
    {
      "entropy": 5.718692016601563,
      "epoch": 0.406907502977372,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989291594175327,
      "loss": 5.6404,
      "mean_token_accuracy": 0.13456545695662497,
      "num_tokens": 9004679.0,
      "step": 4100
    },
    {
      "entropy": 5.861246585845947,
      "epoch": 0.4074037316395395,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989257039454689,
      "loss": 5.7761,
      "mean_token_accuracy": 0.13330629467964172,
      "num_tokens": 9016422.0,
      "step": 4105
    },
    {
      "entropy": 5.722641658782959,
      "epoch": 0.40789996030170705,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049892224292053,
      "loss": 5.6967,
      "mean_token_accuracy": 0.13267009183764458,
      "num_tokens": 9027914.0,
      "step": 4110
    },
    {
      "entropy": 5.695920372009278,
      "epoch": 0.40839618896387453,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004989187763428021,
      "loss": 5.6558,
      "mean_token_accuracy": 0.14080327674746512,
      "num_tokens": 9038701.0,
      "step": 4115
    },
    {
      "entropy": 5.735637044906616,
      "epoch": 0.40889241762604206,
      "grad_norm": 1.109375,
      "learning_rate": 0.000498915304212371,
      "loss": 5.7198,
      "mean_token_accuracy": 0.13315169736742974,
      "num_tokens": 9050620.0,
      "step": 4120
    },
    {
      "entropy": 5.794487714767456,
      "epoch": 0.4093886462882096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989118265293228,
      "loss": 5.7439,
      "mean_token_accuracy": 0.13364142403006554,
      "num_tokens": 9060900.0,
      "step": 4125
    },
    {
      "entropy": 5.77433934211731,
      "epoch": 0.4098848749503771,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004989083432937439,
      "loss": 5.6609,
      "mean_token_accuracy": 0.13388703539967536,
      "num_tokens": 9072122.0,
      "step": 4130
    },
    {
      "entropy": 5.710920810699463,
      "epoch": 0.41038110361254465,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989048545057204,
      "loss": 5.6995,
      "mean_token_accuracy": 0.13454304486513138,
      "num_tokens": 9082780.0,
      "step": 4135
    },
    {
      "entropy": 5.70075044631958,
      "epoch": 0.4108773322747122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989013601653391,
      "loss": 5.7259,
      "mean_token_accuracy": 0.13173099905252456,
      "num_tokens": 9093518.0,
      "step": 4140
    },
    {
      "entropy": 5.779675149917603,
      "epoch": 0.4113735609368797,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988978602726866,
      "loss": 5.7615,
      "mean_token_accuracy": 0.1367797940969467,
      "num_tokens": 9105035.0,
      "step": 4145
    },
    {
      "entropy": 5.7428912162780765,
      "epoch": 0.41186978959904724,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988943548278494,
      "loss": 5.6694,
      "mean_token_accuracy": 0.12952224314212799,
      "num_tokens": 9117154.0,
      "step": 4150
    },
    {
      "entropy": 5.698440980911255,
      "epoch": 0.4123660182612148,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988908438309149,
      "loss": 5.7325,
      "mean_token_accuracy": 0.1313714452087879,
      "num_tokens": 9128818.0,
      "step": 4155
    },
    {
      "entropy": 5.804380702972412,
      "epoch": 0.4128622469233823,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004988873272819697,
      "loss": 5.7558,
      "mean_token_accuracy": 0.13293662592768668,
      "num_tokens": 9140188.0,
      "step": 4160
    },
    {
      "entropy": 5.897729969024658,
      "epoch": 0.41335847558554983,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004988838051811014,
      "loss": 5.8471,
      "mean_token_accuracy": 0.1278106726706028,
      "num_tokens": 9150565.0,
      "step": 4165
    },
    {
      "entropy": 5.7736671447753904,
      "epoch": 0.41385470424771736,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498880277528397,
      "loss": 5.684,
      "mean_token_accuracy": 0.13222259730100633,
      "num_tokens": 9160130.0,
      "step": 4170
    },
    {
      "entropy": 5.690751600265503,
      "epoch": 0.4143509329098849,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988767443239443,
      "loss": 5.6746,
      "mean_token_accuracy": 0.13625940755009652,
      "num_tokens": 9171740.0,
      "step": 4175
    },
    {
      "entropy": 5.721519517898559,
      "epoch": 0.4148471615720524,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988732055678307,
      "loss": 5.6384,
      "mean_token_accuracy": 0.14912208542227745,
      "num_tokens": 9182459.0,
      "step": 4180
    },
    {
      "entropy": 5.836207294464112,
      "epoch": 0.41534339023421996,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498869661260144,
      "loss": 5.6811,
      "mean_token_accuracy": 0.13373003005981446,
      "num_tokens": 9192284.0,
      "step": 4185
    },
    {
      "entropy": 5.731276655197144,
      "epoch": 0.41583961889638743,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000498866111400972,
      "loss": 5.6921,
      "mean_token_accuracy": 0.1359886907041073,
      "num_tokens": 9202700.0,
      "step": 4190
    },
    {
      "entropy": 5.68711485862732,
      "epoch": 0.41633584755855496,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004988625559904029,
      "loss": 5.6895,
      "mean_token_accuracy": 0.13072032630443572,
      "num_tokens": 9213013.0,
      "step": 4195
    },
    {
      "entropy": 5.8586445331573485,
      "epoch": 0.4168320762207225,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988589950285248,
      "loss": 5.7644,
      "mean_token_accuracy": 0.13009681701660156,
      "num_tokens": 9224568.0,
      "step": 4200
    },
    {
      "entropy": 5.7014992237091064,
      "epoch": 0.41732830488289,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988554285154259,
      "loss": 5.6349,
      "mean_token_accuracy": 0.1338179163634777,
      "num_tokens": 9233791.0,
      "step": 4205
    },
    {
      "entropy": 5.699745559692383,
      "epoch": 0.41782453354505755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988518564511948,
      "loss": 5.7055,
      "mean_token_accuracy": 0.1301266260445118,
      "num_tokens": 9245048.0,
      "step": 4210
    },
    {
      "entropy": 5.8795177936553955,
      "epoch": 0.4183207622072251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988482788359202,
      "loss": 5.7239,
      "mean_token_accuracy": 0.13848262578248977,
      "num_tokens": 9256508.0,
      "step": 4215
    },
    {
      "entropy": 5.765008163452149,
      "epoch": 0.4188169908693926,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988446956696904,
      "loss": 5.7431,
      "mean_token_accuracy": 0.13248386681079866,
      "num_tokens": 9267871.0,
      "step": 4220
    },
    {
      "entropy": 5.804168462753296,
      "epoch": 0.41931321953156014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988411069525947,
      "loss": 5.8064,
      "mean_token_accuracy": 0.12762014716863632,
      "num_tokens": 9279287.0,
      "step": 4225
    },
    {
      "entropy": 5.8505165576934814,
      "epoch": 0.4198094481937277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004988375126847216,
      "loss": 5.81,
      "mean_token_accuracy": 0.12756233513355256,
      "num_tokens": 9291257.0,
      "step": 4230
    },
    {
      "entropy": 5.800893497467041,
      "epoch": 0.4203056768558952,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004988339128661607,
      "loss": 5.7671,
      "mean_token_accuracy": 0.1308151975274086,
      "num_tokens": 9302590.0,
      "step": 4235
    },
    {
      "entropy": 5.663939094543457,
      "epoch": 0.42080190551806274,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498830307497001,
      "loss": 5.6354,
      "mean_token_accuracy": 0.1342343732714653,
      "num_tokens": 9313174.0,
      "step": 4240
    },
    {
      "entropy": 5.75346884727478,
      "epoch": 0.42129813418023027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000498826696577332,
      "loss": 5.7124,
      "mean_token_accuracy": 0.13228978142142295,
      "num_tokens": 9325034.0,
      "step": 4245
    },
    {
      "entropy": 5.772179889678955,
      "epoch": 0.4217943628423978,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988230801072432,
      "loss": 5.6846,
      "mean_token_accuracy": 0.1347103200852871,
      "num_tokens": 9336834.0,
      "step": 4250
    },
    {
      "entropy": 5.712495279312134,
      "epoch": 0.4222905915045653,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988194580868244,
      "loss": 5.7062,
      "mean_token_accuracy": 0.13381339460611344,
      "num_tokens": 9347747.0,
      "step": 4255
    },
    {
      "entropy": 5.727442264556885,
      "epoch": 0.42278682016673286,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988158305161654,
      "loss": 5.7096,
      "mean_token_accuracy": 0.13135963007807733,
      "num_tokens": 9359186.0,
      "step": 4260
    },
    {
      "entropy": 5.726292657852173,
      "epoch": 0.42328304882890033,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498812197395356,
      "loss": 5.643,
      "mean_token_accuracy": 0.13420115262269974,
      "num_tokens": 9369485.0,
      "step": 4265
    },
    {
      "entropy": 5.756616640090942,
      "epoch": 0.42377927749106786,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004988085587244863,
      "loss": 5.7421,
      "mean_token_accuracy": 0.1320301651954651,
      "num_tokens": 9382658.0,
      "step": 4270
    },
    {
      "entropy": 5.704425382614136,
      "epoch": 0.4242755061532354,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988049145036466,
      "loss": 5.7081,
      "mean_token_accuracy": 0.1268060877919197,
      "num_tokens": 9394151.0,
      "step": 4275
    },
    {
      "entropy": 5.870201253890992,
      "epoch": 0.4247717348154029,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004988012647329273,
      "loss": 5.7866,
      "mean_token_accuracy": 0.12479467168450356,
      "num_tokens": 9405956.0,
      "step": 4280
    },
    {
      "entropy": 5.763204097747803,
      "epoch": 0.42526796347757045,
      "grad_norm": 1.03125,
      "learning_rate": 0.000498797609412419,
      "loss": 5.6691,
      "mean_token_accuracy": 0.13346681147813796,
      "num_tokens": 9416889.0,
      "step": 4285
    },
    {
      "entropy": 5.74954628944397,
      "epoch": 0.425764192139738,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987939485422122,
      "loss": 5.7826,
      "mean_token_accuracy": 0.12909511998295783,
      "num_tokens": 9427602.0,
      "step": 4290
    },
    {
      "entropy": 5.825077533721924,
      "epoch": 0.4262604208019055,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987902821223976,
      "loss": 5.6888,
      "mean_token_accuracy": 0.13714735433459282,
      "num_tokens": 9438557.0,
      "step": 4295
    },
    {
      "entropy": 5.731916427612305,
      "epoch": 0.42675664946407305,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004987866101530663,
      "loss": 5.7532,
      "mean_token_accuracy": 0.13179253563284873,
      "num_tokens": 9449598.0,
      "step": 4300
    },
    {
      "entropy": 5.765017986297607,
      "epoch": 0.4272528781262406,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987829326343092,
      "loss": 5.6573,
      "mean_token_accuracy": 0.14001653492450714,
      "num_tokens": 9459740.0,
      "step": 4305
    },
    {
      "entropy": 5.717455673217773,
      "epoch": 0.4277491067884081,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004987792495662177,
      "loss": 5.6955,
      "mean_token_accuracy": 0.1312531128525734,
      "num_tokens": 9469835.0,
      "step": 4310
    },
    {
      "entropy": 5.793053722381591,
      "epoch": 0.42824533545057564,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498775560948883,
      "loss": 5.7655,
      "mean_token_accuracy": 0.1313070349395275,
      "num_tokens": 9480735.0,
      "step": 4315
    },
    {
      "entropy": 5.741236734390259,
      "epoch": 0.42874156411274317,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004987718667823965,
      "loss": 5.668,
      "mean_token_accuracy": 0.13478718027472497,
      "num_tokens": 9491321.0,
      "step": 4320
    },
    {
      "entropy": 5.744964456558227,
      "epoch": 0.4292377927749107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049876816706685,
      "loss": 5.809,
      "mean_token_accuracy": 0.12865864261984825,
      "num_tokens": 9503507.0,
      "step": 4325
    },
    {
      "entropy": 5.820233726501465,
      "epoch": 0.42973402143707823,
      "grad_norm": 1.84375,
      "learning_rate": 0.000498764461802335,
      "loss": 5.8134,
      "mean_token_accuracy": 0.13732830286026002,
      "num_tokens": 9514404.0,
      "step": 4330
    },
    {
      "entropy": 5.733412551879883,
      "epoch": 0.43023025009924576,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004987607509889437,
      "loss": 5.6735,
      "mean_token_accuracy": 0.13512502685189248,
      "num_tokens": 9525555.0,
      "step": 4335
    },
    {
      "entropy": 5.790317058563232,
      "epoch": 0.43072647876141323,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987570346267677,
      "loss": 5.6567,
      "mean_token_accuracy": 0.14046468064188958,
      "num_tokens": 9535084.0,
      "step": 4340
    },
    {
      "entropy": 5.746789455413818,
      "epoch": 0.43122270742358076,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987533127158996,
      "loss": 5.6852,
      "mean_token_accuracy": 0.13589171767234803,
      "num_tokens": 9546170.0,
      "step": 4345
    },
    {
      "entropy": 5.726515817642212,
      "epoch": 0.4317189360857483,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987495852564314,
      "loss": 5.6739,
      "mean_token_accuracy": 0.13515314161777497,
      "num_tokens": 9556793.0,
      "step": 4350
    },
    {
      "entropy": 5.713581562042236,
      "epoch": 0.4322151647479158,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004987458522484556,
      "loss": 5.6383,
      "mean_token_accuracy": 0.1377846635878086,
      "num_tokens": 9568114.0,
      "step": 4355
    },
    {
      "entropy": 5.778633546829224,
      "epoch": 0.43271139341008336,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004987421136920648,
      "loss": 5.633,
      "mean_token_accuracy": 0.1307872377336025,
      "num_tokens": 9579119.0,
      "step": 4360
    },
    {
      "entropy": 5.659013366699218,
      "epoch": 0.4332076220722509,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004987383695873517,
      "loss": 5.6973,
      "mean_token_accuracy": 0.13660963848233224,
      "num_tokens": 9590259.0,
      "step": 4365
    },
    {
      "entropy": 5.7897804260253904,
      "epoch": 0.4337038507344184,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987346199344092,
      "loss": 5.7071,
      "mean_token_accuracy": 0.13325831666588783,
      "num_tokens": 9600431.0,
      "step": 4370
    },
    {
      "entropy": 5.736825799942016,
      "epoch": 0.43420007939658595,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987308647333303,
      "loss": 5.6958,
      "mean_token_accuracy": 0.132701276242733,
      "num_tokens": 9610286.0,
      "step": 4375
    },
    {
      "entropy": 5.765596675872803,
      "epoch": 0.4346963080587535,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004987271039842079,
      "loss": 5.7085,
      "mean_token_accuracy": 0.13259144350886345,
      "num_tokens": 9621148.0,
      "step": 4380
    },
    {
      "entropy": 5.761705350875855,
      "epoch": 0.435192536720921,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987233376871355,
      "loss": 5.5976,
      "mean_token_accuracy": 0.13539344295859337,
      "num_tokens": 9631300.0,
      "step": 4385
    },
    {
      "entropy": 5.7267553329467775,
      "epoch": 0.43568876538308854,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987195658422063,
      "loss": 5.708,
      "mean_token_accuracy": 0.1294101282954216,
      "num_tokens": 9641986.0,
      "step": 4390
    },
    {
      "entropy": 5.728924798965454,
      "epoch": 0.43618499404525607,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987157884495141,
      "loss": 5.7197,
      "mean_token_accuracy": 0.13337371572852136,
      "num_tokens": 9652962.0,
      "step": 4395
    },
    {
      "entropy": 5.735528469085693,
      "epoch": 0.4366812227074236,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987120055091523,
      "loss": 5.7407,
      "mean_token_accuracy": 0.13252369463443756,
      "num_tokens": 9664084.0,
      "step": 4400
    },
    {
      "entropy": 5.70850019454956,
      "epoch": 0.43717745136959113,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987082170212149,
      "loss": 5.653,
      "mean_token_accuracy": 0.13495920971035957,
      "num_tokens": 9675821.0,
      "step": 4405
    },
    {
      "entropy": 5.791933012008667,
      "epoch": 0.43767368003175866,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004987044229857957,
      "loss": 5.7684,
      "mean_token_accuracy": 0.13300022408366202,
      "num_tokens": 9687964.0,
      "step": 4410
    },
    {
      "entropy": 5.7553457736969,
      "epoch": 0.43816990869392614,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987006234029889,
      "loss": 5.7364,
      "mean_token_accuracy": 0.13509790748357772,
      "num_tokens": 9699653.0,
      "step": 4415
    },
    {
      "entropy": 5.820305919647216,
      "epoch": 0.43866613735609367,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986968182728885,
      "loss": 5.7504,
      "mean_token_accuracy": 0.13183074593544006,
      "num_tokens": 9709872.0,
      "step": 4420
    },
    {
      "entropy": 5.804308652877808,
      "epoch": 0.4391623660182612,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004986930075955892,
      "loss": 5.6777,
      "mean_token_accuracy": 0.12841553315520288,
      "num_tokens": 9720589.0,
      "step": 4425
    },
    {
      "entropy": 5.678828144073487,
      "epoch": 0.4396585946804287,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986891913711853,
      "loss": 5.6132,
      "mean_token_accuracy": 0.1378442943096161,
      "num_tokens": 9731862.0,
      "step": 4430
    },
    {
      "entropy": 5.726953935623169,
      "epoch": 0.44015482334259626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986853695997715,
      "loss": 5.7703,
      "mean_token_accuracy": 0.1355280302464962,
      "num_tokens": 9743710.0,
      "step": 4435
    },
    {
      "entropy": 5.76676754951477,
      "epoch": 0.4406510520047638,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004986815422814425,
      "loss": 5.7003,
      "mean_token_accuracy": 0.12463414892554284,
      "num_tokens": 9753855.0,
      "step": 4440
    },
    {
      "entropy": 5.8478355884552,
      "epoch": 0.4411472806669313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986777094162932,
      "loss": 5.7266,
      "mean_token_accuracy": 0.1364635020494461,
      "num_tokens": 9763883.0,
      "step": 4445
    },
    {
      "entropy": 5.7128133296966555,
      "epoch": 0.44164350932909885,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986738710044187,
      "loss": 5.6756,
      "mean_token_accuracy": 0.13460996970534325,
      "num_tokens": 9774890.0,
      "step": 4450
    },
    {
      "entropy": 5.650014448165893,
      "epoch": 0.4421397379912664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986700270459142,
      "loss": 5.6621,
      "mean_token_accuracy": 0.13714212104678153,
      "num_tokens": 9785249.0,
      "step": 4455
    },
    {
      "entropy": 5.740856885910034,
      "epoch": 0.4426359666534339,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986661775408749,
      "loss": 5.6623,
      "mean_token_accuracy": 0.13843592032790183,
      "num_tokens": 9795723.0,
      "step": 4460
    },
    {
      "entropy": 5.7767352104187015,
      "epoch": 0.44313219531560144,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986623224893965,
      "loss": 5.623,
      "mean_token_accuracy": 0.13589017540216447,
      "num_tokens": 9807637.0,
      "step": 4465
    },
    {
      "entropy": 5.7211939811706545,
      "epoch": 0.44362842397776897,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986584618915744,
      "loss": 5.7314,
      "mean_token_accuracy": 0.13530121594667435,
      "num_tokens": 9820298.0,
      "step": 4470
    },
    {
      "entropy": 5.7044487476348875,
      "epoch": 0.4441246526399365,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004986545957475045,
      "loss": 5.6198,
      "mean_token_accuracy": 0.13783113807439804,
      "num_tokens": 9831251.0,
      "step": 4475
    },
    {
      "entropy": 5.665576314926147,
      "epoch": 0.44462088130210403,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986507240572825,
      "loss": 5.6883,
      "mean_token_accuracy": 0.1373862199485302,
      "num_tokens": 9842884.0,
      "step": 4480
    },
    {
      "entropy": 5.705218696594239,
      "epoch": 0.44511710996427156,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004986468468210042,
      "loss": 5.582,
      "mean_token_accuracy": 0.13796113431453705,
      "num_tokens": 9854413.0,
      "step": 4485
    },
    {
      "entropy": 5.7773439407348635,
      "epoch": 0.44561333862643904,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986429640387663,
      "loss": 5.7264,
      "mean_token_accuracy": 0.13711052760481834,
      "num_tokens": 9865771.0,
      "step": 4490
    },
    {
      "entropy": 5.729163122177124,
      "epoch": 0.44610956728860657,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986390757106648,
      "loss": 5.6396,
      "mean_token_accuracy": 0.14091018065810204,
      "num_tokens": 9876717.0,
      "step": 4495
    },
    {
      "entropy": 5.699306201934815,
      "epoch": 0.4466057959507741,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986351818367959,
      "loss": 5.7232,
      "mean_token_accuracy": 0.13762547001242637,
      "num_tokens": 9887616.0,
      "step": 4500
    },
    {
      "entropy": 5.774582195281982,
      "epoch": 0.44710202461294163,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986312824172566,
      "loss": 5.6812,
      "mean_token_accuracy": 0.13563110381364823,
      "num_tokens": 9898979.0,
      "step": 4505
    },
    {
      "entropy": 5.78033857345581,
      "epoch": 0.44759825327510916,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986273774521433,
      "loss": 5.7292,
      "mean_token_accuracy": 0.12634325847029687,
      "num_tokens": 9912158.0,
      "step": 4510
    },
    {
      "entropy": 5.724444484710693,
      "epoch": 0.4480944819372767,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004986234669415529,
      "loss": 5.7164,
      "mean_token_accuracy": 0.13096265345811844,
      "num_tokens": 9923886.0,
      "step": 4515
    },
    {
      "entropy": 5.743840265274048,
      "epoch": 0.4485907105994442,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986195508855824,
      "loss": 5.6961,
      "mean_token_accuracy": 0.13753404170274736,
      "num_tokens": 9934979.0,
      "step": 4520
    },
    {
      "entropy": 5.798821640014649,
      "epoch": 0.44908693926161175,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498615629284329,
      "loss": 5.7964,
      "mean_token_accuracy": 0.1303584486246109,
      "num_tokens": 9944715.0,
      "step": 4525
    },
    {
      "entropy": 5.751795530319214,
      "epoch": 0.4495831679237793,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986117021378896,
      "loss": 5.6463,
      "mean_token_accuracy": 0.12978143915534018,
      "num_tokens": 9954736.0,
      "step": 4530
    },
    {
      "entropy": 5.774712133407593,
      "epoch": 0.4500793965859468,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000498607769446362,
      "loss": 5.7416,
      "mean_token_accuracy": 0.13690666407346724,
      "num_tokens": 9965952.0,
      "step": 4535
    },
    {
      "entropy": 5.77537169456482,
      "epoch": 0.45057562524811434,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004986038312098435,
      "loss": 5.6995,
      "mean_token_accuracy": 0.13444727882742882,
      "num_tokens": 9976428.0,
      "step": 4540
    },
    {
      "entropy": 5.748169374465943,
      "epoch": 0.4510718539102819,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985998874284319,
      "loss": 5.6355,
      "mean_token_accuracy": 0.1407597541809082,
      "num_tokens": 9986100.0,
      "step": 4545
    },
    {
      "entropy": 5.620346879959106,
      "epoch": 0.4515680825724494,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004985959381022248,
      "loss": 5.6185,
      "mean_token_accuracy": 0.1355469599366188,
      "num_tokens": 9996584.0,
      "step": 4550
    },
    {
      "entropy": 5.725736522674561,
      "epoch": 0.45206431123461693,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004985919832313203,
      "loss": 5.6334,
      "mean_token_accuracy": 0.1394061729311943,
      "num_tokens": 10007792.0,
      "step": 4555
    },
    {
      "entropy": 5.884342670440674,
      "epoch": 0.45256053989678446,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985880228158164,
      "loss": 5.8379,
      "mean_token_accuracy": 0.12560807019472123,
      "num_tokens": 10019615.0,
      "step": 4560
    },
    {
      "entropy": 5.680107164382934,
      "epoch": 0.45305676855895194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004985840568558112,
      "loss": 5.6466,
      "mean_token_accuracy": 0.1384553760290146,
      "num_tokens": 10030223.0,
      "step": 4565
    },
    {
      "entropy": 5.679829692840576,
      "epoch": 0.45355299722111947,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985800853514033,
      "loss": 5.6585,
      "mean_token_accuracy": 0.13285320699214936,
      "num_tokens": 10041522.0,
      "step": 4570
    },
    {
      "entropy": 5.752073955535889,
      "epoch": 0.454049225883287,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985761083026911,
      "loss": 5.6156,
      "mean_token_accuracy": 0.13825916424393653,
      "num_tokens": 10052847.0,
      "step": 4575
    },
    {
      "entropy": 5.654062080383301,
      "epoch": 0.45454545454545453,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004985721257097731,
      "loss": 5.63,
      "mean_token_accuracy": 0.1342723213136196,
      "num_tokens": 10062259.0,
      "step": 4580
    },
    {
      "entropy": 5.811122560501099,
      "epoch": 0.45504168320762206,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004985681375727482,
      "loss": 5.7899,
      "mean_token_accuracy": 0.1263584315776825,
      "num_tokens": 10073341.0,
      "step": 4585
    },
    {
      "entropy": 5.754662370681762,
      "epoch": 0.4555379118697896,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985641438917151,
      "loss": 5.6554,
      "mean_token_accuracy": 0.13732756152749062,
      "num_tokens": 10083992.0,
      "step": 4590
    },
    {
      "entropy": 5.713476657867432,
      "epoch": 0.4560341405319571,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985601446667731,
      "loss": 5.6417,
      "mean_token_accuracy": 0.13909700736403466,
      "num_tokens": 10094245.0,
      "step": 4595
    },
    {
      "entropy": 5.715325260162354,
      "epoch": 0.45653036919412465,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004985561398980213,
      "loss": 5.6345,
      "mean_token_accuracy": 0.13592055737972258,
      "num_tokens": 10105976.0,
      "step": 4600
    },
    {
      "entropy": 5.653140258789063,
      "epoch": 0.4570265978562922,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985521295855588,
      "loss": 5.7124,
      "mean_token_accuracy": 0.13366659879684448,
      "num_tokens": 10117373.0,
      "step": 4605
    },
    {
      "entropy": 5.737657976150513,
      "epoch": 0.4575228265184597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985481137294853,
      "loss": 5.6485,
      "mean_token_accuracy": 0.12762533873319626,
      "num_tokens": 10128470.0,
      "step": 4610
    },
    {
      "entropy": 5.667543649673462,
      "epoch": 0.45801905518062724,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985440923299002,
      "loss": 5.6324,
      "mean_token_accuracy": 0.14285274744033813,
      "num_tokens": 10140217.0,
      "step": 4615
    },
    {
      "entropy": 5.714731311798095,
      "epoch": 0.4585152838427948,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004985400653869034,
      "loss": 5.7311,
      "mean_token_accuracy": 0.13409728705883026,
      "num_tokens": 10152308.0,
      "step": 4620
    },
    {
      "entropy": 5.830925703048706,
      "epoch": 0.4590115125049623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004985360329005945,
      "loss": 5.74,
      "mean_token_accuracy": 0.13180570527911187,
      "num_tokens": 10164234.0,
      "step": 4625
    },
    {
      "entropy": 5.854005765914917,
      "epoch": 0.45950774116712984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985319948710738,
      "loss": 5.779,
      "mean_token_accuracy": 0.1286759592592716,
      "num_tokens": 10175287.0,
      "step": 4630
    },
    {
      "entropy": 5.739871406555176,
      "epoch": 0.46000396982929737,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004985279512984413,
      "loss": 5.6726,
      "mean_token_accuracy": 0.1389988750219345,
      "num_tokens": 10184658.0,
      "step": 4635
    },
    {
      "entropy": 5.756584978103637,
      "epoch": 0.46050019849146484,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004985239021827972,
      "loss": 5.6701,
      "mean_token_accuracy": 0.13657966926693915,
      "num_tokens": 10194593.0,
      "step": 4640
    },
    {
      "entropy": 5.682445383071899,
      "epoch": 0.46099642715363237,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498519847524242,
      "loss": 5.5901,
      "mean_token_accuracy": 0.13688813745975495,
      "num_tokens": 10205200.0,
      "step": 4645
    },
    {
      "entropy": 5.660930824279785,
      "epoch": 0.4614926558157999,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985157873228762,
      "loss": 5.6176,
      "mean_token_accuracy": 0.13350621685385705,
      "num_tokens": 10216848.0,
      "step": 4650
    },
    {
      "entropy": 5.756806135177612,
      "epoch": 0.46198888447796743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004985117215788005,
      "loss": 5.6687,
      "mean_token_accuracy": 0.1331123724579811,
      "num_tokens": 10228188.0,
      "step": 4655
    },
    {
      "entropy": 5.7153270721435545,
      "epoch": 0.46248511314013496,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004985076502921158,
      "loss": 5.742,
      "mean_token_accuracy": 0.13385683670639992,
      "num_tokens": 10238253.0,
      "step": 4660
    },
    {
      "entropy": 5.795754623413086,
      "epoch": 0.4629813418023025,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000498503573462923,
      "loss": 5.6746,
      "mean_token_accuracy": 0.1363489605486393,
      "num_tokens": 10248169.0,
      "step": 4665
    },
    {
      "entropy": 5.669560050964355,
      "epoch": 0.46347757046447,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984994910913232,
      "loss": 5.5654,
      "mean_token_accuracy": 0.13844399079680442,
      "num_tokens": 10258962.0,
      "step": 4670
    },
    {
      "entropy": 5.718635940551758,
      "epoch": 0.46397379912663755,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004984954031774175,
      "loss": 5.6804,
      "mean_token_accuracy": 0.13743941560387612,
      "num_tokens": 10270303.0,
      "step": 4675
    },
    {
      "entropy": 5.634719276428223,
      "epoch": 0.4644700277888051,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984913097213075,
      "loss": 5.5954,
      "mean_token_accuracy": 0.13856030106544495,
      "num_tokens": 10281782.0,
      "step": 4680
    },
    {
      "entropy": 5.744327068328857,
      "epoch": 0.4649662564509726,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004984872107230945,
      "loss": 5.6372,
      "mean_token_accuracy": 0.13823656737804413,
      "num_tokens": 10292937.0,
      "step": 4685
    },
    {
      "entropy": 5.709350728988648,
      "epoch": 0.46546248511314015,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004984831061828803,
      "loss": 5.6914,
      "mean_token_accuracy": 0.1315615229308605,
      "num_tokens": 10305443.0,
      "step": 4690
    },
    {
      "entropy": 5.712571716308593,
      "epoch": 0.4659587137753077,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984789961007667,
      "loss": 5.6408,
      "mean_token_accuracy": 0.1365855559706688,
      "num_tokens": 10316489.0,
      "step": 4695
    },
    {
      "entropy": 5.68888144493103,
      "epoch": 0.4664549424374752,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004984748804768554,
      "loss": 5.6768,
      "mean_token_accuracy": 0.13440565168857574,
      "num_tokens": 10326690.0,
      "step": 4700
    },
    {
      "entropy": 5.7611644744873045,
      "epoch": 0.46695117109964274,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004984707593112487,
      "loss": 5.6832,
      "mean_token_accuracy": 0.13210530802607537,
      "num_tokens": 10337059.0,
      "step": 4705
    },
    {
      "entropy": 5.71557879447937,
      "epoch": 0.46744739976181027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004984666326040487,
      "loss": 5.5398,
      "mean_token_accuracy": 0.1411898709833622,
      "num_tokens": 10347035.0,
      "step": 4710
    },
    {
      "entropy": 5.6951854705810545,
      "epoch": 0.46794362842397774,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004984625003553578,
      "loss": 5.6881,
      "mean_token_accuracy": 0.13463103100657464,
      "num_tokens": 10359492.0,
      "step": 4715
    },
    {
      "entropy": 5.743189573287964,
      "epoch": 0.4684398570861453,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004984583625652783,
      "loss": 5.6612,
      "mean_token_accuracy": 0.13574108779430388,
      "num_tokens": 10370128.0,
      "step": 4720
    },
    {
      "entropy": 5.71764907836914,
      "epoch": 0.4689360857483128,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498454219233913,
      "loss": 5.6652,
      "mean_token_accuracy": 0.13303780034184456,
      "num_tokens": 10380276.0,
      "step": 4725
    },
    {
      "entropy": 5.636638927459717,
      "epoch": 0.46943231441048033,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984500703613645,
      "loss": 5.6568,
      "mean_token_accuracy": 0.13573246523737909,
      "num_tokens": 10392489.0,
      "step": 4730
    },
    {
      "entropy": 5.818667268753051,
      "epoch": 0.46992854307264786,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984459159477358,
      "loss": 5.7419,
      "mean_token_accuracy": 0.1304431177675724,
      "num_tokens": 10404259.0,
      "step": 4735
    },
    {
      "entropy": 5.734631967544556,
      "epoch": 0.4704247717348154,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004984417559931298,
      "loss": 5.5915,
      "mean_token_accuracy": 0.13806108832359315,
      "num_tokens": 10415000.0,
      "step": 4740
    },
    {
      "entropy": 5.713548278808593,
      "epoch": 0.4709210003969829,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984375904976498,
      "loss": 5.724,
      "mean_token_accuracy": 0.13026771023869516,
      "num_tokens": 10426078.0,
      "step": 4745
    },
    {
      "entropy": 5.754714393615723,
      "epoch": 0.47141722905915046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984334194613989,
      "loss": 5.66,
      "mean_token_accuracy": 0.13449166119098663,
      "num_tokens": 10436838.0,
      "step": 4750
    },
    {
      "entropy": 5.725654745101929,
      "epoch": 0.471913457721318,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004984292428844809,
      "loss": 5.6405,
      "mean_token_accuracy": 0.13901436030864717,
      "num_tokens": 10448544.0,
      "step": 4755
    },
    {
      "entropy": 5.781258630752563,
      "epoch": 0.4724096863834855,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984250607669989,
      "loss": 5.7021,
      "mean_token_accuracy": 0.1291268825531006,
      "num_tokens": 10459730.0,
      "step": 4760
    },
    {
      "entropy": 5.841650438308716,
      "epoch": 0.47290591504565305,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004984208731090569,
      "loss": 5.7694,
      "mean_token_accuracy": 0.13160821050405502,
      "num_tokens": 10470257.0,
      "step": 4765
    },
    {
      "entropy": 5.788986110687256,
      "epoch": 0.4734021437078206,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984166799107587,
      "loss": 5.7152,
      "mean_token_accuracy": 0.13656962364912034,
      "num_tokens": 10480835.0,
      "step": 4770
    },
    {
      "entropy": 5.691136264801026,
      "epoch": 0.4738983723699881,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004984124811722081,
      "loss": 5.7538,
      "mean_token_accuracy": 0.13584274128079415,
      "num_tokens": 10491720.0,
      "step": 4775
    },
    {
      "entropy": 5.787897634506225,
      "epoch": 0.47439460103215564,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004984082768935096,
      "loss": 5.7623,
      "mean_token_accuracy": 0.1347417004406452,
      "num_tokens": 10502313.0,
      "step": 4780
    },
    {
      "entropy": 5.749018049240112,
      "epoch": 0.47489082969432317,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004984040670747672,
      "loss": 5.6209,
      "mean_token_accuracy": 0.1366601176559925,
      "num_tokens": 10513333.0,
      "step": 4785
    },
    {
      "entropy": 5.652175331115723,
      "epoch": 0.47538705835649064,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004983998517160852,
      "loss": 5.6131,
      "mean_token_accuracy": 0.1452295422554016,
      "num_tokens": 10522766.0,
      "step": 4790
    },
    {
      "entropy": 5.747906017303467,
      "epoch": 0.4758832870186582,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004983956308175684,
      "loss": 5.6529,
      "mean_token_accuracy": 0.1344590365886688,
      "num_tokens": 10534362.0,
      "step": 4795
    },
    {
      "entropy": 5.721209239959717,
      "epoch": 0.4763795156808257,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983914043793212,
      "loss": 5.6664,
      "mean_token_accuracy": 0.13596387356519699,
      "num_tokens": 10545072.0,
      "step": 4800
    },
    {
      "entropy": 5.6192014694213865,
      "epoch": 0.47687574434299324,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004983871724014487,
      "loss": 5.6199,
      "mean_token_accuracy": 0.1320567600429058,
      "num_tokens": 10555953.0,
      "step": 4805
    },
    {
      "entropy": 5.810444021224976,
      "epoch": 0.47737197300516077,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983829348840556,
      "loss": 5.6569,
      "mean_token_accuracy": 0.13939146995544432,
      "num_tokens": 10567757.0,
      "step": 4810
    },
    {
      "entropy": 5.632370805740356,
      "epoch": 0.4778682016673283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498378691827247,
      "loss": 5.5607,
      "mean_token_accuracy": 0.13832593187689782,
      "num_tokens": 10579390.0,
      "step": 4815
    },
    {
      "entropy": 5.69977650642395,
      "epoch": 0.4783644303294958,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004983744432311282,
      "loss": 5.7685,
      "mean_token_accuracy": 0.133011906594038,
      "num_tokens": 10589624.0,
      "step": 4820
    },
    {
      "entropy": 5.833197927474975,
      "epoch": 0.47886065899166336,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983701890958046,
      "loss": 5.7341,
      "mean_token_accuracy": 0.1273055300116539,
      "num_tokens": 10601927.0,
      "step": 4825
    },
    {
      "entropy": 5.666011571884155,
      "epoch": 0.4793568876538309,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983659294213815,
      "loss": 5.6065,
      "mean_token_accuracy": 0.13757556453347206,
      "num_tokens": 10614073.0,
      "step": 4830
    },
    {
      "entropy": 5.661135864257813,
      "epoch": 0.4798531163159984,
      "grad_norm": 1.125,
      "learning_rate": 0.0004983616642079649,
      "loss": 5.6301,
      "mean_token_accuracy": 0.1396137222647667,
      "num_tokens": 10625535.0,
      "step": 4835
    },
    {
      "entropy": 5.815369939804077,
      "epoch": 0.48034934497816595,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004983573934556601,
      "loss": 5.614,
      "mean_token_accuracy": 0.13902199566364287,
      "num_tokens": 10637160.0,
      "step": 4840
    },
    {
      "entropy": 5.777980327606201,
      "epoch": 0.4808455736403335,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004983531171645734,
      "loss": 5.713,
      "mean_token_accuracy": 0.1307061031460762,
      "num_tokens": 10649061.0,
      "step": 4845
    },
    {
      "entropy": 5.596870040893554,
      "epoch": 0.481341802302501,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004983488353348106,
      "loss": 5.574,
      "mean_token_accuracy": 0.13727790862321854,
      "num_tokens": 10661130.0,
      "step": 4850
    },
    {
      "entropy": 5.691119718551636,
      "epoch": 0.48183803096466854,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004983445479664779,
      "loss": 5.6324,
      "mean_token_accuracy": 0.13580886498093606,
      "num_tokens": 10673017.0,
      "step": 4855
    },
    {
      "entropy": 5.755751132965088,
      "epoch": 0.48233425962683607,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004983402550596817,
      "loss": 5.7026,
      "mean_token_accuracy": 0.13476815149188043,
      "num_tokens": 10683268.0,
      "step": 4860
    },
    {
      "entropy": 5.768578052520752,
      "epoch": 0.48283048828900355,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004983359566145285,
      "loss": 5.7126,
      "mean_token_accuracy": 0.1333691321313381,
      "num_tokens": 10694220.0,
      "step": 4865
    },
    {
      "entropy": 5.73603048324585,
      "epoch": 0.4833267169511711,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983316526311248,
      "loss": 5.6282,
      "mean_token_accuracy": 0.1314948871731758,
      "num_tokens": 10704829.0,
      "step": 4870
    },
    {
      "entropy": 5.730405712127686,
      "epoch": 0.4838229456133386,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004983273431095773,
      "loss": 5.8154,
      "mean_token_accuracy": 0.1304888866841793,
      "num_tokens": 10717160.0,
      "step": 4875
    },
    {
      "entropy": 5.7811564922332765,
      "epoch": 0.48431917427550614,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004983230280499929,
      "loss": 5.6262,
      "mean_token_accuracy": 0.13455311432480813,
      "num_tokens": 10728638.0,
      "step": 4880
    },
    {
      "entropy": 5.6971851825714115,
      "epoch": 0.48481540293767367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983187074524787,
      "loss": 5.6748,
      "mean_token_accuracy": 0.13421054780483246,
      "num_tokens": 10740122.0,
      "step": 4885
    },
    {
      "entropy": 5.73472146987915,
      "epoch": 0.4853116315998412,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004983143813171418,
      "loss": 5.6434,
      "mean_token_accuracy": 0.14405334889888763,
      "num_tokens": 10751103.0,
      "step": 4890
    },
    {
      "entropy": 5.732849550247193,
      "epoch": 0.48580786026200873,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004983100496440895,
      "loss": 5.5982,
      "mean_token_accuracy": 0.13946570307016373,
      "num_tokens": 10761937.0,
      "step": 4895
    },
    {
      "entropy": 5.759403657913208,
      "epoch": 0.48630408892417626,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498305712433429,
      "loss": 5.8191,
      "mean_token_accuracy": 0.1310167983174324,
      "num_tokens": 10773527.0,
      "step": 4900
    },
    {
      "entropy": 5.745750427246094,
      "epoch": 0.4868003175863438,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000498301369685268,
      "loss": 5.6581,
      "mean_token_accuracy": 0.13396176546812058,
      "num_tokens": 10783873.0,
      "step": 4905
    },
    {
      "entropy": 5.6830870628356935,
      "epoch": 0.4872965462485113,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982970213997142,
      "loss": 5.5051,
      "mean_token_accuracy": 0.14291448444128035,
      "num_tokens": 10796124.0,
      "step": 4910
    },
    {
      "entropy": 5.6431303977966305,
      "epoch": 0.48779277491067885,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982926675768756,
      "loss": 5.6092,
      "mean_token_accuracy": 0.1361992783844471,
      "num_tokens": 10806468.0,
      "step": 4915
    },
    {
      "entropy": 5.720559692382812,
      "epoch": 0.4882890035728464,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982883082168599,
      "loss": 5.6515,
      "mean_token_accuracy": 0.13573368415236473,
      "num_tokens": 10817126.0,
      "step": 4920
    },
    {
      "entropy": 5.731108808517456,
      "epoch": 0.4887852322350139,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004982839433197752,
      "loss": 5.6821,
      "mean_token_accuracy": 0.12980374544858933,
      "num_tokens": 10828985.0,
      "step": 4925
    },
    {
      "entropy": 5.688351011276245,
      "epoch": 0.48928146089718144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049827957288573,
      "loss": 5.6058,
      "mean_token_accuracy": 0.13472920954227446,
      "num_tokens": 10840528.0,
      "step": 4930
    },
    {
      "entropy": 5.721200561523437,
      "epoch": 0.48977768955934897,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004982751969148324,
      "loss": 5.6168,
      "mean_token_accuracy": 0.13815968558192254,
      "num_tokens": 10851312.0,
      "step": 4935
    },
    {
      "entropy": 5.662455034255982,
      "epoch": 0.49027391822151645,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004982708154071912,
      "loss": 5.5795,
      "mean_token_accuracy": 0.1390117309987545,
      "num_tokens": 10861882.0,
      "step": 4940
    },
    {
      "entropy": 5.7047789096832275,
      "epoch": 0.490770146883684,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004982664283629147,
      "loss": 5.7346,
      "mean_token_accuracy": 0.13037165477871895,
      "num_tokens": 10873187.0,
      "step": 4945
    },
    {
      "entropy": 5.743262815475464,
      "epoch": 0.4912663755458515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982620357821118,
      "loss": 5.6135,
      "mean_token_accuracy": 0.14105834290385247,
      "num_tokens": 10884254.0,
      "step": 4950
    },
    {
      "entropy": 5.622763061523438,
      "epoch": 0.49176260420801904,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004982576376648918,
      "loss": 5.615,
      "mean_token_accuracy": 0.13944189846515656,
      "num_tokens": 10895770.0,
      "step": 4955
    },
    {
      "entropy": 5.711516904830932,
      "epoch": 0.49225883287018657,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982532340113633,
      "loss": 5.6244,
      "mean_token_accuracy": 0.14097289890050888,
      "num_tokens": 10906436.0,
      "step": 4960
    },
    {
      "entropy": 5.741786670684815,
      "epoch": 0.4927550615323541,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982488248216357,
      "loss": 5.5906,
      "mean_token_accuracy": 0.1392080157995224,
      "num_tokens": 10917655.0,
      "step": 4965
    },
    {
      "entropy": 5.617937469482422,
      "epoch": 0.49325129019452163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982444100958182,
      "loss": 5.6049,
      "mean_token_accuracy": 0.14341603815555573,
      "num_tokens": 10928359.0,
      "step": 4970
    },
    {
      "entropy": 5.716538715362549,
      "epoch": 0.49374751885668916,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004982399898340205,
      "loss": 5.7679,
      "mean_token_accuracy": 0.1270577371120453,
      "num_tokens": 10940494.0,
      "step": 4975
    },
    {
      "entropy": 5.857838869094849,
      "epoch": 0.4942437475188567,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498235564036352,
      "loss": 5.7631,
      "mean_token_accuracy": 0.1268538475036621,
      "num_tokens": 10951298.0,
      "step": 4980
    },
    {
      "entropy": 5.717168092727661,
      "epoch": 0.4947399761810242,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004982311327029226,
      "loss": 5.6547,
      "mean_token_accuracy": 0.13692860156297684,
      "num_tokens": 10961680.0,
      "step": 4985
    },
    {
      "entropy": 5.787375116348267,
      "epoch": 0.49523620484319175,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004982266958338421,
      "loss": 5.7866,
      "mean_token_accuracy": 0.13215267434716224,
      "num_tokens": 10974968.0,
      "step": 4990
    },
    {
      "entropy": 5.722283601760864,
      "epoch": 0.4957324335053593,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982222534292205,
      "loss": 5.715,
      "mean_token_accuracy": 0.12861865088343621,
      "num_tokens": 10986383.0,
      "step": 4995
    },
    {
      "entropy": 5.815975666046143,
      "epoch": 0.4962286621675268,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004982178054891681,
      "loss": 5.7407,
      "mean_token_accuracy": 0.1366094656288624,
      "num_tokens": 10998384.0,
      "step": 5000
    },
    {
      "entropy": 5.669220447540283,
      "epoch": 0.49672489082969434,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004982133520137951,
      "loss": 5.5079,
      "mean_token_accuracy": 0.14211993515491486,
      "num_tokens": 11009420.0,
      "step": 5005
    },
    {
      "entropy": 5.699658346176148,
      "epoch": 0.4972211194918619,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982088930032119,
      "loss": 5.6802,
      "mean_token_accuracy": 0.1344943329691887,
      "num_tokens": 11019601.0,
      "step": 5010
    },
    {
      "entropy": 5.678679418563843,
      "epoch": 0.4977173481540294,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004982044284575293,
      "loss": 5.6103,
      "mean_token_accuracy": 0.13872682452201843,
      "num_tokens": 11031138.0,
      "step": 5015
    },
    {
      "entropy": 5.736675214767456,
      "epoch": 0.4982135768161969,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981999583768576,
      "loss": 5.6224,
      "mean_token_accuracy": 0.13761773854494094,
      "num_tokens": 11041643.0,
      "step": 5020
    },
    {
      "entropy": 5.688022089004517,
      "epoch": 0.4987098054783644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498195482761308,
      "loss": 5.6359,
      "mean_token_accuracy": 0.12926055118441582,
      "num_tokens": 11052116.0,
      "step": 5025
    },
    {
      "entropy": 5.82230315208435,
      "epoch": 0.49920603414053194,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004981910016109914,
      "loss": 5.8455,
      "mean_token_accuracy": 0.12513070702552795,
      "num_tokens": 11064796.0,
      "step": 5030
    },
    {
      "entropy": 5.823551273345947,
      "epoch": 0.49970226280269947,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004981865149260189,
      "loss": 5.6722,
      "mean_token_accuracy": 0.13358607441186904,
      "num_tokens": 11075029.0,
      "step": 5035
    },
    {
      "entropy": 5.739116048812866,
      "epoch": 0.5001984914648671,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981820227065016,
      "loss": 5.6742,
      "mean_token_accuracy": 0.13163042291998864,
      "num_tokens": 11086702.0,
      "step": 5040
    },
    {
      "entropy": 5.684933805465699,
      "epoch": 0.5006947201270345,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981775249525513,
      "loss": 5.6758,
      "mean_token_accuracy": 0.13385781049728393,
      "num_tokens": 11097417.0,
      "step": 5045
    },
    {
      "entropy": 5.679539537429809,
      "epoch": 0.5011909487892021,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004981730216642791,
      "loss": 5.5834,
      "mean_token_accuracy": 0.14295208007097243,
      "num_tokens": 11108458.0,
      "step": 5050
    },
    {
      "entropy": 5.757784986495972,
      "epoch": 0.5016871774513696,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981685128417969,
      "loss": 5.7534,
      "mean_token_accuracy": 0.12978242710232735,
      "num_tokens": 11121466.0,
      "step": 5055
    },
    {
      "entropy": 5.712614822387695,
      "epoch": 0.5021834061135371,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004981639984852165,
      "loss": 5.7019,
      "mean_token_accuracy": 0.12968832328915597,
      "num_tokens": 11132489.0,
      "step": 5060
    },
    {
      "entropy": 5.697144508361816,
      "epoch": 0.5026796347757047,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004981594785946498,
      "loss": 5.6217,
      "mean_token_accuracy": 0.13766293674707414,
      "num_tokens": 11143087.0,
      "step": 5065
    },
    {
      "entropy": 5.766337108612061,
      "epoch": 0.5031758634378721,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004981549531702089,
      "loss": 5.6869,
      "mean_token_accuracy": 0.13343412950634956,
      "num_tokens": 11154229.0,
      "step": 5070
    },
    {
      "entropy": 5.72542028427124,
      "epoch": 0.5036720921000397,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498150422212006,
      "loss": 5.653,
      "mean_token_accuracy": 0.13460973873734475,
      "num_tokens": 11165698.0,
      "step": 5075
    },
    {
      "entropy": 5.74044508934021,
      "epoch": 0.5041683207622072,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004981458857201534,
      "loss": 5.6408,
      "mean_token_accuracy": 0.1373266398906708,
      "num_tokens": 11177025.0,
      "step": 5080
    },
    {
      "entropy": 5.747441053390503,
      "epoch": 0.5046645494243748,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004981413436947639,
      "loss": 5.7135,
      "mean_token_accuracy": 0.13659801632165908,
      "num_tokens": 11188766.0,
      "step": 5085
    },
    {
      "entropy": 5.725644111633301,
      "epoch": 0.5051607780865423,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981367961359496,
      "loss": 5.7435,
      "mean_token_accuracy": 0.12596021741628646,
      "num_tokens": 11201626.0,
      "step": 5090
    },
    {
      "entropy": 5.773015403747559,
      "epoch": 0.5056570067487098,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004981322430438239,
      "loss": 5.6693,
      "mean_token_accuracy": 0.13133040070533752,
      "num_tokens": 11213975.0,
      "step": 5095
    },
    {
      "entropy": 5.69825325012207,
      "epoch": 0.5061532354108773,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004981276844184991,
      "loss": 5.6164,
      "mean_token_accuracy": 0.14137618467211724,
      "num_tokens": 11223636.0,
      "step": 5100
    },
    {
      "entropy": 5.637740516662598,
      "epoch": 0.5066494640730449,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981231202600886,
      "loss": 5.5744,
      "mean_token_accuracy": 0.14177348613739013,
      "num_tokens": 11233549.0,
      "step": 5105
    },
    {
      "entropy": 5.65978512763977,
      "epoch": 0.5071456927352124,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004981185505687056,
      "loss": 5.5895,
      "mean_token_accuracy": 0.1379475213587284,
      "num_tokens": 11242915.0,
      "step": 5110
    },
    {
      "entropy": 5.7940751075744625,
      "epoch": 0.50764192139738,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981139753444632,
      "loss": 5.7119,
      "mean_token_accuracy": 0.13013848885893822,
      "num_tokens": 11254370.0,
      "step": 5115
    },
    {
      "entropy": 5.705103826522827,
      "epoch": 0.5081381500595474,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004981093945874752,
      "loss": 5.6356,
      "mean_token_accuracy": 0.14226910769939421,
      "num_tokens": 11264226.0,
      "step": 5120
    },
    {
      "entropy": 5.608145618438721,
      "epoch": 0.508634378721715,
      "grad_norm": 1.125,
      "learning_rate": 0.0004981048082978547,
      "loss": 5.538,
      "mean_token_accuracy": 0.14511876031756402,
      "num_tokens": 11275304.0,
      "step": 5125
    },
    {
      "entropy": 5.62836709022522,
      "epoch": 0.5091306073838825,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004981002164757158,
      "loss": 5.609,
      "mean_token_accuracy": 0.13693601936101912,
      "num_tokens": 11286179.0,
      "step": 5130
    },
    {
      "entropy": 5.747485017776489,
      "epoch": 0.50962683604605,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980956191211723,
      "loss": 5.6286,
      "mean_token_accuracy": 0.1369728185236454,
      "num_tokens": 11297244.0,
      "step": 5135
    },
    {
      "entropy": 5.7410609245300295,
      "epoch": 0.5101230647082176,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004980910162343382,
      "loss": 5.6362,
      "mean_token_accuracy": 0.13777609616518022,
      "num_tokens": 11307251.0,
      "step": 5140
    },
    {
      "entropy": 5.737069845199585,
      "epoch": 0.510619293370385,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980864078153275,
      "loss": 5.6163,
      "mean_token_accuracy": 0.13710779026150705,
      "num_tokens": 11318441.0,
      "step": 5145
    },
    {
      "entropy": 5.618476247787475,
      "epoch": 0.5111155220325526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004980817938642548,
      "loss": 5.6013,
      "mean_token_accuracy": 0.1396390289068222,
      "num_tokens": 11328743.0,
      "step": 5150
    },
    {
      "entropy": 5.592562484741211,
      "epoch": 0.5116117506947201,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498077174381234,
      "loss": 5.5838,
      "mean_token_accuracy": 0.13907210081815718,
      "num_tokens": 11340745.0,
      "step": 5155
    },
    {
      "entropy": 5.721065616607666,
      "epoch": 0.5121079793568877,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004980725493663802,
      "loss": 5.6591,
      "mean_token_accuracy": 0.13914958611130715,
      "num_tokens": 11352060.0,
      "step": 5160
    },
    {
      "entropy": 5.704443025588989,
      "epoch": 0.5126042080190552,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004980679188198078,
      "loss": 5.666,
      "mean_token_accuracy": 0.1363778106868267,
      "num_tokens": 11363097.0,
      "step": 5165
    },
    {
      "entropy": 5.727260255813599,
      "epoch": 0.5131004366812227,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004980632827416317,
      "loss": 5.6933,
      "mean_token_accuracy": 0.14730945602059364,
      "num_tokens": 11373410.0,
      "step": 5170
    },
    {
      "entropy": 5.768536138534546,
      "epoch": 0.5135966653433902,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004980586411319669,
      "loss": 5.7924,
      "mean_token_accuracy": 0.1284305416047573,
      "num_tokens": 11385615.0,
      "step": 5175
    },
    {
      "entropy": 5.652726125717163,
      "epoch": 0.5140928940055578,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980539939909283,
      "loss": 5.5513,
      "mean_token_accuracy": 0.13596862852573394,
      "num_tokens": 11396492.0,
      "step": 5180
    },
    {
      "entropy": 5.777144432067871,
      "epoch": 0.5145891226677253,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980493413186314,
      "loss": 5.6847,
      "mean_token_accuracy": 0.13327620327472686,
      "num_tokens": 11406377.0,
      "step": 5185
    },
    {
      "entropy": 5.754715394973755,
      "epoch": 0.5150853513298929,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004980446831151914,
      "loss": 5.6536,
      "mean_token_accuracy": 0.13574032932519914,
      "num_tokens": 11416862.0,
      "step": 5190
    },
    {
      "entropy": 5.678183650970459,
      "epoch": 0.5155815799920603,
      "grad_norm": 1.125,
      "learning_rate": 0.0004980400193807239,
      "loss": 5.5927,
      "mean_token_accuracy": 0.13788124099373816,
      "num_tokens": 11428246.0,
      "step": 5195
    },
    {
      "entropy": 5.726457595825195,
      "epoch": 0.5160778086542279,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004980353501153446,
      "loss": 5.6287,
      "mean_token_accuracy": 0.13608812317252159,
      "num_tokens": 11438147.0,
      "step": 5200
    },
    {
      "entropy": 5.77124056816101,
      "epoch": 0.5165740373163954,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004980306753191691,
      "loss": 5.7454,
      "mean_token_accuracy": 0.12761850655078888,
      "num_tokens": 11449154.0,
      "step": 5205
    },
    {
      "entropy": 5.737915945053101,
      "epoch": 0.5170702659785629,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004980259949923135,
      "loss": 5.6579,
      "mean_token_accuracy": 0.1306302972137928,
      "num_tokens": 11459262.0,
      "step": 5210
    },
    {
      "entropy": 5.699729919433594,
      "epoch": 0.5175664946407305,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980213091348938,
      "loss": 5.6464,
      "mean_token_accuracy": 0.13351952880620957,
      "num_tokens": 11471991.0,
      "step": 5215
    },
    {
      "entropy": 5.65016541481018,
      "epoch": 0.5180627233028979,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004980166177470262,
      "loss": 5.5818,
      "mean_token_accuracy": 0.13733325749635697,
      "num_tokens": 11483176.0,
      "step": 5220
    },
    {
      "entropy": 5.730344581604004,
      "epoch": 0.5185589519650655,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004980119208288271,
      "loss": 5.6823,
      "mean_token_accuracy": 0.13771358504891396,
      "num_tokens": 11494015.0,
      "step": 5225
    },
    {
      "entropy": 5.649020433425903,
      "epoch": 0.519055180627233,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980072183804128,
      "loss": 5.5902,
      "mean_token_accuracy": 0.13536551743745803,
      "num_tokens": 11506365.0,
      "step": 5230
    },
    {
      "entropy": 5.696982908248901,
      "epoch": 0.5195514092894006,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004980025104019,
      "loss": 5.6678,
      "mean_token_accuracy": 0.13850680142641067,
      "num_tokens": 11517199.0,
      "step": 5235
    },
    {
      "entropy": 5.665837192535401,
      "epoch": 0.520047637951568,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004979977968934056,
      "loss": 5.5321,
      "mean_token_accuracy": 0.14265239983797073,
      "num_tokens": 11527591.0,
      "step": 5240
    },
    {
      "entropy": 5.718610525131226,
      "epoch": 0.5205438666137356,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004979930778550463,
      "loss": 5.7121,
      "mean_token_accuracy": 0.13773501440882682,
      "num_tokens": 11538657.0,
      "step": 5245
    },
    {
      "entropy": 5.688341856002808,
      "epoch": 0.5210400952759031,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004979883532869392,
      "loss": 5.5253,
      "mean_token_accuracy": 0.1551755487918854,
      "num_tokens": 11549513.0,
      "step": 5250
    },
    {
      "entropy": 5.6611675262451175,
      "epoch": 0.5215363239380707,
      "grad_norm": 1.25,
      "learning_rate": 0.0004979836231892013,
      "loss": 5.619,
      "mean_token_accuracy": 0.1370332196354866,
      "num_tokens": 11560376.0,
      "step": 5255
    },
    {
      "entropy": 5.632587194442749,
      "epoch": 0.5220325526002382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979788875619501,
      "loss": 5.592,
      "mean_token_accuracy": 0.13320594951510428,
      "num_tokens": 11571218.0,
      "step": 5260
    },
    {
      "entropy": 5.762354516983033,
      "epoch": 0.5225287812624058,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000497974146405303,
      "loss": 5.654,
      "mean_token_accuracy": 0.1349178060889244,
      "num_tokens": 11581981.0,
      "step": 5265
    },
    {
      "entropy": 5.6612245559692385,
      "epoch": 0.5230250099245732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979693997193776,
      "loss": 5.6226,
      "mean_token_accuracy": 0.13676370978355407,
      "num_tokens": 11592561.0,
      "step": 5270
    },
    {
      "entropy": 5.706000423431396,
      "epoch": 0.5235212385867408,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004979646475042914,
      "loss": 5.5981,
      "mean_token_accuracy": 0.13818725645542146,
      "num_tokens": 11601933.0,
      "step": 5275
    },
    {
      "entropy": 5.635031986236572,
      "epoch": 0.5240174672489083,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979598897601625,
      "loss": 5.5193,
      "mean_token_accuracy": 0.14932274371385573,
      "num_tokens": 11613938.0,
      "step": 5280
    },
    {
      "entropy": 5.739613437652588,
      "epoch": 0.5245136959110758,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004979551264871088,
      "loss": 5.7201,
      "mean_token_accuracy": 0.1296633265912533,
      "num_tokens": 11624890.0,
      "step": 5285
    },
    {
      "entropy": 5.702748775482178,
      "epoch": 0.5250099245732434,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979503576852482,
      "loss": 5.6881,
      "mean_token_accuracy": 0.13752070665359498,
      "num_tokens": 11635080.0,
      "step": 5290
    },
    {
      "entropy": 5.576523113250732,
      "epoch": 0.5255061532354108,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004979455833546994,
      "loss": 5.5969,
      "mean_token_accuracy": 0.13823814019560815,
      "num_tokens": 11645180.0,
      "step": 5295
    },
    {
      "entropy": 5.721831560134888,
      "epoch": 0.5260023818975784,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979408034955804,
      "loss": 5.6493,
      "mean_token_accuracy": 0.13646110743284226,
      "num_tokens": 11657599.0,
      "step": 5300
    },
    {
      "entropy": 5.783962821960449,
      "epoch": 0.5264986105597459,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979360181080099,
      "loss": 5.6992,
      "mean_token_accuracy": 0.12464715465903282,
      "num_tokens": 11669177.0,
      "step": 5305
    },
    {
      "entropy": 5.630827951431274,
      "epoch": 0.5269948392219135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979312271921065,
      "loss": 5.5352,
      "mean_token_accuracy": 0.13785774409770965,
      "num_tokens": 11679756.0,
      "step": 5310
    },
    {
      "entropy": 5.6718080043792725,
      "epoch": 0.527491067884081,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497926430747989,
      "loss": 5.599,
      "mean_token_accuracy": 0.1393997550010681,
      "num_tokens": 11689805.0,
      "step": 5315
    },
    {
      "entropy": 5.688294124603272,
      "epoch": 0.5279872965462485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004979216287757764,
      "loss": 5.641,
      "mean_token_accuracy": 0.13995739743113517,
      "num_tokens": 11701150.0,
      "step": 5320
    },
    {
      "entropy": 5.594429349899292,
      "epoch": 0.528483525208416,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497916821275588,
      "loss": 5.5622,
      "mean_token_accuracy": 0.1381341226398945,
      "num_tokens": 11712256.0,
      "step": 5325
    },
    {
      "entropy": 5.716231107711792,
      "epoch": 0.5289797538705836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979120082475424,
      "loss": 5.5984,
      "mean_token_accuracy": 0.14262946993112563,
      "num_tokens": 11722547.0,
      "step": 5330
    },
    {
      "entropy": 5.754014587402343,
      "epoch": 0.5294759825327511,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004979071896917595,
      "loss": 5.6221,
      "mean_token_accuracy": 0.13919249400496483,
      "num_tokens": 11732309.0,
      "step": 5335
    },
    {
      "entropy": 5.730963134765625,
      "epoch": 0.5299722111949187,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004979023656083586,
      "loss": 5.6814,
      "mean_token_accuracy": 0.13742681220173836,
      "num_tokens": 11743982.0,
      "step": 5340
    },
    {
      "entropy": 5.62978343963623,
      "epoch": 0.5304684398570861,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004978975359974594,
      "loss": 5.6221,
      "mean_token_accuracy": 0.1360566258430481,
      "num_tokens": 11755506.0,
      "step": 5345
    },
    {
      "entropy": 5.712881660461425,
      "epoch": 0.5309646685192537,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004978927008591816,
      "loss": 5.6342,
      "mean_token_accuracy": 0.13947277292609214,
      "num_tokens": 11767024.0,
      "step": 5350
    },
    {
      "entropy": 5.759839773178101,
      "epoch": 0.5314608971814212,
      "grad_norm": 1.125,
      "learning_rate": 0.0004978878601936451,
      "loss": 5.6332,
      "mean_token_accuracy": 0.13160441145300866,
      "num_tokens": 11777625.0,
      "step": 5355
    },
    {
      "entropy": 5.656109571456909,
      "epoch": 0.5319571258435887,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004978830140009698,
      "loss": 5.5953,
      "mean_token_accuracy": 0.14661457240581513,
      "num_tokens": 11788610.0,
      "step": 5360
    },
    {
      "entropy": 5.680585765838623,
      "epoch": 0.5324533545057563,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004978781622812762,
      "loss": 5.7022,
      "mean_token_accuracy": 0.14207389056682587,
      "num_tokens": 11800828.0,
      "step": 5365
    },
    {
      "entropy": 5.801745271682739,
      "epoch": 0.5329495831679237,
      "grad_norm": 1.125,
      "learning_rate": 0.0004978733050346844,
      "loss": 5.6902,
      "mean_token_accuracy": 0.13995519876480103,
      "num_tokens": 11811729.0,
      "step": 5370
    },
    {
      "entropy": 5.692205333709717,
      "epoch": 0.5334458118300913,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004978684422613149,
      "loss": 5.7299,
      "mean_token_accuracy": 0.13896377608180047,
      "num_tokens": 11822081.0,
      "step": 5375
    },
    {
      "entropy": 5.687965822219849,
      "epoch": 0.5339420404922588,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978635739612882,
      "loss": 5.5952,
      "mean_token_accuracy": 0.13786420002579688,
      "num_tokens": 11832984.0,
      "step": 5380
    },
    {
      "entropy": 5.674173641204834,
      "epoch": 0.5344382691544264,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004978587001347252,
      "loss": 5.5537,
      "mean_token_accuracy": 0.13452147021889688,
      "num_tokens": 11842918.0,
      "step": 5385
    },
    {
      "entropy": 5.637296724319458,
      "epoch": 0.5349344978165939,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978538207817465,
      "loss": 5.5927,
      "mean_token_accuracy": 0.13891708105802536,
      "num_tokens": 11854364.0,
      "step": 5390
    },
    {
      "entropy": 5.69665207862854,
      "epoch": 0.5354307264787614,
      "grad_norm": 1.125,
      "learning_rate": 0.0004978489359024734,
      "loss": 5.6741,
      "mean_token_accuracy": 0.13736676722764968,
      "num_tokens": 11866400.0,
      "step": 5395
    },
    {
      "entropy": 5.77194619178772,
      "epoch": 0.5359269551409289,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978440454970268,
      "loss": 5.6643,
      "mean_token_accuracy": 0.1373192198574543,
      "num_tokens": 11877094.0,
      "step": 5400
    },
    {
      "entropy": 5.712098503112793,
      "epoch": 0.5364231838030965,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004978391495655281,
      "loss": 5.6265,
      "mean_token_accuracy": 0.13823723345994948,
      "num_tokens": 11887289.0,
      "step": 5405
    },
    {
      "entropy": 5.696305465698242,
      "epoch": 0.536919412465264,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004978342481080987,
      "loss": 5.6021,
      "mean_token_accuracy": 0.13555959314107896,
      "num_tokens": 11897700.0,
      "step": 5410
    },
    {
      "entropy": 5.591624641418457,
      "epoch": 0.5374156411274316,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004978293411248601,
      "loss": 5.5265,
      "mean_token_accuracy": 0.14417028427124023,
      "num_tokens": 11908623.0,
      "step": 5415
    },
    {
      "entropy": 5.655980730056763,
      "epoch": 0.537911869789599,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004978244286159339,
      "loss": 5.5597,
      "mean_token_accuracy": 0.14498542994260788,
      "num_tokens": 11917883.0,
      "step": 5420
    },
    {
      "entropy": 5.656184291839599,
      "epoch": 0.5384080984517666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978195105814421,
      "loss": 5.5927,
      "mean_token_accuracy": 0.13965989127755166,
      "num_tokens": 11929145.0,
      "step": 5425
    },
    {
      "entropy": 5.682995748519898,
      "epoch": 0.5389043271139341,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978145870215065,
      "loss": 5.577,
      "mean_token_accuracy": 0.13799101039767264,
      "num_tokens": 11940073.0,
      "step": 5430
    },
    {
      "entropy": 5.711254978179932,
      "epoch": 0.5394005557761016,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004978096579362493,
      "loss": 5.7123,
      "mean_token_accuracy": 0.13857175707817077,
      "num_tokens": 11953140.0,
      "step": 5435
    },
    {
      "entropy": 5.723056030273438,
      "epoch": 0.5398967844382692,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004978047233257927,
      "loss": 5.6317,
      "mean_token_accuracy": 0.1420987643301487,
      "num_tokens": 11963853.0,
      "step": 5440
    },
    {
      "entropy": 5.667485523223877,
      "epoch": 0.5403930131004366,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977997831902591,
      "loss": 5.5927,
      "mean_token_accuracy": 0.14859448075294496,
      "num_tokens": 11975469.0,
      "step": 5445
    },
    {
      "entropy": 5.6678722381591795,
      "epoch": 0.5408892417626042,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004977948375297709,
      "loss": 5.5808,
      "mean_token_accuracy": 0.14363450706005096,
      "num_tokens": 11985227.0,
      "step": 5450
    },
    {
      "entropy": 5.57526273727417,
      "epoch": 0.5413854704247717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004977898863444509,
      "loss": 5.5922,
      "mean_token_accuracy": 0.13635839521884918,
      "num_tokens": 11995459.0,
      "step": 5455
    },
    {
      "entropy": 5.646705865859985,
      "epoch": 0.5418816990869393,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004977849296344217,
      "loss": 5.6519,
      "mean_token_accuracy": 0.1372312895953655,
      "num_tokens": 12007705.0,
      "step": 5460
    },
    {
      "entropy": 5.632894992828369,
      "epoch": 0.5423779277491068,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977799673998063,
      "loss": 5.5496,
      "mean_token_accuracy": 0.14118488058447837,
      "num_tokens": 12019729.0,
      "step": 5465
    },
    {
      "entropy": 5.699683809280396,
      "epoch": 0.5428741564112743,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004977749996407278,
      "loss": 5.6303,
      "mean_token_accuracy": 0.13460145369172097,
      "num_tokens": 12031009.0,
      "step": 5470
    },
    {
      "entropy": 5.765320444107056,
      "epoch": 0.5433703850734418,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004977700263573094,
      "loss": 5.5798,
      "mean_token_accuracy": 0.13716082349419595,
      "num_tokens": 12041396.0,
      "step": 5475
    },
    {
      "entropy": 5.692982959747314,
      "epoch": 0.5438666137356094,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004977650475496743,
      "loss": 5.7342,
      "mean_token_accuracy": 0.13190877363085746,
      "num_tokens": 12053112.0,
      "step": 5480
    },
    {
      "entropy": 5.73151273727417,
      "epoch": 0.5443628423977769,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004977600632179459,
      "loss": 5.5996,
      "mean_token_accuracy": 0.13548437431454657,
      "num_tokens": 12064823.0,
      "step": 5485
    },
    {
      "entropy": 5.780025148391724,
      "epoch": 0.5448590710599445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977550733622482,
      "loss": 5.7334,
      "mean_token_accuracy": 0.13658748045563698,
      "num_tokens": 12076656.0,
      "step": 5490
    },
    {
      "entropy": 5.697851705551147,
      "epoch": 0.5453552997221119,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004977500779827045,
      "loss": 5.5117,
      "mean_token_accuracy": 0.14188292101025582,
      "num_tokens": 12086324.0,
      "step": 5495
    },
    {
      "entropy": 5.588416719436646,
      "epoch": 0.5458515283842795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977450770794389,
      "loss": 5.5597,
      "mean_token_accuracy": 0.1385554626584053,
      "num_tokens": 12098076.0,
      "step": 5500
    },
    {
      "entropy": 5.649808597564697,
      "epoch": 0.546347757046447,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004977400706525755,
      "loss": 5.6227,
      "mean_token_accuracy": 0.1343136951327324,
      "num_tokens": 12111282.0,
      "step": 5505
    },
    {
      "entropy": 5.66321382522583,
      "epoch": 0.5468439857086145,
      "grad_norm": 1.125,
      "learning_rate": 0.0004977350587022381,
      "loss": 5.5651,
      "mean_token_accuracy": 0.14411219134926795,
      "num_tokens": 12121932.0,
      "step": 5510
    },
    {
      "entropy": 5.653674602508545,
      "epoch": 0.5473402143707821,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004977300412285513,
      "loss": 5.5735,
      "mean_token_accuracy": 0.14246958196163179,
      "num_tokens": 12133040.0,
      "step": 5515
    },
    {
      "entropy": 5.701460266113282,
      "epoch": 0.5478364430329495,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004977250182316394,
      "loss": 5.6379,
      "mean_token_accuracy": 0.13948213756084443,
      "num_tokens": 12142577.0,
      "step": 5520
    },
    {
      "entropy": 5.702146053314209,
      "epoch": 0.5483326716951171,
      "grad_norm": 1.171875,
      "learning_rate": 0.000497719989711627,
      "loss": 5.599,
      "mean_token_accuracy": 0.1384103365242481,
      "num_tokens": 12153941.0,
      "step": 5525
    },
    {
      "entropy": 5.652317714691162,
      "epoch": 0.5488289003572846,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004977149556686387,
      "loss": 5.6555,
      "mean_token_accuracy": 0.13867533653974534,
      "num_tokens": 12165347.0,
      "step": 5530
    },
    {
      "entropy": 5.608773899078369,
      "epoch": 0.5493251290194522,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004977099161027996,
      "loss": 5.5386,
      "mean_token_accuracy": 0.13975136280059813,
      "num_tokens": 12175705.0,
      "step": 5535
    },
    {
      "entropy": 5.700317764282227,
      "epoch": 0.5498213576816197,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004977048710142342,
      "loss": 5.5895,
      "mean_token_accuracy": 0.13888199031352996,
      "num_tokens": 12186781.0,
      "step": 5540
    },
    {
      "entropy": 5.747708368301391,
      "epoch": 0.5503175863437872,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004976998204030681,
      "loss": 5.6178,
      "mean_token_accuracy": 0.13247620835900306,
      "num_tokens": 12197145.0,
      "step": 5545
    },
    {
      "entropy": 5.624990844726563,
      "epoch": 0.5508138150059547,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004976947642694262,
      "loss": 5.6263,
      "mean_token_accuracy": 0.1409397877752781,
      "num_tokens": 12207637.0,
      "step": 5550
    },
    {
      "entropy": 5.633298254013061,
      "epoch": 0.5513100436681223,
      "grad_norm": 1.125,
      "learning_rate": 0.0004976897026134341,
      "loss": 5.5564,
      "mean_token_accuracy": 0.13853972479701043,
      "num_tokens": 12218156.0,
      "step": 5555
    },
    {
      "entropy": 5.7252274513244625,
      "epoch": 0.5518062723302898,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004976846354352172,
      "loss": 5.6283,
      "mean_token_accuracy": 0.13936149179935456,
      "num_tokens": 12229987.0,
      "step": 5560
    },
    {
      "entropy": 5.722540760040284,
      "epoch": 0.5523025009924574,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004976795627349012,
      "loss": 5.6081,
      "mean_token_accuracy": 0.1349262535572052,
      "num_tokens": 12240508.0,
      "step": 5565
    },
    {
      "entropy": 5.62272539138794,
      "epoch": 0.5527987296546248,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004976744845126118,
      "loss": 5.5405,
      "mean_token_accuracy": 0.1421356350183487,
      "num_tokens": 12250317.0,
      "step": 5570
    },
    {
      "entropy": 5.614935493469238,
      "epoch": 0.5532949583167924,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004976694007684751,
      "loss": 5.5891,
      "mean_token_accuracy": 0.1400021217763424,
      "num_tokens": 12262104.0,
      "step": 5575
    },
    {
      "entropy": 5.761466026306152,
      "epoch": 0.5537911869789599,
      "grad_norm": 1.25,
      "learning_rate": 0.000497664311502617,
      "loss": 5.6366,
      "mean_token_accuracy": 0.14045115411281586,
      "num_tokens": 12273259.0,
      "step": 5580
    },
    {
      "entropy": 5.7229087352752686,
      "epoch": 0.5542874156411274,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004976592167151638,
      "loss": 5.6723,
      "mean_token_accuracy": 0.1320815458893776,
      "num_tokens": 12285504.0,
      "step": 5585
    },
    {
      "entropy": 5.715719747543335,
      "epoch": 0.554783644303295,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004976541164062418,
      "loss": 5.6055,
      "mean_token_accuracy": 0.13502155393362045,
      "num_tokens": 12296512.0,
      "step": 5590
    },
    {
      "entropy": 5.545258665084839,
      "epoch": 0.5552798729654624,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004976490105759774,
      "loss": 5.4642,
      "mean_token_accuracy": 0.1421860933303833,
      "num_tokens": 12307948.0,
      "step": 5595
    },
    {
      "entropy": 5.641209363937378,
      "epoch": 0.55577610162763,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004976438992244976,
      "loss": 5.633,
      "mean_token_accuracy": 0.13229038044810296,
      "num_tokens": 12320020.0,
      "step": 5600
    },
    {
      "entropy": 5.724086093902588,
      "epoch": 0.5562723302897975,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004976387823519286,
      "loss": 5.5138,
      "mean_token_accuracy": 0.14413542300462723,
      "num_tokens": 12331091.0,
      "step": 5605
    },
    {
      "entropy": 5.681919384002685,
      "epoch": 0.5567685589519651,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004976336599583975,
      "loss": 5.6127,
      "mean_token_accuracy": 0.13598761186003686,
      "num_tokens": 12342621.0,
      "step": 5610
    },
    {
      "entropy": 5.748515462875366,
      "epoch": 0.5572647876141326,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004976285320440315,
      "loss": 5.8738,
      "mean_token_accuracy": 0.12385975420475007,
      "num_tokens": 12353330.0,
      "step": 5615
    },
    {
      "entropy": 5.765871143341064,
      "epoch": 0.5577610162763001,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004976233986089577,
      "loss": 5.6063,
      "mean_token_accuracy": 0.1373237006366253,
      "num_tokens": 12364418.0,
      "step": 5620
    },
    {
      "entropy": 5.650335502624512,
      "epoch": 0.5582572449384676,
      "grad_norm": 1.125,
      "learning_rate": 0.0004976182596533031,
      "loss": 5.5913,
      "mean_token_accuracy": 0.13827567622065545,
      "num_tokens": 12377022.0,
      "step": 5625
    },
    {
      "entropy": 5.680367279052734,
      "epoch": 0.5587534736006352,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004976131151771955,
      "loss": 5.6671,
      "mean_token_accuracy": 0.13575168177485467,
      "num_tokens": 12387799.0,
      "step": 5630
    },
    {
      "entropy": 5.630795526504516,
      "epoch": 0.5592497022628027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004976079651807623,
      "loss": 5.6079,
      "mean_token_accuracy": 0.1334557980298996,
      "num_tokens": 12397387.0,
      "step": 5635
    },
    {
      "entropy": 5.640039873123169,
      "epoch": 0.5597459309249703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976028096641313,
      "loss": 5.5243,
      "mean_token_accuracy": 0.1420644722878933,
      "num_tokens": 12407994.0,
      "step": 5640
    },
    {
      "entropy": 5.816545867919922,
      "epoch": 0.5602421595871377,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004975976486274302,
      "loss": 5.6984,
      "mean_token_accuracy": 0.13436274826526642,
      "num_tokens": 12419124.0,
      "step": 5645
    },
    {
      "entropy": 5.715925931930542,
      "epoch": 0.5607383882493053,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004975924820707872,
      "loss": 5.5958,
      "mean_token_accuracy": 0.14045625925064087,
      "num_tokens": 12428699.0,
      "step": 5650
    },
    {
      "entropy": 5.690898752212524,
      "epoch": 0.5612346169114728,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004975873099943302,
      "loss": 5.6495,
      "mean_token_accuracy": 0.13534027561545373,
      "num_tokens": 12439025.0,
      "step": 5655
    },
    {
      "entropy": 5.691838169097901,
      "epoch": 0.5617308455736403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004975821323981875,
      "loss": 5.6325,
      "mean_token_accuracy": 0.1347513109445572,
      "num_tokens": 12449439.0,
      "step": 5660
    },
    {
      "entropy": 5.722136497497559,
      "epoch": 0.5622270742358079,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004975769492824877,
      "loss": 5.5762,
      "mean_token_accuracy": 0.13460683524608613,
      "num_tokens": 12460066.0,
      "step": 5665
    },
    {
      "entropy": 5.635313987731934,
      "epoch": 0.5627233028979753,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004975717606473589,
      "loss": 5.5847,
      "mean_token_accuracy": 0.13959030508995057,
      "num_tokens": 12471093.0,
      "step": 5670
    },
    {
      "entropy": 5.682146787643433,
      "epoch": 0.5632195315601429,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004975665664929302,
      "loss": 5.5914,
      "mean_token_accuracy": 0.13997285813093185,
      "num_tokens": 12481181.0,
      "step": 5675
    },
    {
      "entropy": 5.685751628875733,
      "epoch": 0.5637157602223104,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004975613668193303,
      "loss": 5.6326,
      "mean_token_accuracy": 0.13502067476511,
      "num_tokens": 12493068.0,
      "step": 5680
    },
    {
      "entropy": 5.689381217956543,
      "epoch": 0.564211988884478,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000497556161626688,
      "loss": 5.5862,
      "mean_token_accuracy": 0.14100849404931068,
      "num_tokens": 12504025.0,
      "step": 5685
    },
    {
      "entropy": 5.658110618591309,
      "epoch": 0.5647082175466455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004975509509151325,
      "loss": 5.5948,
      "mean_token_accuracy": 0.1377375118434429,
      "num_tokens": 12514567.0,
      "step": 5690
    },
    {
      "entropy": 5.766994333267212,
      "epoch": 0.565204446208813,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004975457346847929,
      "loss": 5.6375,
      "mean_token_accuracy": 0.13393894359469413,
      "num_tokens": 12525088.0,
      "step": 5695
    },
    {
      "entropy": 5.696994304656982,
      "epoch": 0.5657006748709805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004975405129357986,
      "loss": 5.5992,
      "mean_token_accuracy": 0.13823219984769822,
      "num_tokens": 12536002.0,
      "step": 5700
    },
    {
      "entropy": 5.6891721248626705,
      "epoch": 0.5661969035331481,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004975352856682792,
      "loss": 5.6193,
      "mean_token_accuracy": 0.13830287829041482,
      "num_tokens": 12547468.0,
      "step": 5705
    },
    {
      "entropy": 5.711158084869385,
      "epoch": 0.5666931321953156,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004975300528823642,
      "loss": 5.5768,
      "mean_token_accuracy": 0.1377693384885788,
      "num_tokens": 12557670.0,
      "step": 5710
    },
    {
      "entropy": 5.638330888748169,
      "epoch": 0.5671893608574832,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004975248145781834,
      "loss": 5.627,
      "mean_token_accuracy": 0.13948642015457152,
      "num_tokens": 12568425.0,
      "step": 5715
    },
    {
      "entropy": 5.595584917068481,
      "epoch": 0.5676855895196506,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004975195707558667,
      "loss": 5.5961,
      "mean_token_accuracy": 0.13865942060947417,
      "num_tokens": 12579138.0,
      "step": 5720
    },
    {
      "entropy": 5.670900630950928,
      "epoch": 0.5681818181818182,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004975143214155443,
      "loss": 5.6068,
      "mean_token_accuracy": 0.14321467131376267,
      "num_tokens": 12589743.0,
      "step": 5725
    },
    {
      "entropy": 5.722610092163086,
      "epoch": 0.5686780468439857,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004975090665573459,
      "loss": 5.5971,
      "mean_token_accuracy": 0.14421911761164666,
      "num_tokens": 12600648.0,
      "step": 5730
    },
    {
      "entropy": 5.657762145996093,
      "epoch": 0.5691742755061532,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004975038061814023,
      "loss": 5.5964,
      "mean_token_accuracy": 0.13604630902409554,
      "num_tokens": 12611369.0,
      "step": 5735
    },
    {
      "entropy": 5.671044635772705,
      "epoch": 0.5696705041683208,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004974985402878437,
      "loss": 5.5645,
      "mean_token_accuracy": 0.13776811808347703,
      "num_tokens": 12621992.0,
      "step": 5740
    },
    {
      "entropy": 5.714522027969361,
      "epoch": 0.5701667328304882,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004974932688768007,
      "loss": 5.6411,
      "mean_token_accuracy": 0.1357453465461731,
      "num_tokens": 12632463.0,
      "step": 5745
    },
    {
      "entropy": 5.658202314376831,
      "epoch": 0.5706629614926558,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004974879919484042,
      "loss": 5.5086,
      "mean_token_accuracy": 0.13810196816921233,
      "num_tokens": 12642282.0,
      "step": 5750
    },
    {
      "entropy": 5.6354608058929445,
      "epoch": 0.5711591901548233,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004974827095027848,
      "loss": 5.5478,
      "mean_token_accuracy": 0.14347190037369728,
      "num_tokens": 12653495.0,
      "step": 5755
    },
    {
      "entropy": 5.685864782333374,
      "epoch": 0.5716554188169909,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004974774215400736,
      "loss": 5.4919,
      "mean_token_accuracy": 0.14302333667874337,
      "num_tokens": 12664322.0,
      "step": 5760
    },
    {
      "entropy": 5.578514575958252,
      "epoch": 0.5721516474791584,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004974721280604019,
      "loss": 5.5722,
      "mean_token_accuracy": 0.13698524311184884,
      "num_tokens": 12673981.0,
      "step": 5765
    },
    {
      "entropy": 5.542292690277099,
      "epoch": 0.572647876141326,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004974668290639007,
      "loss": 5.5871,
      "mean_token_accuracy": 0.14008723422884942,
      "num_tokens": 12684778.0,
      "step": 5770
    },
    {
      "entropy": 5.777960014343262,
      "epoch": 0.5731441048034934,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004974615245507014,
      "loss": 5.5635,
      "mean_token_accuracy": 0.14317553788423537,
      "num_tokens": 12695287.0,
      "step": 5775
    },
    {
      "entropy": 5.666398906707764,
      "epoch": 0.573640333465661,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004974562145209359,
      "loss": 5.6195,
      "mean_token_accuracy": 0.13330471217632295,
      "num_tokens": 12705773.0,
      "step": 5780
    },
    {
      "entropy": 5.635389184951782,
      "epoch": 0.5741365621278285,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004974508989747354,
      "loss": 5.5509,
      "mean_token_accuracy": 0.1432520866394043,
      "num_tokens": 12716783.0,
      "step": 5785
    },
    {
      "entropy": 5.659649658203125,
      "epoch": 0.5746327907899961,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004974455779122321,
      "loss": 5.6025,
      "mean_token_accuracy": 0.13743339031934737,
      "num_tokens": 12728146.0,
      "step": 5790
    },
    {
      "entropy": 5.6869384288787845,
      "epoch": 0.5751290194521635,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004974402513335577,
      "loss": 5.6155,
      "mean_token_accuracy": 0.1398487702012062,
      "num_tokens": 12739683.0,
      "step": 5795
    },
    {
      "entropy": 5.711162948608399,
      "epoch": 0.5756252481143311,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004974349192388444,
      "loss": 5.6065,
      "mean_token_accuracy": 0.13632605522871016,
      "num_tokens": 12750257.0,
      "step": 5800
    },
    {
      "entropy": 5.788808107376099,
      "epoch": 0.5761214767764986,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004974295816282245,
      "loss": 5.6892,
      "mean_token_accuracy": 0.13611532747745514,
      "num_tokens": 12761055.0,
      "step": 5805
    },
    {
      "entropy": 5.688528108596802,
      "epoch": 0.5766177054386661,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004974242385018303,
      "loss": 5.6104,
      "mean_token_accuracy": 0.14297330379486084,
      "num_tokens": 12771014.0,
      "step": 5810
    },
    {
      "entropy": 5.609748506546021,
      "epoch": 0.5771139341008337,
      "grad_norm": 1.3125,
      "learning_rate": 0.000497418889859794,
      "loss": 5.5335,
      "mean_token_accuracy": 0.14075086861848832,
      "num_tokens": 12782623.0,
      "step": 5815
    },
    {
      "entropy": 5.740034294128418,
      "epoch": 0.5776101627630011,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004974135357022488,
      "loss": 5.6505,
      "mean_token_accuracy": 0.1401318684220314,
      "num_tokens": 12793533.0,
      "step": 5820
    },
    {
      "entropy": 5.61648964881897,
      "epoch": 0.5781063914251687,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004974081760293269,
      "loss": 5.512,
      "mean_token_accuracy": 0.1384097568690777,
      "num_tokens": 12804389.0,
      "step": 5825
    },
    {
      "entropy": 5.618610429763794,
      "epoch": 0.5786026200873362,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004974028108411617,
      "loss": 5.6248,
      "mean_token_accuracy": 0.14029746502637863,
      "num_tokens": 12815672.0,
      "step": 5830
    },
    {
      "entropy": 5.741160535812378,
      "epoch": 0.5790988487495038,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973974401378859,
      "loss": 5.607,
      "mean_token_accuracy": 0.14122308045625687,
      "num_tokens": 12826113.0,
      "step": 5835
    },
    {
      "entropy": 5.631668663024902,
      "epoch": 0.5795950774116713,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004973920639196329,
      "loss": 5.5487,
      "mean_token_accuracy": 0.14111461639404296,
      "num_tokens": 12835941.0,
      "step": 5840
    },
    {
      "entropy": 5.592073154449463,
      "epoch": 0.5800913060738389,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004973866821865358,
      "loss": 5.6247,
      "mean_token_accuracy": 0.13312872648239135,
      "num_tokens": 12846186.0,
      "step": 5845
    },
    {
      "entropy": 5.74661545753479,
      "epoch": 0.5805875347360063,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004973812949387283,
      "loss": 5.6047,
      "mean_token_accuracy": 0.14106664285063744,
      "num_tokens": 12856380.0,
      "step": 5850
    },
    {
      "entropy": 5.635701704025268,
      "epoch": 0.5810837633981739,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004973759021763439,
      "loss": 5.6399,
      "mean_token_accuracy": 0.13828295320272446,
      "num_tokens": 12865999.0,
      "step": 5855
    },
    {
      "entropy": 5.644204998016358,
      "epoch": 0.5815799920603414,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004973705038995162,
      "loss": 5.517,
      "mean_token_accuracy": 0.14061327129602433,
      "num_tokens": 12876388.0,
      "step": 5860
    },
    {
      "entropy": 5.691782808303833,
      "epoch": 0.582076220722509,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004973651001083792,
      "loss": 5.5245,
      "mean_token_accuracy": 0.14006526470184327,
      "num_tokens": 12885862.0,
      "step": 5865
    },
    {
      "entropy": 5.646957874298096,
      "epoch": 0.5825724493846764,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004973596908030669,
      "loss": 5.6013,
      "mean_token_accuracy": 0.1346560686826706,
      "num_tokens": 12897489.0,
      "step": 5870
    },
    {
      "entropy": 5.69366307258606,
      "epoch": 0.583068678046844,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004973542759837135,
      "loss": 5.6249,
      "mean_token_accuracy": 0.13276098296046257,
      "num_tokens": 12907395.0,
      "step": 5875
    },
    {
      "entropy": 5.6319990158081055,
      "epoch": 0.5835649067090115,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000497348855650453,
      "loss": 5.5742,
      "mean_token_accuracy": 0.14042129814624787,
      "num_tokens": 12919439.0,
      "step": 5880
    },
    {
      "entropy": 5.719336223602295,
      "epoch": 0.584061135371179,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004973434298034201,
      "loss": 5.6786,
      "mean_token_accuracy": 0.12929542362689972,
      "num_tokens": 12931892.0,
      "step": 5885
    },
    {
      "entropy": 5.791540145874023,
      "epoch": 0.5845573640333466,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004973379984427492,
      "loss": 5.7252,
      "mean_token_accuracy": 0.13445741534233094,
      "num_tokens": 12943189.0,
      "step": 5890
    },
    {
      "entropy": 5.697743034362793,
      "epoch": 0.585053592695514,
      "grad_norm": 1.421875,
      "learning_rate": 0.000497332561568575,
      "loss": 5.6548,
      "mean_token_accuracy": 0.13218288421630858,
      "num_tokens": 12954662.0,
      "step": 5895
    },
    {
      "entropy": 5.677671623229981,
      "epoch": 0.5855498213576816,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004973271191810325,
      "loss": 5.5874,
      "mean_token_accuracy": 0.13747621178627015,
      "num_tokens": 12964924.0,
      "step": 5900
    },
    {
      "entropy": 5.711654758453369,
      "epoch": 0.5860460500198491,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004973216712802564,
      "loss": 5.667,
      "mean_token_accuracy": 0.1371985673904419,
      "num_tokens": 12976483.0,
      "step": 5905
    },
    {
      "entropy": 5.62073769569397,
      "epoch": 0.5865422786820167,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000497316217866382,
      "loss": 5.5676,
      "mean_token_accuracy": 0.14329530149698258,
      "num_tokens": 12988054.0,
      "step": 5910
    },
    {
      "entropy": 5.7230644702911375,
      "epoch": 0.5870385073441842,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004973107589395444,
      "loss": 5.6311,
      "mean_token_accuracy": 0.14080853536725044,
      "num_tokens": 13000596.0,
      "step": 5915
    },
    {
      "entropy": 5.587391328811646,
      "epoch": 0.5875347360063518,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004973052944998791,
      "loss": 5.4717,
      "mean_token_accuracy": 0.14157656133174895,
      "num_tokens": 13012360.0,
      "step": 5920
    },
    {
      "entropy": 5.644783878326416,
      "epoch": 0.5880309646685192,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004972998245475214,
      "loss": 5.6258,
      "mean_token_accuracy": 0.13704633936285973,
      "num_tokens": 13023806.0,
      "step": 5925
    },
    {
      "entropy": 5.673040199279785,
      "epoch": 0.5885271933306868,
      "grad_norm": 1.140625,
      "learning_rate": 0.000497294349082607,
      "loss": 5.5491,
      "mean_token_accuracy": 0.14386843293905258,
      "num_tokens": 13034049.0,
      "step": 5930
    },
    {
      "entropy": 5.67236328125,
      "epoch": 0.5890234219928543,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004972888681052719,
      "loss": 5.6343,
      "mean_token_accuracy": 0.13715939074754716,
      "num_tokens": 13045006.0,
      "step": 5935
    },
    {
      "entropy": 5.7274645328521725,
      "epoch": 0.5895196506550219,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004972833816156518,
      "loss": 5.5657,
      "mean_token_accuracy": 0.13989174738526344,
      "num_tokens": 13055337.0,
      "step": 5940
    },
    {
      "entropy": 5.6145857810974125,
      "epoch": 0.5900158793171894,
      "grad_norm": 1.375,
      "learning_rate": 0.0004972778896138828,
      "loss": 5.5417,
      "mean_token_accuracy": 0.14450987577438354,
      "num_tokens": 13065267.0,
      "step": 5945
    },
    {
      "entropy": 5.630779361724853,
      "epoch": 0.5905121079793569,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004972723921001011,
      "loss": 5.6165,
      "mean_token_accuracy": 0.13902801647782326,
      "num_tokens": 13075173.0,
      "step": 5950
    },
    {
      "entropy": 5.7163426876068115,
      "epoch": 0.5910083366415244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004972668890744431,
      "loss": 5.587,
      "mean_token_accuracy": 0.13847742676734925,
      "num_tokens": 13086848.0,
      "step": 5955
    },
    {
      "entropy": 5.637412977218628,
      "epoch": 0.5915045653036919,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004972613805370452,
      "loss": 5.5737,
      "mean_token_accuracy": 0.14273878633975984,
      "num_tokens": 13096865.0,
      "step": 5960
    },
    {
      "entropy": 5.6107714653015135,
      "epoch": 0.5920007939658595,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004972558664880439,
      "loss": 5.5551,
      "mean_token_accuracy": 0.13893564268946648,
      "num_tokens": 13107907.0,
      "step": 5965
    },
    {
      "entropy": 5.6265466690063475,
      "epoch": 0.592497022628027,
      "grad_norm": 1.421875,
      "learning_rate": 0.000497250346927576,
      "loss": 5.4825,
      "mean_token_accuracy": 0.14717274457216262,
      "num_tokens": 13118171.0,
      "step": 5970
    },
    {
      "entropy": 5.6349586009979244,
      "epoch": 0.5929932512901945,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004972448218557785,
      "loss": 5.4692,
      "mean_token_accuracy": 0.14425964802503585,
      "num_tokens": 13127042.0,
      "step": 5975
    },
    {
      "entropy": 5.712845277786255,
      "epoch": 0.593489479952362,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004972392912727882,
      "loss": 5.7227,
      "mean_token_accuracy": 0.14010459184646606,
      "num_tokens": 13137026.0,
      "step": 5980
    },
    {
      "entropy": 5.684154605865478,
      "epoch": 0.5939857086145296,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004972337551787425,
      "loss": 5.5856,
      "mean_token_accuracy": 0.14441057294607162,
      "num_tokens": 13146947.0,
      "step": 5985
    },
    {
      "entropy": 5.752909183502197,
      "epoch": 0.5944819372766971,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004972282135737785,
      "loss": 5.6296,
      "mean_token_accuracy": 0.1390555150806904,
      "num_tokens": 13158714.0,
      "step": 5990
    },
    {
      "entropy": 5.725433969497681,
      "epoch": 0.5949781659388647,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004972226664580337,
      "loss": 5.6347,
      "mean_token_accuracy": 0.14057021215558052,
      "num_tokens": 13170156.0,
      "step": 5995
    },
    {
      "entropy": 5.599172878265381,
      "epoch": 0.5954743946010321,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004972171138316456,
      "loss": 5.5435,
      "mean_token_accuracy": 0.14058107063174247,
      "num_tokens": 13179919.0,
      "step": 6000
    },
    {
      "epoch": 0.5954743946010321,
      "eval_entropy": 5.539261898934402,
      "eval_loss": 5.676332473754883,
      "eval_mean_token_accuracy": 0.13925191743069687,
      "eval_num_tokens": 13179919.0,
      "eval_runtime": 27.2141,
      "eval_samples_per_second": 1237.189,
      "eval_steps_per_second": 154.662,
      "step": 6000
    },
    {
      "entropy": 5.674650955200195,
      "epoch": 0.5959706232631997,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004972115556947518,
      "loss": 5.6249,
      "mean_token_accuracy": 0.13901146575808526,
      "num_tokens": 13190777.0,
      "step": 6005
    },
    {
      "entropy": 5.756080913543701,
      "epoch": 0.5964668519253672,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004972059920474904,
      "loss": 5.6127,
      "mean_token_accuracy": 0.1370018996298313,
      "num_tokens": 13202179.0,
      "step": 6010
    },
    {
      "entropy": 5.667721080780029,
      "epoch": 0.5969630805875348,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004972004228899991,
      "loss": 5.5609,
      "mean_token_accuracy": 0.14518504366278648,
      "num_tokens": 13212764.0,
      "step": 6015
    },
    {
      "entropy": 5.638906669616699,
      "epoch": 0.5974593092497023,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004971948482224162,
      "loss": 5.5963,
      "mean_token_accuracy": 0.13697412461042405,
      "num_tokens": 13223343.0,
      "step": 6020
    },
    {
      "entropy": 5.687494373321533,
      "epoch": 0.5979555379118698,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004971892680448798,
      "loss": 5.5771,
      "mean_token_accuracy": 0.14278190582990646,
      "num_tokens": 13233839.0,
      "step": 6025
    },
    {
      "entropy": 5.725995826721191,
      "epoch": 0.5984517665740373,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004971836823575283,
      "loss": 5.6752,
      "mean_token_accuracy": 0.13207956105470658,
      "num_tokens": 13243854.0,
      "step": 6030
    },
    {
      "entropy": 5.717099523544311,
      "epoch": 0.5989479952362048,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004971780911605003,
      "loss": 5.5877,
      "mean_token_accuracy": 0.14032131507992746,
      "num_tokens": 13256073.0,
      "step": 6035
    },
    {
      "entropy": 5.6954559803009035,
      "epoch": 0.5994442238983724,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004971724944539343,
      "loss": 5.6402,
      "mean_token_accuracy": 0.13555241674184798,
      "num_tokens": 13267568.0,
      "step": 6040
    },
    {
      "entropy": 5.717910718917847,
      "epoch": 0.5999404525605398,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004971668922379692,
      "loss": 5.6138,
      "mean_token_accuracy": 0.13373507261276246,
      "num_tokens": 13278829.0,
      "step": 6045
    },
    {
      "entropy": 5.715434503555298,
      "epoch": 0.6004366812227074,
      "grad_norm": 1.21875,
      "learning_rate": 0.000497161284512744,
      "loss": 5.5727,
      "mean_token_accuracy": 0.13908127769827844,
      "num_tokens": 13289156.0,
      "step": 6050
    },
    {
      "entropy": 5.565212440490723,
      "epoch": 0.6009329098848749,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004971556712783976,
      "loss": 5.437,
      "mean_token_accuracy": 0.15020594894886016,
      "num_tokens": 13299817.0,
      "step": 6055
    },
    {
      "entropy": 5.570635747909546,
      "epoch": 0.6014291385470425,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004971500525350692,
      "loss": 5.5152,
      "mean_token_accuracy": 0.1415119782090187,
      "num_tokens": 13310290.0,
      "step": 6060
    },
    {
      "entropy": 5.611606311798096,
      "epoch": 0.60192536720921,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004971444282828983,
      "loss": 5.6316,
      "mean_token_accuracy": 0.14053241312503814,
      "num_tokens": 13321747.0,
      "step": 6065
    },
    {
      "entropy": 5.770630168914795,
      "epoch": 0.6024215958713776,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004971387985220242,
      "loss": 5.5698,
      "mean_token_accuracy": 0.13983482047915458,
      "num_tokens": 13333249.0,
      "step": 6070
    },
    {
      "entropy": 5.579962587356567,
      "epoch": 0.602917824533545,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004971331632525866,
      "loss": 5.5425,
      "mean_token_accuracy": 0.14256804585456848,
      "num_tokens": 13345336.0,
      "step": 6075
    },
    {
      "entropy": 5.7523644924163815,
      "epoch": 0.6034140531957126,
      "grad_norm": 1.328125,
      "learning_rate": 0.000497127522474725,
      "loss": 5.6939,
      "mean_token_accuracy": 0.12900094762444497,
      "num_tokens": 13355797.0,
      "step": 6080
    },
    {
      "entropy": 5.762563323974609,
      "epoch": 0.6039102818578801,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004971218761885797,
      "loss": 5.6477,
      "mean_token_accuracy": 0.13897108808159828,
      "num_tokens": 13367058.0,
      "step": 6085
    },
    {
      "entropy": 5.672767210006714,
      "epoch": 0.6044065105200477,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004971162243942904,
      "loss": 5.5998,
      "mean_token_accuracy": 0.14195925369858742,
      "num_tokens": 13378579.0,
      "step": 6090
    },
    {
      "entropy": 5.662849092483521,
      "epoch": 0.6049027391822152,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004971105670919975,
      "loss": 5.5746,
      "mean_token_accuracy": 0.13392864167690277,
      "num_tokens": 13389240.0,
      "step": 6095
    },
    {
      "entropy": 5.732813835144043,
      "epoch": 0.6053989678443827,
      "grad_norm": 1.203125,
      "learning_rate": 0.000497104904281841,
      "loss": 5.7113,
      "mean_token_accuracy": 0.13209655806422232,
      "num_tokens": 13401895.0,
      "step": 6100
    },
    {
      "entropy": 5.758761978149414,
      "epoch": 0.6058951965065502,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004970992359639616,
      "loss": 5.7002,
      "mean_token_accuracy": 0.13254112601280213,
      "num_tokens": 13413334.0,
      "step": 6105
    },
    {
      "entropy": 5.666363096237182,
      "epoch": 0.6063914251687177,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004970935621384996,
      "loss": 5.6009,
      "mean_token_accuracy": 0.13594173789024352,
      "num_tokens": 13423371.0,
      "step": 6110
    },
    {
      "entropy": 5.707225322723389,
      "epoch": 0.6068876538308853,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004970878828055959,
      "loss": 5.5712,
      "mean_token_accuracy": 0.14353752508759499,
      "num_tokens": 13434782.0,
      "step": 6115
    },
    {
      "entropy": 5.667262220382691,
      "epoch": 0.6073838824930528,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004970821979653912,
      "loss": 5.5675,
      "mean_token_accuracy": 0.13443252593278884,
      "num_tokens": 13444493.0,
      "step": 6120
    },
    {
      "entropy": 5.677046823501587,
      "epoch": 0.6078801111552203,
      "grad_norm": 1.25,
      "learning_rate": 0.0004970765076180265,
      "loss": 5.5606,
      "mean_token_accuracy": 0.14075927138328553,
      "num_tokens": 13455518.0,
      "step": 6125
    },
    {
      "entropy": 5.6904298782348635,
      "epoch": 0.6083763398173878,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004970708117636431,
      "loss": 5.6519,
      "mean_token_accuracy": 0.13924447149038316,
      "num_tokens": 13467631.0,
      "step": 6130
    },
    {
      "entropy": 5.626682472229004,
      "epoch": 0.6088725684795554,
      "grad_norm": 1.3125,
      "learning_rate": 0.000497065110402382,
      "loss": 5.5835,
      "mean_token_accuracy": 0.13781617507338523,
      "num_tokens": 13478217.0,
      "step": 6135
    },
    {
      "entropy": 5.707493734359741,
      "epoch": 0.6093687971417229,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004970594035343846,
      "loss": 5.5498,
      "mean_token_accuracy": 0.13723395764827728,
      "num_tokens": 13489169.0,
      "step": 6140
    },
    {
      "entropy": 5.661879825592041,
      "epoch": 0.6098650258038905,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004970536911597926,
      "loss": 5.5769,
      "mean_token_accuracy": 0.13313405066728592,
      "num_tokens": 13499273.0,
      "step": 6145
    },
    {
      "entropy": 5.68490777015686,
      "epoch": 0.6103612544660579,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004970479732787476,
      "loss": 5.6029,
      "mean_token_accuracy": 0.13875373601913452,
      "num_tokens": 13509664.0,
      "step": 6150
    },
    {
      "entropy": 5.635036277770996,
      "epoch": 0.6108574831282255,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004970422498913911,
      "loss": 5.5095,
      "mean_token_accuracy": 0.13770228177309035,
      "num_tokens": 13520288.0,
      "step": 6155
    },
    {
      "entropy": 5.652883291244507,
      "epoch": 0.611353711790393,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004970365209978655,
      "loss": 5.5962,
      "mean_token_accuracy": 0.1385968416929245,
      "num_tokens": 13531167.0,
      "step": 6160
    },
    {
      "entropy": 5.669186305999756,
      "epoch": 0.6118499404525606,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004970307865983124,
      "loss": 5.6452,
      "mean_token_accuracy": 0.1389533519744873,
      "num_tokens": 13544109.0,
      "step": 6165
    },
    {
      "entropy": 5.670026063919067,
      "epoch": 0.6123461691147281,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004970250466928744,
      "loss": 5.5559,
      "mean_token_accuracy": 0.1415662385523319,
      "num_tokens": 13554935.0,
      "step": 6170
    },
    {
      "entropy": 5.690758848190308,
      "epoch": 0.6128423977768956,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970193012816936,
      "loss": 5.6017,
      "mean_token_accuracy": 0.13809868767857553,
      "num_tokens": 13565959.0,
      "step": 6175
    },
    {
      "entropy": 5.655741214752197,
      "epoch": 0.6133386264390631,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004970135503649125,
      "loss": 5.5941,
      "mean_token_accuracy": 0.134732249379158,
      "num_tokens": 13577883.0,
      "step": 6180
    },
    {
      "entropy": 5.661329746246338,
      "epoch": 0.6138348551012306,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004970077939426738,
      "loss": 5.5512,
      "mean_token_accuracy": 0.14186619818210602,
      "num_tokens": 13588597.0,
      "step": 6185
    },
    {
      "entropy": 5.636578178405761,
      "epoch": 0.6143310837633982,
      "grad_norm": 1.125,
      "learning_rate": 0.0004970020320151202,
      "loss": 5.4979,
      "mean_token_accuracy": 0.1422838196158409,
      "num_tokens": 13598772.0,
      "step": 6190
    },
    {
      "entropy": 5.648239517211914,
      "epoch": 0.6148273124255657,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004969962645823946,
      "loss": 5.5874,
      "mean_token_accuracy": 0.14099250435829164,
      "num_tokens": 13609696.0,
      "step": 6195
    },
    {
      "entropy": 5.65839467048645,
      "epoch": 0.6153235410877332,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00049699049164464,
      "loss": 5.5748,
      "mean_token_accuracy": 0.13995693773031234,
      "num_tokens": 13619389.0,
      "step": 6200
    },
    {
      "entropy": 5.664949941635132,
      "epoch": 0.6158197697499007,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004969847132019995,
      "loss": 5.5538,
      "mean_token_accuracy": 0.13774232640862466,
      "num_tokens": 13632133.0,
      "step": 6205
    },
    {
      "entropy": 5.664448404312134,
      "epoch": 0.6163159984120683,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004969789292546164,
      "loss": 5.4893,
      "mean_token_accuracy": 0.14898946434259414,
      "num_tokens": 13642715.0,
      "step": 6210
    },
    {
      "entropy": 5.596500825881958,
      "epoch": 0.6168122270742358,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004969731398026343,
      "loss": 5.6021,
      "mean_token_accuracy": 0.14063864797353745,
      "num_tokens": 13653223.0,
      "step": 6215
    },
    {
      "entropy": 5.610008573532104,
      "epoch": 0.6173084557364034,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004969673448461966,
      "loss": 5.5759,
      "mean_token_accuracy": 0.1375471271574497,
      "num_tokens": 13664206.0,
      "step": 6220
    },
    {
      "entropy": 5.656976890563965,
      "epoch": 0.6178046843985708,
      "grad_norm": 1.265625,
      "learning_rate": 0.000496961544385447,
      "loss": 5.5963,
      "mean_token_accuracy": 0.13967818394303322,
      "num_tokens": 13673974.0,
      "step": 6225
    },
    {
      "entropy": 5.656727123260498,
      "epoch": 0.6183009130607384,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004969557384205295,
      "loss": 5.5275,
      "mean_token_accuracy": 0.14233259111642838,
      "num_tokens": 13685189.0,
      "step": 6230
    },
    {
      "entropy": 5.7195563316345215,
      "epoch": 0.6187971417229059,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004969499269515878,
      "loss": 5.5487,
      "mean_token_accuracy": 0.14223597347736358,
      "num_tokens": 13696181.0,
      "step": 6235
    },
    {
      "entropy": 5.705056238174438,
      "epoch": 0.6192933703850735,
      "grad_norm": 1.375,
      "learning_rate": 0.0004969441099787663,
      "loss": 5.6531,
      "mean_token_accuracy": 0.13462742269039155,
      "num_tokens": 13706719.0,
      "step": 6240
    },
    {
      "entropy": 5.732485389709472,
      "epoch": 0.619789599047241,
      "grad_norm": 1.171875,
      "learning_rate": 0.000496938287502209,
      "loss": 5.6186,
      "mean_token_accuracy": 0.1415099397301674,
      "num_tokens": 13716653.0,
      "step": 6245
    },
    {
      "entropy": 5.6523768424987795,
      "epoch": 0.6202858277094085,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004969324595220605,
      "loss": 5.5196,
      "mean_token_accuracy": 0.14459524899721146,
      "num_tokens": 13728220.0,
      "step": 6250
    },
    {
      "entropy": 5.570815372467041,
      "epoch": 0.620782056371576,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004969266260384652,
      "loss": 5.4999,
      "mean_token_accuracy": 0.14312912449240683,
      "num_tokens": 13738103.0,
      "step": 6255
    },
    {
      "entropy": 5.660464859008789,
      "epoch": 0.6212782850337435,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004969207870515678,
      "loss": 5.5633,
      "mean_token_accuracy": 0.13947368934750556,
      "num_tokens": 13749845.0,
      "step": 6260
    },
    {
      "entropy": 5.710503196716308,
      "epoch": 0.6217745136959111,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000496914942561513,
      "loss": 5.5918,
      "mean_token_accuracy": 0.1427766501903534,
      "num_tokens": 13759983.0,
      "step": 6265
    },
    {
      "entropy": 5.694309282302856,
      "epoch": 0.6222707423580786,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004969090925684459,
      "loss": 5.6674,
      "mean_token_accuracy": 0.13780342638492585,
      "num_tokens": 13770045.0,
      "step": 6270
    },
    {
      "entropy": 5.577391815185547,
      "epoch": 0.6227669710202461,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004969032370725114,
      "loss": 5.548,
      "mean_token_accuracy": 0.1407869666814804,
      "num_tokens": 13779823.0,
      "step": 6275
    },
    {
      "entropy": 5.648318815231323,
      "epoch": 0.6232631996824136,
      "grad_norm": 1.125,
      "learning_rate": 0.0004968973760738548,
      "loss": 5.5478,
      "mean_token_accuracy": 0.1380795456469059,
      "num_tokens": 13791165.0,
      "step": 6280
    },
    {
      "entropy": 5.745862674713135,
      "epoch": 0.6237594283445812,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004968915095726216,
      "loss": 5.6411,
      "mean_token_accuracy": 0.13508215472102164,
      "num_tokens": 13802872.0,
      "step": 6285
    },
    {
      "entropy": 5.660447835922241,
      "epoch": 0.6242556570067487,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004968856375689568,
      "loss": 5.5716,
      "mean_token_accuracy": 0.14541017040610313,
      "num_tokens": 13813893.0,
      "step": 6290
    },
    {
      "entropy": 5.657851648330689,
      "epoch": 0.6247518856689163,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004968797600630065,
      "loss": 5.6222,
      "mean_token_accuracy": 0.14212016984820366,
      "num_tokens": 13823613.0,
      "step": 6295
    },
    {
      "entropy": 5.745026111602783,
      "epoch": 0.6252481143310837,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004968738770549162,
      "loss": 5.6324,
      "mean_token_accuracy": 0.14466911256313325,
      "num_tokens": 13835413.0,
      "step": 6300
    },
    {
      "entropy": 5.653292179107666,
      "epoch": 0.6257443429932513,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004968679885448318,
      "loss": 5.5666,
      "mean_token_accuracy": 0.1379884511232376,
      "num_tokens": 13845806.0,
      "step": 6305
    },
    {
      "entropy": 5.715186929702758,
      "epoch": 0.6262405716554188,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004968620945328995,
      "loss": 5.6673,
      "mean_token_accuracy": 0.1320894718170166,
      "num_tokens": 13856540.0,
      "step": 6310
    },
    {
      "entropy": 5.666464328765869,
      "epoch": 0.6267368003175864,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004968561950192652,
      "loss": 5.5357,
      "mean_token_accuracy": 0.1421516492962837,
      "num_tokens": 13867609.0,
      "step": 6315
    },
    {
      "entropy": 5.6684606075286865,
      "epoch": 0.6272330289797539,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004968502900040753,
      "loss": 5.5959,
      "mean_token_accuracy": 0.13551198020577432,
      "num_tokens": 13878299.0,
      "step": 6320
    },
    {
      "entropy": 5.706673192977905,
      "epoch": 0.6277292576419214,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004968443794874764,
      "loss": 5.5993,
      "mean_token_accuracy": 0.1370138332247734,
      "num_tokens": 13889625.0,
      "step": 6325
    },
    {
      "entropy": 5.623292779922485,
      "epoch": 0.6282254863040889,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004968384634696149,
      "loss": 5.5543,
      "mean_token_accuracy": 0.13414983823895454,
      "num_tokens": 13899680.0,
      "step": 6330
    },
    {
      "entropy": 5.563479089736939,
      "epoch": 0.6287217149662565,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004968325419506375,
      "loss": 5.5621,
      "mean_token_accuracy": 0.14465661495923995,
      "num_tokens": 13910883.0,
      "step": 6335
    },
    {
      "entropy": 5.654639530181885,
      "epoch": 0.629217943628424,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004968266149306909,
      "loss": 5.568,
      "mean_token_accuracy": 0.1406630113720894,
      "num_tokens": 13921442.0,
      "step": 6340
    },
    {
      "entropy": 5.665995740890503,
      "epoch": 0.6297141722905915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004968206824099224,
      "loss": 5.5819,
      "mean_token_accuracy": 0.14578820019960403,
      "num_tokens": 13930309.0,
      "step": 6345
    },
    {
      "entropy": 5.704039812088013,
      "epoch": 0.630210400952759,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004968147443884789,
      "loss": 5.5844,
      "mean_token_accuracy": 0.1445680007338524,
      "num_tokens": 13941346.0,
      "step": 6350
    },
    {
      "entropy": 5.648365449905396,
      "epoch": 0.6307066296149265,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004968088008665078,
      "loss": 5.5487,
      "mean_token_accuracy": 0.14456001222133635,
      "num_tokens": 13951933.0,
      "step": 6355
    },
    {
      "entropy": 5.611508464813232,
      "epoch": 0.6312028582770941,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004968028518441563,
      "loss": 5.6062,
      "mean_token_accuracy": 0.13924315497279166,
      "num_tokens": 13963112.0,
      "step": 6360
    },
    {
      "entropy": 5.738792657852173,
      "epoch": 0.6316990869392616,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004967968973215721,
      "loss": 5.6379,
      "mean_token_accuracy": 0.14186515137553216,
      "num_tokens": 13974218.0,
      "step": 6365
    },
    {
      "entropy": 5.7108982563018795,
      "epoch": 0.6321953156014292,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004967909372989027,
      "loss": 5.5614,
      "mean_token_accuracy": 0.1465187832713127,
      "num_tokens": 13984841.0,
      "step": 6370
    },
    {
      "entropy": 5.632450580596924,
      "epoch": 0.6326915442635966,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004967849717762959,
      "loss": 5.5929,
      "mean_token_accuracy": 0.1414507046341896,
      "num_tokens": 13997504.0,
      "step": 6375
    },
    {
      "entropy": 5.609023666381836,
      "epoch": 0.6331877729257642,
      "grad_norm": 1.125,
      "learning_rate": 0.0004967790007538999,
      "loss": 5.5203,
      "mean_token_accuracy": 0.14872804284095764,
      "num_tokens": 14009244.0,
      "step": 6380
    },
    {
      "entropy": 5.636469125747681,
      "epoch": 0.6336840015879317,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004967730242318624,
      "loss": 5.5521,
      "mean_token_accuracy": 0.1430388480424881,
      "num_tokens": 14021190.0,
      "step": 6385
    },
    {
      "entropy": 5.585435390472412,
      "epoch": 0.6341802302500993,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000496767042210332,
      "loss": 5.496,
      "mean_token_accuracy": 0.14339879602193834,
      "num_tokens": 14032191.0,
      "step": 6390
    },
    {
      "entropy": 5.5543266296386715,
      "epoch": 0.6346764589122668,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004967610546894566,
      "loss": 5.4956,
      "mean_token_accuracy": 0.14951492398977279,
      "num_tokens": 14042817.0,
      "step": 6395
    },
    {
      "entropy": 5.5756786346435545,
      "epoch": 0.6351726875744343,
      "grad_norm": 1.171875,
      "learning_rate": 0.000496755061669385,
      "loss": 5.4978,
      "mean_token_accuracy": 0.14426037520170212,
      "num_tokens": 14052725.0,
      "step": 6400
    },
    {
      "entropy": 5.647441148757935,
      "epoch": 0.6356689162366018,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004967490631502656,
      "loss": 5.5907,
      "mean_token_accuracy": 0.13421930372714996,
      "num_tokens": 14063621.0,
      "step": 6405
    },
    {
      "entropy": 5.804601240158081,
      "epoch": 0.6361651448987694,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004967430591322474,
      "loss": 5.703,
      "mean_token_accuracy": 0.13386939838528633,
      "num_tokens": 14074350.0,
      "step": 6410
    },
    {
      "entropy": 5.729103040695191,
      "epoch": 0.6366613735609369,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004967370496154791,
      "loss": 5.6279,
      "mean_token_accuracy": 0.1327170081436634,
      "num_tokens": 14084600.0,
      "step": 6415
    },
    {
      "entropy": 5.691579008102417,
      "epoch": 0.6371576022231044,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004967310346001098,
      "loss": 5.554,
      "mean_token_accuracy": 0.14277902543544768,
      "num_tokens": 14094739.0,
      "step": 6420
    },
    {
      "entropy": 5.757460069656372,
      "epoch": 0.6376538308852719,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004967250140862885,
      "loss": 5.7146,
      "mean_token_accuracy": 0.13419883623719214,
      "num_tokens": 14106020.0,
      "step": 6425
    },
    {
      "entropy": 5.801743650436402,
      "epoch": 0.6381500595474394,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004967189880741647,
      "loss": 5.6907,
      "mean_token_accuracy": 0.13547368422150613,
      "num_tokens": 14117699.0,
      "step": 6430
    },
    {
      "entropy": 5.642472600936889,
      "epoch": 0.638646288209607,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004967129565638878,
      "loss": 5.5483,
      "mean_token_accuracy": 0.1437451995909214,
      "num_tokens": 14128901.0,
      "step": 6435
    },
    {
      "entropy": 5.62691297531128,
      "epoch": 0.6391425168717745,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004967069195556073,
      "loss": 5.5958,
      "mean_token_accuracy": 0.14107475355267524,
      "num_tokens": 14140832.0,
      "step": 6440
    },
    {
      "entropy": 5.692808389663696,
      "epoch": 0.6396387455339421,
      "grad_norm": 1.375,
      "learning_rate": 0.000496700877049473,
      "loss": 5.642,
      "mean_token_accuracy": 0.136774842441082,
      "num_tokens": 14151936.0,
      "step": 6445
    },
    {
      "entropy": 5.741881799697876,
      "epoch": 0.6401349741961095,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004966948290456345,
      "loss": 5.6264,
      "mean_token_accuracy": 0.13769101351499557,
      "num_tokens": 14160945.0,
      "step": 6450
    },
    {
      "entropy": 5.630609369277954,
      "epoch": 0.6406312028582771,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004966887755442421,
      "loss": 5.5296,
      "mean_token_accuracy": 0.14113252535462378,
      "num_tokens": 14172052.0,
      "step": 6455
    },
    {
      "entropy": 5.6526648044586185,
      "epoch": 0.6411274315204446,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004966827165454456,
      "loss": 5.5777,
      "mean_token_accuracy": 0.1402004323899746,
      "num_tokens": 14181788.0,
      "step": 6460
    },
    {
      "entropy": 5.73133339881897,
      "epoch": 0.6416236601826122,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004966766520493956,
      "loss": 5.6755,
      "mean_token_accuracy": 0.13516734316945075,
      "num_tokens": 14193668.0,
      "step": 6465
    },
    {
      "entropy": 5.737149620056153,
      "epoch": 0.6421198888447797,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004966705820562422,
      "loss": 5.5519,
      "mean_token_accuracy": 0.1432420514523983,
      "num_tokens": 14203665.0,
      "step": 6470
    },
    {
      "entropy": 5.658433151245117,
      "epoch": 0.6426161175069472,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004966645065661361,
      "loss": 5.588,
      "mean_token_accuracy": 0.13887179046869277,
      "num_tokens": 14214213.0,
      "step": 6475
    },
    {
      "entropy": 5.692167949676514,
      "epoch": 0.6431123461691147,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004966584255792279,
      "loss": 5.6708,
      "mean_token_accuracy": 0.13620780408382416,
      "num_tokens": 14225754.0,
      "step": 6480
    },
    {
      "entropy": 5.7162415981292725,
      "epoch": 0.6436085748312823,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004966523390956683,
      "loss": 5.5838,
      "mean_token_accuracy": 0.13970621898770333,
      "num_tokens": 14236508.0,
      "step": 6485
    },
    {
      "entropy": 5.762499284744263,
      "epoch": 0.6441048034934498,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004966462471156083,
      "loss": 5.5935,
      "mean_token_accuracy": 0.14537082836031914,
      "num_tokens": 14247719.0,
      "step": 6490
    },
    {
      "entropy": 5.684523582458496,
      "epoch": 0.6446010321556173,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004966401496391989,
      "loss": 5.6694,
      "mean_token_accuracy": 0.13051126524806023,
      "num_tokens": 14258667.0,
      "step": 6495
    },
    {
      "entropy": 5.66899733543396,
      "epoch": 0.6450972608177848,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004966340466665915,
      "loss": 5.6793,
      "mean_token_accuracy": 0.1355142705142498,
      "num_tokens": 14269666.0,
      "step": 6500
    },
    {
      "entropy": 5.615774822235108,
      "epoch": 0.6455934894799523,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004966279381979373,
      "loss": 5.5651,
      "mean_token_accuracy": 0.1401240684092045,
      "num_tokens": 14280862.0,
      "step": 6505
    },
    {
      "entropy": 5.7421032905578615,
      "epoch": 0.6460897181421199,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004966218242333879,
      "loss": 5.7063,
      "mean_token_accuracy": 0.12791309878230095,
      "num_tokens": 14292505.0,
      "step": 6510
    },
    {
      "entropy": 5.742855167388916,
      "epoch": 0.6465859468042874,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004966157047730945,
      "loss": 5.5103,
      "mean_token_accuracy": 0.1449418306350708,
      "num_tokens": 14302547.0,
      "step": 6515
    },
    {
      "entropy": 5.614678859710693,
      "epoch": 0.647082175466455,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004966095798172095,
      "loss": 5.6538,
      "mean_token_accuracy": 0.13585017248988152,
      "num_tokens": 14313585.0,
      "step": 6520
    },
    {
      "entropy": 5.6123490810394285,
      "epoch": 0.6475784041286224,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966034493658843,
      "loss": 5.5176,
      "mean_token_accuracy": 0.13744375556707383,
      "num_tokens": 14323665.0,
      "step": 6525
    },
    {
      "entropy": 5.671282529830933,
      "epoch": 0.64807463279079,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004965973134192711,
      "loss": 5.5439,
      "mean_token_accuracy": 0.142033139616251,
      "num_tokens": 14334257.0,
      "step": 6530
    },
    {
      "entropy": 5.749910449981689,
      "epoch": 0.6485708614529575,
      "grad_norm": 1.171875,
      "learning_rate": 0.000496591171977522,
      "loss": 5.5703,
      "mean_token_accuracy": 0.14235752671957017,
      "num_tokens": 14345350.0,
      "step": 6535
    },
    {
      "entropy": 5.708539628982544,
      "epoch": 0.6490670901151251,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004965850250407893,
      "loss": 5.5925,
      "mean_token_accuracy": 0.14330435395240784,
      "num_tokens": 14355593.0,
      "step": 6540
    },
    {
      "entropy": 5.616926860809326,
      "epoch": 0.6495633187772926,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004965788726092256,
      "loss": 5.4756,
      "mean_token_accuracy": 0.14465712308883666,
      "num_tokens": 14366405.0,
      "step": 6545
    },
    {
      "entropy": 5.634476613998413,
      "epoch": 0.6500595474394602,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004965727146829832,
      "loss": 5.5546,
      "mean_token_accuracy": 0.13652571961283683,
      "num_tokens": 14377228.0,
      "step": 6550
    },
    {
      "entropy": 5.772863578796387,
      "epoch": 0.6505557761016276,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004965665512622148,
      "loss": 5.6893,
      "mean_token_accuracy": 0.13742679879069328,
      "num_tokens": 14387113.0,
      "step": 6555
    },
    {
      "entropy": 5.691931819915771,
      "epoch": 0.6510520047637952,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004965603823470735,
      "loss": 5.5737,
      "mean_token_accuracy": 0.1436733514070511,
      "num_tokens": 14397980.0,
      "step": 6560
    },
    {
      "entropy": 5.629791212081909,
      "epoch": 0.6515482334259627,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004965542079377122,
      "loss": 5.4931,
      "mean_token_accuracy": 0.14434098154306413,
      "num_tokens": 14409063.0,
      "step": 6565
    },
    {
      "entropy": 5.575025939941407,
      "epoch": 0.6520444620881302,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004965480280342839,
      "loss": 5.4802,
      "mean_token_accuracy": 0.15252825766801834,
      "num_tokens": 14421056.0,
      "step": 6570
    },
    {
      "entropy": 5.630699586868286,
      "epoch": 0.6525406907502977,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004965418426369419,
      "loss": 5.4697,
      "mean_token_accuracy": 0.15299968495965005,
      "num_tokens": 14430778.0,
      "step": 6575
    },
    {
      "entropy": 5.602815914154053,
      "epoch": 0.6530369194124652,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004965356517458395,
      "loss": 5.48,
      "mean_token_accuracy": 0.14320947229862213,
      "num_tokens": 14441544.0,
      "step": 6580
    },
    {
      "entropy": 5.649829483032226,
      "epoch": 0.6535331480746328,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004965294553611303,
      "loss": 5.6544,
      "mean_token_accuracy": 0.1364649511873722,
      "num_tokens": 14453062.0,
      "step": 6585
    },
    {
      "entropy": 5.675595712661743,
      "epoch": 0.6540293767368003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000496523253482968,
      "loss": 5.5654,
      "mean_token_accuracy": 0.1394916407763958,
      "num_tokens": 14464402.0,
      "step": 6590
    },
    {
      "entropy": 5.731576204299927,
      "epoch": 0.6545256053989679,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004965170461115065,
      "loss": 5.6484,
      "mean_token_accuracy": 0.1412472739815712,
      "num_tokens": 14476595.0,
      "step": 6595
    },
    {
      "entropy": 5.751549243927002,
      "epoch": 0.6550218340611353,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004965108332468994,
      "loss": 5.6443,
      "mean_token_accuracy": 0.13668947592377662,
      "num_tokens": 14487298.0,
      "step": 6600
    },
    {
      "entropy": 5.675912952423095,
      "epoch": 0.6555180627233029,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004965046148893011,
      "loss": 5.5498,
      "mean_token_accuracy": 0.1487780675292015,
      "num_tokens": 14499708.0,
      "step": 6605
    },
    {
      "entropy": 5.6672505855560305,
      "epoch": 0.6560142913854704,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004964983910388654,
      "loss": 5.5976,
      "mean_token_accuracy": 0.13831101059913636,
      "num_tokens": 14511773.0,
      "step": 6610
    },
    {
      "entropy": 5.619940280914307,
      "epoch": 0.656510520047638,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004964921616957471,
      "loss": 5.5022,
      "mean_token_accuracy": 0.14120615869760514,
      "num_tokens": 14523024.0,
      "step": 6615
    },
    {
      "entropy": 5.689584398269654,
      "epoch": 0.6570067487098055,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004964859268601005,
      "loss": 5.5733,
      "mean_token_accuracy": 0.14072625637054442,
      "num_tokens": 14532915.0,
      "step": 6620
    },
    {
      "entropy": 5.666746330261231,
      "epoch": 0.657502977371973,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004964796865320801,
      "loss": 5.5671,
      "mean_token_accuracy": 0.14149462580680847,
      "num_tokens": 14543627.0,
      "step": 6625
    },
    {
      "entropy": 5.629714298248291,
      "epoch": 0.6579992060341405,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004964734407118406,
      "loss": 5.5306,
      "mean_token_accuracy": 0.14535292536020278,
      "num_tokens": 14555181.0,
      "step": 6630
    },
    {
      "entropy": 5.638975334167481,
      "epoch": 0.6584954346963081,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004964671893995372,
      "loss": 5.5308,
      "mean_token_accuracy": 0.14256928563117982,
      "num_tokens": 14565542.0,
      "step": 6635
    },
    {
      "entropy": 5.703400754928589,
      "epoch": 0.6589916633584756,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004964609325953245,
      "loss": 5.5883,
      "mean_token_accuracy": 0.1413099080324173,
      "num_tokens": 14575395.0,
      "step": 6640
    },
    {
      "entropy": 5.730867099761963,
      "epoch": 0.6594878920206431,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004964546702993581,
      "loss": 5.6376,
      "mean_token_accuracy": 0.13729515075683593,
      "num_tokens": 14587206.0,
      "step": 6645
    },
    {
      "entropy": 5.698408222198486,
      "epoch": 0.6599841206828106,
      "grad_norm": 1.15625,
      "learning_rate": 0.000496448402511793,
      "loss": 5.6415,
      "mean_token_accuracy": 0.13848037272691727,
      "num_tokens": 14599048.0,
      "step": 6650
    },
    {
      "entropy": 5.748895883560181,
      "epoch": 0.6604803493449781,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004964421292327847,
      "loss": 5.5967,
      "mean_token_accuracy": 0.13868438228964805,
      "num_tokens": 14609633.0,
      "step": 6655
    },
    {
      "entropy": 5.662390947341919,
      "epoch": 0.6609765780071457,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004964358504624887,
      "loss": 5.4976,
      "mean_token_accuracy": 0.14763383120298385,
      "num_tokens": 14620373.0,
      "step": 6660
    },
    {
      "entropy": 5.575328254699707,
      "epoch": 0.6614728066693132,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004964295662010608,
      "loss": 5.4393,
      "mean_token_accuracy": 0.14503514766693115,
      "num_tokens": 14630463.0,
      "step": 6665
    },
    {
      "entropy": 5.780710697174072,
      "epoch": 0.6619690353314808,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004964232764486568,
      "loss": 5.6943,
      "mean_token_accuracy": 0.1360047772526741,
      "num_tokens": 14641523.0,
      "step": 6670
    },
    {
      "entropy": 5.65576958656311,
      "epoch": 0.6624652639936482,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004964169812054326,
      "loss": 5.5178,
      "mean_token_accuracy": 0.143309985101223,
      "num_tokens": 14651812.0,
      "step": 6675
    },
    {
      "entropy": 5.594173526763916,
      "epoch": 0.6629614926558158,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004964106804715444,
      "loss": 5.5172,
      "mean_token_accuracy": 0.14531162083148957,
      "num_tokens": 14661982.0,
      "step": 6680
    },
    {
      "entropy": 5.60432767868042,
      "epoch": 0.6634577213179833,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004964043742471486,
      "loss": 5.5335,
      "mean_token_accuracy": 0.1462668791413307,
      "num_tokens": 14671981.0,
      "step": 6685
    },
    {
      "entropy": 5.65592999458313,
      "epoch": 0.6639539499801509,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004963980625324012,
      "loss": 5.5383,
      "mean_token_accuracy": 0.13972821086645126,
      "num_tokens": 14682478.0,
      "step": 6690
    },
    {
      "entropy": 5.647305583953857,
      "epoch": 0.6644501786423184,
      "grad_norm": 1.140625,
      "learning_rate": 0.000496391745327459,
      "loss": 5.6314,
      "mean_token_accuracy": 0.13244865760207175,
      "num_tokens": 14695197.0,
      "step": 6695
    },
    {
      "entropy": 5.654128742218018,
      "epoch": 0.664946407304486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004963854226324785,
      "loss": 5.5348,
      "mean_token_accuracy": 0.13978346884250642,
      "num_tokens": 14705702.0,
      "step": 6700
    },
    {
      "entropy": 5.684941673278809,
      "epoch": 0.6654426359666534,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004963790944476167,
      "loss": 5.5372,
      "mean_token_accuracy": 0.14601366221904755,
      "num_tokens": 14715370.0,
      "step": 6705
    },
    {
      "entropy": 5.70877103805542,
      "epoch": 0.665938864628821,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004963727607730303,
      "loss": 5.5696,
      "mean_token_accuracy": 0.14533966779708862,
      "num_tokens": 14725481.0,
      "step": 6710
    },
    {
      "entropy": 5.647214078903199,
      "epoch": 0.6664350932909885,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004963664216088764,
      "loss": 5.4937,
      "mean_token_accuracy": 0.14059368669986724,
      "num_tokens": 14735250.0,
      "step": 6715
    },
    {
      "entropy": 5.623666477203369,
      "epoch": 0.666931321953156,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004963600769553124,
      "loss": 5.5031,
      "mean_token_accuracy": 0.1456894040107727,
      "num_tokens": 14744741.0,
      "step": 6720
    },
    {
      "entropy": 5.6850409507751465,
      "epoch": 0.6674275506153236,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004963537268124953,
      "loss": 5.5403,
      "mean_token_accuracy": 0.14254993498325347,
      "num_tokens": 14755865.0,
      "step": 6725
    },
    {
      "entropy": 5.562895202636719,
      "epoch": 0.667923779277491,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004963473711805829,
      "loss": 5.4957,
      "mean_token_accuracy": 0.14778019413352012,
      "num_tokens": 14766154.0,
      "step": 6730
    },
    {
      "entropy": 5.640674448013305,
      "epoch": 0.6684200079396586,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004963410100597326,
      "loss": 5.5385,
      "mean_token_accuracy": 0.13713145777583122,
      "num_tokens": 14776940.0,
      "step": 6735
    },
    {
      "entropy": 5.683514404296875,
      "epoch": 0.6689162366018261,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004963346434501021,
      "loss": 5.645,
      "mean_token_accuracy": 0.13701337948441505,
      "num_tokens": 14788367.0,
      "step": 6740
    },
    {
      "entropy": 5.664901876449585,
      "epoch": 0.6694124652639937,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004963282713518495,
      "loss": 5.5544,
      "mean_token_accuracy": 0.14326379597187042,
      "num_tokens": 14798751.0,
      "step": 6745
    },
    {
      "entropy": 5.589773416519165,
      "epoch": 0.6699086939261611,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004963218937651326,
      "loss": 5.4872,
      "mean_token_accuracy": 0.1464078515768051,
      "num_tokens": 14808748.0,
      "step": 6750
    },
    {
      "entropy": 5.691519403457642,
      "epoch": 0.6704049225883287,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004963155106901098,
      "loss": 5.6509,
      "mean_token_accuracy": 0.13549503087997436,
      "num_tokens": 14820524.0,
      "step": 6755
    },
    {
      "entropy": 5.787349843978882,
      "epoch": 0.6709011512504962,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004963091221269391,
      "loss": 5.6894,
      "mean_token_accuracy": 0.13060433566570281,
      "num_tokens": 14832485.0,
      "step": 6760
    },
    {
      "entropy": 5.663033580780029,
      "epoch": 0.6713973799126638,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000496302728075779,
      "loss": 5.4845,
      "mean_token_accuracy": 0.1408195547759533,
      "num_tokens": 14843951.0,
      "step": 6765
    },
    {
      "entropy": 5.6637598991394045,
      "epoch": 0.6718936085748313,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004962963285367882,
      "loss": 5.6021,
      "mean_token_accuracy": 0.1400584824383259,
      "num_tokens": 14855502.0,
      "step": 6770
    },
    {
      "entropy": 5.64642333984375,
      "epoch": 0.6723898372369989,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004962899235101253,
      "loss": 5.5038,
      "mean_token_accuracy": 0.14453912377357483,
      "num_tokens": 14866305.0,
      "step": 6775
    },
    {
      "entropy": 5.690909481048584,
      "epoch": 0.6728860658991663,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004962835129959491,
      "loss": 5.5257,
      "mean_token_accuracy": 0.14146189987659455,
      "num_tokens": 14877222.0,
      "step": 6780
    },
    {
      "entropy": 5.613544654846192,
      "epoch": 0.6733822945613339,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004962770969944186,
      "loss": 5.5968,
      "mean_token_accuracy": 0.14108360260725022,
      "num_tokens": 14888099.0,
      "step": 6785
    },
    {
      "entropy": 5.652590942382813,
      "epoch": 0.6738785232235014,
      "grad_norm": 1.359375,
      "learning_rate": 0.000496270675505693,
      "loss": 5.5311,
      "mean_token_accuracy": 0.14845846667885781,
      "num_tokens": 14899042.0,
      "step": 6790
    },
    {
      "entropy": 5.6786455631256105,
      "epoch": 0.6743747518856689,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004962642485299314,
      "loss": 5.593,
      "mean_token_accuracy": 0.13456349894404412,
      "num_tokens": 14908837.0,
      "step": 6795
    },
    {
      "entropy": 5.635013437271118,
      "epoch": 0.6748709805478365,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004962578160672931,
      "loss": 5.5291,
      "mean_token_accuracy": 0.14173650741577148,
      "num_tokens": 14920773.0,
      "step": 6800
    },
    {
      "entropy": 5.6272735595703125,
      "epoch": 0.6753672092100039,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004962513781179379,
      "loss": 5.5153,
      "mean_token_accuracy": 0.14413926750421524,
      "num_tokens": 14930970.0,
      "step": 6805
    },
    {
      "entropy": 5.634784173965454,
      "epoch": 0.6758634378721715,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004962449346820251,
      "loss": 5.5727,
      "mean_token_accuracy": 0.13537643030285834,
      "num_tokens": 14942378.0,
      "step": 6810
    },
    {
      "entropy": 5.667113494873047,
      "epoch": 0.676359666534339,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004962384857597148,
      "loss": 5.6142,
      "mean_token_accuracy": 0.13616188243031502,
      "num_tokens": 14953020.0,
      "step": 6815
    },
    {
      "entropy": 5.721668100357055,
      "epoch": 0.6768558951965066,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004962320313511668,
      "loss": 5.6692,
      "mean_token_accuracy": 0.13877316564321518,
      "num_tokens": 14962340.0,
      "step": 6820
    },
    {
      "entropy": 5.77312822341919,
      "epoch": 0.677352123858674,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004962255714565409,
      "loss": 5.5629,
      "mean_token_accuracy": 0.14356503784656524,
      "num_tokens": 14972792.0,
      "step": 6825
    },
    {
      "entropy": 5.712197542190552,
      "epoch": 0.6778483525208416,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004962191060759978,
      "loss": 5.566,
      "mean_token_accuracy": 0.14011918753385544,
      "num_tokens": 14983698.0,
      "step": 6830
    },
    {
      "entropy": 5.615923500061035,
      "epoch": 0.6783445811830091,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004962126352096973,
      "loss": 5.5205,
      "mean_token_accuracy": 0.14521263763308526,
      "num_tokens": 14994583.0,
      "step": 6835
    },
    {
      "entropy": 5.560826587677002,
      "epoch": 0.6788408098451767,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004962061588578004,
      "loss": 5.4634,
      "mean_token_accuracy": 0.1445203959941864,
      "num_tokens": 15005325.0,
      "step": 6840
    },
    {
      "entropy": 5.720466232299804,
      "epoch": 0.6793370385073442,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004961996770204673,
      "loss": 5.5341,
      "mean_token_accuracy": 0.14516704827547072,
      "num_tokens": 15015155.0,
      "step": 6845
    },
    {
      "entropy": 5.663384675979614,
      "epoch": 0.6798332671695118,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004961931896978588,
      "loss": 5.6054,
      "mean_token_accuracy": 0.14135198444128036,
      "num_tokens": 15026529.0,
      "step": 6850
    },
    {
      "entropy": 5.576613330841065,
      "epoch": 0.6803294958316792,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004961866968901359,
      "loss": 5.5353,
      "mean_token_accuracy": 0.14139796197414398,
      "num_tokens": 15037632.0,
      "step": 6855
    },
    {
      "entropy": 5.630209732055664,
      "epoch": 0.6808257244938468,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004961801985974595,
      "loss": 5.6066,
      "mean_token_accuracy": 0.14088872745633124,
      "num_tokens": 15050397.0,
      "step": 6860
    },
    {
      "entropy": 5.6089049816131595,
      "epoch": 0.6813219531560143,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004961736948199907,
      "loss": 5.5162,
      "mean_token_accuracy": 0.14164044559001923,
      "num_tokens": 15061505.0,
      "step": 6865
    },
    {
      "entropy": 5.743947601318359,
      "epoch": 0.6818181818181818,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004961671855578909,
      "loss": 5.7115,
      "mean_token_accuracy": 0.13932714164257048,
      "num_tokens": 15073752.0,
      "step": 6870
    },
    {
      "entropy": 5.737188816070557,
      "epoch": 0.6823144104803494,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004961606708113216,
      "loss": 5.6021,
      "mean_token_accuracy": 0.13797509521245957,
      "num_tokens": 15084730.0,
      "step": 6875
    },
    {
      "entropy": 5.672350978851318,
      "epoch": 0.6828106391425168,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004961541505804441,
      "loss": 5.4947,
      "mean_token_accuracy": 0.14610698968172073,
      "num_tokens": 15095811.0,
      "step": 6880
    },
    {
      "entropy": 5.638359546661377,
      "epoch": 0.6833068678046844,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004961476248654202,
      "loss": 5.4873,
      "mean_token_accuracy": 0.14188987910747528,
      "num_tokens": 15106102.0,
      "step": 6885
    },
    {
      "entropy": 5.659850788116455,
      "epoch": 0.6838030964668519,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004961410936664118,
      "loss": 5.5832,
      "mean_token_accuracy": 0.1400137349963188,
      "num_tokens": 15115907.0,
      "step": 6890
    },
    {
      "entropy": 5.662350511550903,
      "epoch": 0.6842993251290195,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004961345569835807,
      "loss": 5.5593,
      "mean_token_accuracy": 0.1420107327401638,
      "num_tokens": 15126788.0,
      "step": 6895
    },
    {
      "entropy": 5.668096208572388,
      "epoch": 0.684795553791187,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004961280148170892,
      "loss": 5.5177,
      "mean_token_accuracy": 0.14150577709078788,
      "num_tokens": 15137814.0,
      "step": 6900
    },
    {
      "entropy": 5.714989280700683,
      "epoch": 0.6852917824533545,
      "grad_norm": 1.625,
      "learning_rate": 0.0004961214671670994,
      "loss": 5.7345,
      "mean_token_accuracy": 0.1334407478570938,
      "num_tokens": 15148597.0,
      "step": 6905
    },
    {
      "entropy": 5.678326559066773,
      "epoch": 0.685788011115522,
      "grad_norm": 1.375,
      "learning_rate": 0.0004961149140337737,
      "loss": 5.5661,
      "mean_token_accuracy": 0.13754147216677665,
      "num_tokens": 15158985.0,
      "step": 6910
    },
    {
      "entropy": 5.67069640159607,
      "epoch": 0.6862842397776896,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004961083554172745,
      "loss": 5.5544,
      "mean_token_accuracy": 0.14061735421419144,
      "num_tokens": 15170242.0,
      "step": 6915
    },
    {
      "entropy": 5.657867908477783,
      "epoch": 0.6867804684398571,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004961017913177646,
      "loss": 5.5111,
      "mean_token_accuracy": 0.1444934755563736,
      "num_tokens": 15180353.0,
      "step": 6920
    },
    {
      "entropy": 5.649337482452393,
      "epoch": 0.6872766971020247,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004960952217354066,
      "loss": 5.4646,
      "mean_token_accuracy": 0.146437818557024,
      "num_tokens": 15191955.0,
      "step": 6925
    },
    {
      "entropy": 5.6114787578582765,
      "epoch": 0.6877729257641921,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004960886466703636,
      "loss": 5.5485,
      "mean_token_accuracy": 0.14177767410874367,
      "num_tokens": 15202088.0,
      "step": 6930
    },
    {
      "entropy": 5.524368381500244,
      "epoch": 0.6882691544263597,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004960820661227987,
      "loss": 5.4892,
      "mean_token_accuracy": 0.14549337849020957,
      "num_tokens": 15212291.0,
      "step": 6935
    },
    {
      "entropy": 5.632409954071045,
      "epoch": 0.6887653830885272,
      "grad_norm": 1.5,
      "learning_rate": 0.0004960754800928748,
      "loss": 5.5471,
      "mean_token_accuracy": 0.1425930991768837,
      "num_tokens": 15223315.0,
      "step": 6940
    },
    {
      "entropy": 5.7817751407623295,
      "epoch": 0.6892616117506947,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004960688885807553,
      "loss": 5.6371,
      "mean_token_accuracy": 0.1389908581972122,
      "num_tokens": 15233416.0,
      "step": 6945
    },
    {
      "entropy": 5.68838357925415,
      "epoch": 0.6897578404128623,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004960622915866038,
      "loss": 5.4801,
      "mean_token_accuracy": 0.14652693271636963,
      "num_tokens": 15242951.0,
      "step": 6950
    },
    {
      "entropy": 5.641556310653686,
      "epoch": 0.6902540690750297,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004960556891105839,
      "loss": 5.5553,
      "mean_token_accuracy": 0.14459084421396257,
      "num_tokens": 15253145.0,
      "step": 6955
    },
    {
      "entropy": 5.636109256744385,
      "epoch": 0.6907502977371973,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004960490811528591,
      "loss": 5.5192,
      "mean_token_accuracy": 0.14341564252972602,
      "num_tokens": 15263712.0,
      "step": 6960
    },
    {
      "entropy": 5.709726572036743,
      "epoch": 0.6912465263993648,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004960424677135935,
      "loss": 5.6146,
      "mean_token_accuracy": 0.14260344952344894,
      "num_tokens": 15273609.0,
      "step": 6965
    },
    {
      "entropy": 5.619313478469849,
      "epoch": 0.6917427550615324,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004960358487929511,
      "loss": 5.4997,
      "mean_token_accuracy": 0.1498698242008686,
      "num_tokens": 15283934.0,
      "step": 6970
    },
    {
      "entropy": 5.595314025878906,
      "epoch": 0.6922389837236999,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004960292243910958,
      "loss": 5.556,
      "mean_token_accuracy": 0.1388765126466751,
      "num_tokens": 15294394.0,
      "step": 6975
    },
    {
      "entropy": 5.633488368988037,
      "epoch": 0.6927352123858674,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000496022594508192,
      "loss": 5.5153,
      "mean_token_accuracy": 0.1490698203444481,
      "num_tokens": 15304622.0,
      "step": 6980
    },
    {
      "entropy": 5.634347820281983,
      "epoch": 0.6932314410480349,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004960159591444041,
      "loss": 5.4291,
      "mean_token_accuracy": 0.1443915270268917,
      "num_tokens": 15314094.0,
      "step": 6985
    },
    {
      "entropy": 5.657189416885376,
      "epoch": 0.6937276697102025,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004960093182998968,
      "loss": 5.5491,
      "mean_token_accuracy": 0.14283886551856995,
      "num_tokens": 15324313.0,
      "step": 6990
    },
    {
      "entropy": 5.614422035217285,
      "epoch": 0.69422389837237,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004960026719748346,
      "loss": 5.4347,
      "mean_token_accuracy": 0.16277465969324112,
      "num_tokens": 15335195.0,
      "step": 6995
    },
    {
      "entropy": 5.740299320220947,
      "epoch": 0.6947201270345376,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004959960201693822,
      "loss": 5.648,
      "mean_token_accuracy": 0.13526506498456,
      "num_tokens": 15345798.0,
      "step": 7000
    },
    {
      "entropy": 5.682719659805298,
      "epoch": 0.695216355696705,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004959893628837048,
      "loss": 5.4993,
      "mean_token_accuracy": 0.14467317312955857,
      "num_tokens": 15356340.0,
      "step": 7005
    },
    {
      "entropy": 5.684999656677246,
      "epoch": 0.6957125843588726,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004959827001179675,
      "loss": 5.5908,
      "mean_token_accuracy": 0.13729521483182908,
      "num_tokens": 15367795.0,
      "step": 7010
    },
    {
      "entropy": 5.690419864654541,
      "epoch": 0.6962088130210401,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004959760318723353,
      "loss": 5.7171,
      "mean_token_accuracy": 0.1323306903243065,
      "num_tokens": 15379214.0,
      "step": 7015
    },
    {
      "entropy": 5.64797682762146,
      "epoch": 0.6967050416832076,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004959693581469736,
      "loss": 5.5806,
      "mean_token_accuracy": 0.14593891128897668,
      "num_tokens": 15390242.0,
      "step": 7020
    },
    {
      "entropy": 5.666845226287842,
      "epoch": 0.6972012703453752,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004959626789420481,
      "loss": 5.5164,
      "mean_token_accuracy": 0.14486017376184462,
      "num_tokens": 15402600.0,
      "step": 7025
    },
    {
      "entropy": 5.642095708847046,
      "epoch": 0.6976974990075426,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004959559942577243,
      "loss": 5.5366,
      "mean_token_accuracy": 0.14455777853727342,
      "num_tokens": 15412883.0,
      "step": 7030
    },
    {
      "entropy": 5.716643142700195,
      "epoch": 0.6981937276697102,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004959493040941678,
      "loss": 5.595,
      "mean_token_accuracy": 0.13957008123397827,
      "num_tokens": 15425919.0,
      "step": 7035
    },
    {
      "entropy": 5.649199104309082,
      "epoch": 0.6986899563318777,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004959426084515448,
      "loss": 5.5128,
      "mean_token_accuracy": 0.14191488474607467,
      "num_tokens": 15437604.0,
      "step": 7040
    },
    {
      "entropy": 5.662557697296142,
      "epoch": 0.6991861849940453,
      "grad_norm": 2.4375,
      "learning_rate": 0.0004959359073300212,
      "loss": 5.5813,
      "mean_token_accuracy": 0.1390751339495182,
      "num_tokens": 15449023.0,
      "step": 7045
    },
    {
      "entropy": 5.726065063476563,
      "epoch": 0.6996824136562128,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004959292007297631,
      "loss": 5.6288,
      "mean_token_accuracy": 0.14274059757590293,
      "num_tokens": 15459567.0,
      "step": 7050
    },
    {
      "entropy": 5.717316389083862,
      "epoch": 0.7001786423183803,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000495922488650937,
      "loss": 5.5548,
      "mean_token_accuracy": 0.14280254542827606,
      "num_tokens": 15470950.0,
      "step": 7055
    },
    {
      "entropy": 5.662448120117188,
      "epoch": 0.7006748709805478,
      "grad_norm": 3.015625,
      "learning_rate": 0.0004959157710937092,
      "loss": 5.5516,
      "mean_token_accuracy": 0.13972592279314994,
      "num_tokens": 15481258.0,
      "step": 7060
    },
    {
      "entropy": 5.656245946884155,
      "epoch": 0.7011710996427154,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004959090480582462,
      "loss": 5.4839,
      "mean_token_accuracy": 0.14634299129247666,
      "num_tokens": 15491672.0,
      "step": 7065
    },
    {
      "entropy": 5.614524078369141,
      "epoch": 0.7016673283048829,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004959023195447148,
      "loss": 5.5485,
      "mean_token_accuracy": 0.1400619551539421,
      "num_tokens": 15502342.0,
      "step": 7070
    },
    {
      "entropy": 5.660572481155396,
      "epoch": 0.7021635569670505,
      "grad_norm": 1.578125,
      "learning_rate": 0.000495895585553282,
      "loss": 5.5747,
      "mean_token_accuracy": 0.1357283189892769,
      "num_tokens": 15513430.0,
      "step": 7075
    },
    {
      "entropy": 5.855673313140869,
      "epoch": 0.7026597856292179,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004958888460841147,
      "loss": 5.6992,
      "mean_token_accuracy": 0.13925954550504685,
      "num_tokens": 15524889.0,
      "step": 7080
    },
    {
      "entropy": 5.645719718933106,
      "epoch": 0.7031560142913855,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004958821011373798,
      "loss": 5.5021,
      "mean_token_accuracy": 0.14169411808252336,
      "num_tokens": 15534998.0,
      "step": 7085
    },
    {
      "entropy": 5.540553855895996,
      "epoch": 0.703652242953553,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004958753507132449,
      "loss": 5.5018,
      "mean_token_accuracy": 0.14497069120407105,
      "num_tokens": 15544752.0,
      "step": 7090
    },
    {
      "entropy": 5.630825757980347,
      "epoch": 0.7041484716157205,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004958685948118772,
      "loss": 5.5101,
      "mean_token_accuracy": 0.14215702265501023,
      "num_tokens": 15555490.0,
      "step": 7095
    },
    {
      "entropy": 5.6324597835540775,
      "epoch": 0.7046447002778881,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004958618334334443,
      "loss": 5.4528,
      "mean_token_accuracy": 0.14548865407705308,
      "num_tokens": 15565752.0,
      "step": 7100
    },
    {
      "entropy": 5.636564779281616,
      "epoch": 0.7051409289400555,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004958550665781138,
      "loss": 5.5937,
      "mean_token_accuracy": 0.13903602063655854,
      "num_tokens": 15576368.0,
      "step": 7105
    },
    {
      "entropy": 5.732804584503174,
      "epoch": 0.7056371576022231,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004958482942460536,
      "loss": 5.6002,
      "mean_token_accuracy": 0.13749876096844674,
      "num_tokens": 15587972.0,
      "step": 7110
    },
    {
      "entropy": 5.760402822494507,
      "epoch": 0.7061333862643906,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004958415164374316,
      "loss": 5.6285,
      "mean_token_accuracy": 0.1405925750732422,
      "num_tokens": 15599951.0,
      "step": 7115
    },
    {
      "entropy": 5.6129536628723145,
      "epoch": 0.7066296149265582,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004958347331524158,
      "loss": 5.6074,
      "mean_token_accuracy": 0.13861899226903915,
      "num_tokens": 15610183.0,
      "step": 7120
    },
    {
      "entropy": 5.62004919052124,
      "epoch": 0.7071258435887257,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004958279443911745,
      "loss": 5.6109,
      "mean_token_accuracy": 0.13311366289854049,
      "num_tokens": 15621074.0,
      "step": 7125
    },
    {
      "entropy": 5.735594987869263,
      "epoch": 0.7076220722508932,
      "grad_norm": 1.328125,
      "learning_rate": 0.000495821150153876,
      "loss": 5.6187,
      "mean_token_accuracy": 0.13785493820905687,
      "num_tokens": 15632752.0,
      "step": 7130
    },
    {
      "entropy": 5.735768508911133,
      "epoch": 0.7081183009130607,
      "grad_norm": 1.234375,
      "learning_rate": 0.000495814350440689,
      "loss": 5.595,
      "mean_token_accuracy": 0.13733669742941856,
      "num_tokens": 15644703.0,
      "step": 7135
    },
    {
      "entropy": 5.678989410400391,
      "epoch": 0.7086145295752283,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004958075452517816,
      "loss": 5.5451,
      "mean_token_accuracy": 0.14232822954654695,
      "num_tokens": 15655278.0,
      "step": 7140
    },
    {
      "entropy": 5.680698251724243,
      "epoch": 0.7091107582373958,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004958007345873232,
      "loss": 5.5887,
      "mean_token_accuracy": 0.142957466840744,
      "num_tokens": 15665998.0,
      "step": 7145
    },
    {
      "entropy": 5.67334885597229,
      "epoch": 0.7096069868995634,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004957939184474821,
      "loss": 5.4847,
      "mean_token_accuracy": 0.1431654192507267,
      "num_tokens": 15676343.0,
      "step": 7150
    },
    {
      "entropy": 5.628778457641602,
      "epoch": 0.7101032155617308,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004957870968324277,
      "loss": 5.5645,
      "mean_token_accuracy": 0.13879983723163605,
      "num_tokens": 15686520.0,
      "step": 7155
    },
    {
      "entropy": 5.669183254241943,
      "epoch": 0.7105994442238984,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004957802697423291,
      "loss": 5.6713,
      "mean_token_accuracy": 0.13498780652880668,
      "num_tokens": 15698302.0,
      "step": 7160
    },
    {
      "entropy": 5.685764312744141,
      "epoch": 0.7110956728860659,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004957734371773554,
      "loss": 5.433,
      "mean_token_accuracy": 0.14448211789131166,
      "num_tokens": 15708462.0,
      "step": 7165
    },
    {
      "entropy": 5.712185096740723,
      "epoch": 0.7115919015482334,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004957665991376763,
      "loss": 5.5399,
      "mean_token_accuracy": 0.1454652190208435,
      "num_tokens": 15719234.0,
      "step": 7170
    },
    {
      "entropy": 5.587440061569214,
      "epoch": 0.712088130210401,
      "grad_norm": 1.125,
      "learning_rate": 0.0004957597556234613,
      "loss": 5.5805,
      "mean_token_accuracy": 0.14065089970827102,
      "num_tokens": 15732245.0,
      "step": 7175
    },
    {
      "entropy": 5.70867280960083,
      "epoch": 0.7125843588725684,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004957529066348799,
      "loss": 5.564,
      "mean_token_accuracy": 0.14204196408390998,
      "num_tokens": 15742403.0,
      "step": 7180
    },
    {
      "entropy": 5.608488512039185,
      "epoch": 0.713080587534736,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004957460521721022,
      "loss": 5.4534,
      "mean_token_accuracy": 0.149733467400074,
      "num_tokens": 15752100.0,
      "step": 7185
    },
    {
      "entropy": 5.6119239807128904,
      "epoch": 0.7135768161969035,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000495739192235298,
      "loss": 5.5344,
      "mean_token_accuracy": 0.1443241611123085,
      "num_tokens": 15762315.0,
      "step": 7190
    },
    {
      "entropy": 5.637025356292725,
      "epoch": 0.7140730448590711,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004957323268246376,
      "loss": 5.5725,
      "mean_token_accuracy": 0.14405375793576242,
      "num_tokens": 15773841.0,
      "step": 7195
    },
    {
      "entropy": 5.645439672470093,
      "epoch": 0.7145692735212386,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004957254559402911,
      "loss": 5.5099,
      "mean_token_accuracy": 0.14171939045190812,
      "num_tokens": 15785511.0,
      "step": 7200
    },
    {
      "entropy": 5.681272315979004,
      "epoch": 0.7150655021834061,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004957185795824288,
      "loss": 5.5413,
      "mean_token_accuracy": 0.14088796973228454,
      "num_tokens": 15795394.0,
      "step": 7205
    },
    {
      "entropy": 5.608623075485229,
      "epoch": 0.7155617308455736,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004957116977512214,
      "loss": 5.5153,
      "mean_token_accuracy": 0.14078716561198235,
      "num_tokens": 15805894.0,
      "step": 7210
    },
    {
      "entropy": 5.721538782119751,
      "epoch": 0.7160579595077412,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004957048104468395,
      "loss": 5.5522,
      "mean_token_accuracy": 0.13552819862961768,
      "num_tokens": 15816595.0,
      "step": 7215
    },
    {
      "entropy": 5.69161581993103,
      "epoch": 0.7165541881699087,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004956979176694539,
      "loss": 5.5046,
      "mean_token_accuracy": 0.14846864938735962,
      "num_tokens": 15827638.0,
      "step": 7220
    },
    {
      "entropy": 5.567067670822143,
      "epoch": 0.7170504168320763,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004956910194192355,
      "loss": 5.5801,
      "mean_token_accuracy": 0.13710641860961914,
      "num_tokens": 15838437.0,
      "step": 7225
    },
    {
      "entropy": 5.706995820999145,
      "epoch": 0.7175466454942437,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004956841156963553,
      "loss": 5.6875,
      "mean_token_accuracy": 0.14104703739285468,
      "num_tokens": 15850110.0,
      "step": 7230
    },
    {
      "entropy": 5.701322555541992,
      "epoch": 0.7180428741564113,
      "grad_norm": 3.03125,
      "learning_rate": 0.0004956772065009846,
      "loss": 5.5656,
      "mean_token_accuracy": 0.13776201754808426,
      "num_tokens": 15861786.0,
      "step": 7235
    },
    {
      "entropy": 5.671795225143432,
      "epoch": 0.7185391028185788,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004956702918332948,
      "loss": 5.5349,
      "mean_token_accuracy": 0.14169813692569733,
      "num_tokens": 15871574.0,
      "step": 7240
    },
    {
      "entropy": 5.6611940383911135,
      "epoch": 0.7190353314807463,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004956633716934572,
      "loss": 5.5119,
      "mean_token_accuracy": 0.15044031888246537,
      "num_tokens": 15881944.0,
      "step": 7245
    },
    {
      "entropy": 5.618998765945435,
      "epoch": 0.7195315601429139,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004956564460816433,
      "loss": 5.5265,
      "mean_token_accuracy": 0.1403558112680912,
      "num_tokens": 15892511.0,
      "step": 7250
    },
    {
      "entropy": 5.635291957855225,
      "epoch": 0.7200277888050813,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004956495149980252,
      "loss": 5.5472,
      "mean_token_accuracy": 0.13567396774888038,
      "num_tokens": 15903387.0,
      "step": 7255
    },
    {
      "entropy": 5.697873115539551,
      "epoch": 0.7205240174672489,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004956425784427745,
      "loss": 5.5649,
      "mean_token_accuracy": 0.1381400115787983,
      "num_tokens": 15915117.0,
      "step": 7260
    },
    {
      "entropy": 5.652171039581299,
      "epoch": 0.7210202461294164,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004956356364160634,
      "loss": 5.5232,
      "mean_token_accuracy": 0.14069996327161788,
      "num_tokens": 15925163.0,
      "step": 7265
    },
    {
      "entropy": 5.534262561798096,
      "epoch": 0.721516474791584,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004956286889180639,
      "loss": 5.4321,
      "mean_token_accuracy": 0.15029175728559493,
      "num_tokens": 15935021.0,
      "step": 7270
    },
    {
      "entropy": 5.592168617248535,
      "epoch": 0.7220127034537515,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004956217359489483,
      "loss": 5.5414,
      "mean_token_accuracy": 0.13656508177518845,
      "num_tokens": 15945416.0,
      "step": 7275
    },
    {
      "entropy": 5.737779712677002,
      "epoch": 0.722508932115919,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000495614777508889,
      "loss": 5.6563,
      "mean_token_accuracy": 0.13640175238251687,
      "num_tokens": 15955388.0,
      "step": 7280
    },
    {
      "entropy": 5.7418968200683596,
      "epoch": 0.7230051607780865,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004956078135980586,
      "loss": 5.6443,
      "mean_token_accuracy": 0.13932420164346696,
      "num_tokens": 15966721.0,
      "step": 7285
    },
    {
      "entropy": 5.642996501922608,
      "epoch": 0.7235013894402541,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004956008442166299,
      "loss": 5.5803,
      "mean_token_accuracy": 0.1386307470500469,
      "num_tokens": 15976388.0,
      "step": 7290
    },
    {
      "entropy": 5.628374862670898,
      "epoch": 0.7239976181024216,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004955938693647756,
      "loss": 5.4845,
      "mean_token_accuracy": 0.14430224001407624,
      "num_tokens": 15986632.0,
      "step": 7295
    },
    {
      "entropy": 5.636146688461304,
      "epoch": 0.7244938467645892,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004955868890426686,
      "loss": 5.5292,
      "mean_token_accuracy": 0.14313902705907822,
      "num_tokens": 15996626.0,
      "step": 7300
    },
    {
      "entropy": 5.613914871215821,
      "epoch": 0.7249900754267566,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004955799032504821,
      "loss": 5.4994,
      "mean_token_accuracy": 0.14351847022771835,
      "num_tokens": 16007055.0,
      "step": 7305
    },
    {
      "entropy": 5.721293926239014,
      "epoch": 0.7254863040889242,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004955729119883893,
      "loss": 5.5683,
      "mean_token_accuracy": 0.13948507755994796,
      "num_tokens": 16019087.0,
      "step": 7310
    },
    {
      "entropy": 5.682937812805176,
      "epoch": 0.7259825327510917,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004955659152565636,
      "loss": 5.5786,
      "mean_token_accuracy": 0.13867438733577728,
      "num_tokens": 16029980.0,
      "step": 7315
    },
    {
      "entropy": 5.670715713500977,
      "epoch": 0.7264787614132592,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004955589130551784,
      "loss": 5.5493,
      "mean_token_accuracy": 0.1418918214738369,
      "num_tokens": 16040932.0,
      "step": 7320
    },
    {
      "entropy": 5.569336795806885,
      "epoch": 0.7269749900754268,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004955519053844076,
      "loss": 5.4942,
      "mean_token_accuracy": 0.14576554149389268,
      "num_tokens": 16051809.0,
      "step": 7325
    },
    {
      "entropy": 5.651855087280273,
      "epoch": 0.7274712187375942,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004955448922444246,
      "loss": 5.5587,
      "mean_token_accuracy": 0.14415163397789002,
      "num_tokens": 16062957.0,
      "step": 7330
    },
    {
      "entropy": 5.694636869430542,
      "epoch": 0.7279674473997618,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004955378736354036,
      "loss": 5.5876,
      "mean_token_accuracy": 0.14314670339226723,
      "num_tokens": 16073592.0,
      "step": 7335
    },
    {
      "entropy": 5.582796335220337,
      "epoch": 0.7284636760619293,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004955308495575185,
      "loss": 5.5451,
      "mean_token_accuracy": 0.14494318068027495,
      "num_tokens": 16085054.0,
      "step": 7340
    },
    {
      "entropy": 5.7231395721435545,
      "epoch": 0.7289599047240969,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004955238200109436,
      "loss": 5.6265,
      "mean_token_accuracy": 0.13964081034064293,
      "num_tokens": 16096936.0,
      "step": 7345
    },
    {
      "entropy": 5.730430269241333,
      "epoch": 0.7294561333862644,
      "grad_norm": 1.25,
      "learning_rate": 0.0004955167849958531,
      "loss": 5.6349,
      "mean_token_accuracy": 0.13439190089702607,
      "num_tokens": 16107673.0,
      "step": 7350
    },
    {
      "entropy": 5.7125009059906,
      "epoch": 0.729952362048432,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004955097445124215,
      "loss": 5.5013,
      "mean_token_accuracy": 0.14364912658929824,
      "num_tokens": 16117941.0,
      "step": 7355
    },
    {
      "entropy": 5.623788690567016,
      "epoch": 0.7304485907105994,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004955026985608235,
      "loss": 5.4744,
      "mean_token_accuracy": 0.14375778734683992,
      "num_tokens": 16128737.0,
      "step": 7360
    },
    {
      "entropy": 5.758817672729492,
      "epoch": 0.730944819372767,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004954956471412336,
      "loss": 5.6364,
      "mean_token_accuracy": 0.1413426525890827,
      "num_tokens": 16139599.0,
      "step": 7365
    },
    {
      "entropy": 5.565500688552857,
      "epoch": 0.7314410480349345,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004954885902538269,
      "loss": 5.4645,
      "mean_token_accuracy": 0.13935800716280938,
      "num_tokens": 16150796.0,
      "step": 7370
    },
    {
      "entropy": 5.59113245010376,
      "epoch": 0.7319372766971021,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004954815278987781,
      "loss": 5.5204,
      "mean_token_accuracy": 0.14462670534849167,
      "num_tokens": 16161236.0,
      "step": 7375
    },
    {
      "entropy": 5.714905166625977,
      "epoch": 0.7324335053592695,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004954744600762627,
      "loss": 5.6138,
      "mean_token_accuracy": 0.1440433919429779,
      "num_tokens": 16172210.0,
      "step": 7380
    },
    {
      "entropy": 5.6536970138549805,
      "epoch": 0.7329297340214371,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004954673867864555,
      "loss": 5.4905,
      "mean_token_accuracy": 0.14692775160074234,
      "num_tokens": 16182950.0,
      "step": 7385
    },
    {
      "entropy": 5.603623485565185,
      "epoch": 0.7334259626836046,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004954603080295323,
      "loss": 5.4175,
      "mean_token_accuracy": 0.1505598358809948,
      "num_tokens": 16193971.0,
      "step": 7390
    },
    {
      "entropy": 5.670416021347046,
      "epoch": 0.7339221913457721,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004954532238056686,
      "loss": 5.5412,
      "mean_token_accuracy": 0.1401501178741455,
      "num_tokens": 16205774.0,
      "step": 7395
    },
    {
      "entropy": 5.6338219165802,
      "epoch": 0.7344184200079397,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004954461341150399,
      "loss": 5.5172,
      "mean_token_accuracy": 0.1410367637872696,
      "num_tokens": 16216626.0,
      "step": 7400
    },
    {
      "entropy": 5.609722328186035,
      "epoch": 0.7349146486701071,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004954390389578222,
      "loss": 5.5946,
      "mean_token_accuracy": 0.13841652125120163,
      "num_tokens": 16227748.0,
      "step": 7405
    },
    {
      "entropy": 5.688519525527954,
      "epoch": 0.7354108773322747,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004954319383341912,
      "loss": 5.5874,
      "mean_token_accuracy": 0.14574253037571908,
      "num_tokens": 16240390.0,
      "step": 7410
    },
    {
      "entropy": 5.652543830871582,
      "epoch": 0.7359071059944422,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004954248322443232,
      "loss": 5.5223,
      "mean_token_accuracy": 0.1477010190486908,
      "num_tokens": 16250563.0,
      "step": 7415
    },
    {
      "entropy": 5.664273929595947,
      "epoch": 0.7364033346566098,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004954177206883943,
      "loss": 5.5947,
      "mean_token_accuracy": 0.1347238227725029,
      "num_tokens": 16260722.0,
      "step": 7420
    },
    {
      "entropy": 5.814026165008545,
      "epoch": 0.7368995633187773,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004954106036665809,
      "loss": 5.5938,
      "mean_token_accuracy": 0.14253473952412604,
      "num_tokens": 16270258.0,
      "step": 7425
    },
    {
      "entropy": 5.656444501876831,
      "epoch": 0.7373957919809448,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004954034811790594,
      "loss": 5.4681,
      "mean_token_accuracy": 0.14684222638607025,
      "num_tokens": 16279504.0,
      "step": 7430
    },
    {
      "entropy": 5.561936283111573,
      "epoch": 0.7378920206431123,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004953963532260067,
      "loss": 5.4833,
      "mean_token_accuracy": 0.1444912225008011,
      "num_tokens": 16291327.0,
      "step": 7435
    },
    {
      "entropy": 5.639579963684082,
      "epoch": 0.7383882493052799,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004953892198075991,
      "loss": 5.6036,
      "mean_token_accuracy": 0.1398332431912422,
      "num_tokens": 16302778.0,
      "step": 7440
    },
    {
      "entropy": 5.753041315078735,
      "epoch": 0.7388844779674474,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004953820809240138,
      "loss": 5.5611,
      "mean_token_accuracy": 0.136615801602602,
      "num_tokens": 16314373.0,
      "step": 7445
    },
    {
      "entropy": 5.669803619384766,
      "epoch": 0.739380706629615,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004953749365754279,
      "loss": 5.5311,
      "mean_token_accuracy": 0.14316530302166938,
      "num_tokens": 16324518.0,
      "step": 7450
    },
    {
      "entropy": 5.618714141845703,
      "epoch": 0.7398769352917824,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004953677867620184,
      "loss": 5.4627,
      "mean_token_accuracy": 0.14124608933925628,
      "num_tokens": 16335657.0,
      "step": 7455
    },
    {
      "entropy": 5.680794334411621,
      "epoch": 0.74037316395395,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004953606314839626,
      "loss": 5.6546,
      "mean_token_accuracy": 0.13675590977072716,
      "num_tokens": 16346497.0,
      "step": 7460
    },
    {
      "entropy": 5.577760505676269,
      "epoch": 0.7408693926161175,
      "grad_norm": 1.28125,
      "learning_rate": 0.000495353470741438,
      "loss": 5.5161,
      "mean_token_accuracy": 0.14445830434560775,
      "num_tokens": 16358175.0,
      "step": 7465
    },
    {
      "entropy": 5.673913908004761,
      "epoch": 0.741365621278285,
      "grad_norm": 1.625,
      "learning_rate": 0.000495346304534622,
      "loss": 5.6426,
      "mean_token_accuracy": 0.13683792799711228,
      "num_tokens": 16369803.0,
      "step": 7470
    },
    {
      "entropy": 5.61613655090332,
      "epoch": 0.7418618499404526,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004953391328636926,
      "loss": 5.3497,
      "mean_token_accuracy": 0.14461807459592818,
      "num_tokens": 16380535.0,
      "step": 7475
    },
    {
      "entropy": 5.614217281341553,
      "epoch": 0.74235807860262,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004953319557288275,
      "loss": 5.5205,
      "mean_token_accuracy": 0.14074980914592744,
      "num_tokens": 16390518.0,
      "step": 7480
    },
    {
      "entropy": 5.6099906921386715,
      "epoch": 0.7428543072647876,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004953247731302046,
      "loss": 5.5489,
      "mean_token_accuracy": 0.13574161306023597,
      "num_tokens": 16400627.0,
      "step": 7485
    },
    {
      "entropy": 5.64672064781189,
      "epoch": 0.7433505359269551,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004953175850680022,
      "loss": 5.6163,
      "mean_token_accuracy": 0.14487088695168496,
      "num_tokens": 16411136.0,
      "step": 7490
    },
    {
      "entropy": 5.746332693099975,
      "epoch": 0.7438467645891227,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004953103915423983,
      "loss": 5.5437,
      "mean_token_accuracy": 0.14299027472734452,
      "num_tokens": 16422830.0,
      "step": 7495
    },
    {
      "entropy": 5.6168067932128904,
      "epoch": 0.7443429932512902,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004953031925535715,
      "loss": 5.4863,
      "mean_token_accuracy": 0.147910675406456,
      "num_tokens": 16434158.0,
      "step": 7500
    },
    {
      "entropy": 5.628999137878418,
      "epoch": 0.7448392219134577,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004952959881017003,
      "loss": 5.525,
      "mean_token_accuracy": 0.14411659091711043,
      "num_tokens": 16444562.0,
      "step": 7505
    },
    {
      "entropy": 5.649853801727295,
      "epoch": 0.7453354505756252,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004952887781869631,
      "loss": 5.5619,
      "mean_token_accuracy": 0.14398169964551927,
      "num_tokens": 16456353.0,
      "step": 7510
    },
    {
      "entropy": 5.65257568359375,
      "epoch": 0.7458316792377928,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004952815628095391,
      "loss": 5.5671,
      "mean_token_accuracy": 0.1359597161412239,
      "num_tokens": 16465756.0,
      "step": 7515
    },
    {
      "entropy": 5.610208749771118,
      "epoch": 0.7463279078999603,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004952743419696069,
      "loss": 5.5214,
      "mean_token_accuracy": 0.14595141112804413,
      "num_tokens": 16477562.0,
      "step": 7520
    },
    {
      "entropy": 5.7025392055511475,
      "epoch": 0.7468241365621279,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004952671156673456,
      "loss": 5.5379,
      "mean_token_accuracy": 0.14130685329437256,
      "num_tokens": 16488134.0,
      "step": 7525
    },
    {
      "entropy": 5.690449857711792,
      "epoch": 0.7473203652242953,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004952598839029345,
      "loss": 5.534,
      "mean_token_accuracy": 0.14104952961206435,
      "num_tokens": 16499146.0,
      "step": 7530
    },
    {
      "entropy": 5.662152290344238,
      "epoch": 0.7478165938864629,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004952526466765529,
      "loss": 5.5845,
      "mean_token_accuracy": 0.14240646362304688,
      "num_tokens": 16510334.0,
      "step": 7535
    },
    {
      "entropy": 5.643217515945435,
      "epoch": 0.7483128225486304,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004952454039883804,
      "loss": 5.4581,
      "mean_token_accuracy": 0.14696277529001237,
      "num_tokens": 16520574.0,
      "step": 7540
    },
    {
      "entropy": 5.64876708984375,
      "epoch": 0.7488090512107979,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004952381558385963,
      "loss": 5.6157,
      "mean_token_accuracy": 0.13727542906999587,
      "num_tokens": 16532246.0,
      "step": 7545
    },
    {
      "entropy": 5.774498224258423,
      "epoch": 0.7493052798729655,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004952309022273805,
      "loss": 5.5924,
      "mean_token_accuracy": 0.1355849213898182,
      "num_tokens": 16541756.0,
      "step": 7550
    },
    {
      "entropy": 5.728652286529541,
      "epoch": 0.7498015085351329,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004952236431549129,
      "loss": 5.5706,
      "mean_token_accuracy": 0.13745858073234557,
      "num_tokens": 16551061.0,
      "step": 7555
    },
    {
      "entropy": 5.587491273880005,
      "epoch": 0.7502977371973005,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004952163786213735,
      "loss": 5.536,
      "mean_token_accuracy": 0.13892893865704536,
      "num_tokens": 16561155.0,
      "step": 7560
    },
    {
      "entropy": 5.599934244155884,
      "epoch": 0.750793965859468,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004952091086269425,
      "loss": 5.4823,
      "mean_token_accuracy": 0.14135693609714509,
      "num_tokens": 16571843.0,
      "step": 7565
    },
    {
      "entropy": 5.620214748382568,
      "epoch": 0.7512901945216356,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004952018331718001,
      "loss": 5.5671,
      "mean_token_accuracy": 0.14064848348498343,
      "num_tokens": 16583395.0,
      "step": 7570
    },
    {
      "entropy": 5.714365148544312,
      "epoch": 0.7517864231838031,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004951945522561268,
      "loss": 5.6238,
      "mean_token_accuracy": 0.13778568878769876,
      "num_tokens": 16594545.0,
      "step": 7575
    },
    {
      "entropy": 5.748218631744384,
      "epoch": 0.7522826518459707,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004951872658801029,
      "loss": 5.6626,
      "mean_token_accuracy": 0.13414991945028304,
      "num_tokens": 16605901.0,
      "step": 7580
    },
    {
      "entropy": 5.714016723632812,
      "epoch": 0.7527788805081381,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004951799740439093,
      "loss": 5.6248,
      "mean_token_accuracy": 0.13783007264137268,
      "num_tokens": 16616783.0,
      "step": 7585
    },
    {
      "entropy": 5.709980201721192,
      "epoch": 0.7532751091703057,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000495172676747727,
      "loss": 5.5507,
      "mean_token_accuracy": 0.14480124562978744,
      "num_tokens": 16628655.0,
      "step": 7590
    },
    {
      "entropy": 5.63602876663208,
      "epoch": 0.7537713378324732,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004951653739917366,
      "loss": 5.5555,
      "mean_token_accuracy": 0.14371190071105958,
      "num_tokens": 16639403.0,
      "step": 7595
    },
    {
      "entropy": 5.723592376708984,
      "epoch": 0.7542675664946408,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004951580657761192,
      "loss": 5.5653,
      "mean_token_accuracy": 0.14568303823471068,
      "num_tokens": 16649937.0,
      "step": 7600
    },
    {
      "entropy": 5.57650785446167,
      "epoch": 0.7547637951568082,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004951507521010563,
      "loss": 5.4744,
      "mean_token_accuracy": 0.14698772504925728,
      "num_tokens": 16661466.0,
      "step": 7605
    },
    {
      "entropy": 5.60304217338562,
      "epoch": 0.7552600238189758,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004951434329667293,
      "loss": 5.5026,
      "mean_token_accuracy": 0.1420058488845825,
      "num_tokens": 16673362.0,
      "step": 7610
    },
    {
      "entropy": 5.621581125259399,
      "epoch": 0.7557562524811433,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004951361083733192,
      "loss": 5.5513,
      "mean_token_accuracy": 0.14048047438263894,
      "num_tokens": 16685555.0,
      "step": 7615
    },
    {
      "entropy": 5.680461406707764,
      "epoch": 0.7562524811433108,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004951287783210082,
      "loss": 5.5161,
      "mean_token_accuracy": 0.13719147592782974,
      "num_tokens": 16696006.0,
      "step": 7620
    },
    {
      "entropy": 5.664773893356323,
      "epoch": 0.7567487098054784,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004951214428099778,
      "loss": 5.5321,
      "mean_token_accuracy": 0.1445251762866974,
      "num_tokens": 16706784.0,
      "step": 7625
    },
    {
      "entropy": 5.5900732517242435,
      "epoch": 0.7572449384676458,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004951141018404099,
      "loss": 5.492,
      "mean_token_accuracy": 0.142139433324337,
      "num_tokens": 16718492.0,
      "step": 7630
    },
    {
      "entropy": 5.6291234493255615,
      "epoch": 0.7577411671298134,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004951067554124865,
      "loss": 5.5611,
      "mean_token_accuracy": 0.13935871571302413,
      "num_tokens": 16729711.0,
      "step": 7635
    },
    {
      "entropy": 5.674110984802246,
      "epoch": 0.7582373957919809,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00049509940352639,
      "loss": 5.5401,
      "mean_token_accuracy": 0.14078266322612762,
      "num_tokens": 16739913.0,
      "step": 7640
    },
    {
      "entropy": 5.7222511768341064,
      "epoch": 0.7587336244541485,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004950920461823024,
      "loss": 5.5741,
      "mean_token_accuracy": 0.13388756290078163,
      "num_tokens": 16752273.0,
      "step": 7645
    },
    {
      "entropy": 5.629803419113159,
      "epoch": 0.759229853116316,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004950846833804064,
      "loss": 5.5227,
      "mean_token_accuracy": 0.1445976808667183,
      "num_tokens": 16762050.0,
      "step": 7650
    },
    {
      "entropy": 5.603659439086914,
      "epoch": 0.7597260817784836,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004950773151208845,
      "loss": 5.5835,
      "mean_token_accuracy": 0.14112742990255356,
      "num_tokens": 16772376.0,
      "step": 7655
    },
    {
      "entropy": 5.637712192535401,
      "epoch": 0.760222310440651,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004950699414039194,
      "loss": 5.5362,
      "mean_token_accuracy": 0.14236859381198883,
      "num_tokens": 16783307.0,
      "step": 7660
    },
    {
      "entropy": 5.644722890853882,
      "epoch": 0.7607185391028186,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000495062562229694,
      "loss": 5.4185,
      "mean_token_accuracy": 0.15151610597968102,
      "num_tokens": 16793582.0,
      "step": 7665
    },
    {
      "entropy": 5.582019853591919,
      "epoch": 0.7612147677649861,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004950551775983912,
      "loss": 5.5423,
      "mean_token_accuracy": 0.14074981063604355,
      "num_tokens": 16805160.0,
      "step": 7670
    },
    {
      "entropy": 5.618256998062134,
      "epoch": 0.7617109964271537,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004950477875101942,
      "loss": 5.5875,
      "mean_token_accuracy": 0.13851356729865075,
      "num_tokens": 16816108.0,
      "step": 7675
    },
    {
      "entropy": 5.612975358963013,
      "epoch": 0.7622072250893211,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004950403919652863,
      "loss": 5.4737,
      "mean_token_accuracy": 0.14750833958387374,
      "num_tokens": 16828126.0,
      "step": 7680
    },
    {
      "entropy": 5.662538337707519,
      "epoch": 0.7627034537514887,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004950329909638507,
      "loss": 5.5507,
      "mean_token_accuracy": 0.13998630419373512,
      "num_tokens": 16838186.0,
      "step": 7685
    },
    {
      "entropy": 5.617666625976563,
      "epoch": 0.7631996824136562,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000495025584506071,
      "loss": 5.495,
      "mean_token_accuracy": 0.14366085231304168,
      "num_tokens": 16848588.0,
      "step": 7690
    },
    {
      "entropy": 5.64614634513855,
      "epoch": 0.7636959110758237,
      "grad_norm": 1.125,
      "learning_rate": 0.0004950181725921311,
      "loss": 5.6076,
      "mean_token_accuracy": 0.1438237413764,
      "num_tokens": 16860269.0,
      "step": 7695
    },
    {
      "entropy": 5.605140256881714,
      "epoch": 0.7641921397379913,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004950107552222145,
      "loss": 5.6063,
      "mean_token_accuracy": 0.14137522503733635,
      "num_tokens": 16872221.0,
      "step": 7700
    },
    {
      "entropy": 5.62051739692688,
      "epoch": 0.7646883684001587,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004950033323965052,
      "loss": 5.4903,
      "mean_token_accuracy": 0.13801852017641067,
      "num_tokens": 16882588.0,
      "step": 7705
    },
    {
      "entropy": 5.720653057098389,
      "epoch": 0.7651845970623263,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004949959041151875,
      "loss": 5.5421,
      "mean_token_accuracy": 0.13992298170924186,
      "num_tokens": 16894278.0,
      "step": 7710
    },
    {
      "entropy": 5.709570741653442,
      "epoch": 0.7656808257244938,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004949884703784453,
      "loss": 5.5144,
      "mean_token_accuracy": 0.13773196712136268,
      "num_tokens": 16905448.0,
      "step": 7715
    },
    {
      "entropy": 5.614786815643311,
      "epoch": 0.7661770543866614,
      "grad_norm": 1.1875,
      "learning_rate": 0.000494981031186463,
      "loss": 5.5471,
      "mean_token_accuracy": 0.1405929893255234,
      "num_tokens": 16914960.0,
      "step": 7720
    },
    {
      "entropy": 5.621784210205078,
      "epoch": 0.7666732830488289,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004949735865394253,
      "loss": 5.6016,
      "mean_token_accuracy": 0.1381474293768406,
      "num_tokens": 16926839.0,
      "step": 7725
    },
    {
      "entropy": 5.658161544799805,
      "epoch": 0.7671695117109965,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004949661364375167,
      "loss": 5.483,
      "mean_token_accuracy": 0.15003880709409714,
      "num_tokens": 16938553.0,
      "step": 7730
    },
    {
      "entropy": 5.702442026138305,
      "epoch": 0.7676657403731639,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004949586808809217,
      "loss": 5.5523,
      "mean_token_accuracy": 0.14368932992219924,
      "num_tokens": 16947987.0,
      "step": 7735
    },
    {
      "entropy": 5.563274669647217,
      "epoch": 0.7681619690353315,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004949512198698256,
      "loss": 5.5136,
      "mean_token_accuracy": 0.1394871860742569,
      "num_tokens": 16959359.0,
      "step": 7740
    },
    {
      "entropy": 5.619741582870484,
      "epoch": 0.768658197697499,
      "grad_norm": 1.34375,
      "learning_rate": 0.000494943753404413,
      "loss": 5.5563,
      "mean_token_accuracy": 0.14314054921269417,
      "num_tokens": 16970731.0,
      "step": 7745
    },
    {
      "entropy": 5.622791481018067,
      "epoch": 0.7691544263596666,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004949362814848693,
      "loss": 5.4888,
      "mean_token_accuracy": 0.14346031695604325,
      "num_tokens": 16981868.0,
      "step": 7750
    },
    {
      "entropy": 5.689385843276978,
      "epoch": 0.769650655021834,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004949288041113799,
      "loss": 5.538,
      "mean_token_accuracy": 0.13591035604476928,
      "num_tokens": 16994639.0,
      "step": 7755
    },
    {
      "entropy": 5.628283739089966,
      "epoch": 0.7701468836840016,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004949213212841299,
      "loss": 5.5455,
      "mean_token_accuracy": 0.13931762129068376,
      "num_tokens": 17005067.0,
      "step": 7760
    },
    {
      "entropy": 5.602070760726929,
      "epoch": 0.7706431123461691,
      "grad_norm": 1.484375,
      "learning_rate": 0.000494913833003305,
      "loss": 5.4596,
      "mean_token_accuracy": 0.14669453501701354,
      "num_tokens": 17015102.0,
      "step": 7765
    },
    {
      "entropy": 5.641578006744385,
      "epoch": 0.7711393410083366,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000494906339269091,
      "loss": 5.5006,
      "mean_token_accuracy": 0.1437997043132782,
      "num_tokens": 17026290.0,
      "step": 7770
    },
    {
      "entropy": 5.628538608551025,
      "epoch": 0.7716355696705042,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004948988400816736,
      "loss": 5.4479,
      "mean_token_accuracy": 0.14013357385993003,
      "num_tokens": 17035445.0,
      "step": 7775
    },
    {
      "entropy": 5.663299036026001,
      "epoch": 0.7721317983326716,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004948913354412388,
      "loss": 5.5597,
      "mean_token_accuracy": 0.14692412912845612,
      "num_tokens": 17046515.0,
      "step": 7780
    },
    {
      "entropy": 5.665937852859497,
      "epoch": 0.7726280269948392,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004948838253479727,
      "loss": 5.5341,
      "mean_token_accuracy": 0.14354851469397545,
      "num_tokens": 17056988.0,
      "step": 7785
    },
    {
      "entropy": 5.645240259170532,
      "epoch": 0.7731242556570067,
      "grad_norm": 1.125,
      "learning_rate": 0.0004948763098020616,
      "loss": 5.5975,
      "mean_token_accuracy": 0.13405250981450081,
      "num_tokens": 17068304.0,
      "step": 7790
    },
    {
      "entropy": 5.615477275848389,
      "epoch": 0.7736204843191743,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004948687888036918,
      "loss": 5.5072,
      "mean_token_accuracy": 0.14080338925123215,
      "num_tokens": 17078405.0,
      "step": 7795
    },
    {
      "entropy": 5.621860647201538,
      "epoch": 0.7741167129813418,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004948612623530497,
      "loss": 5.4073,
      "mean_token_accuracy": 0.14988256618380547,
      "num_tokens": 17087839.0,
      "step": 7800
    },
    {
      "entropy": 5.525861644744873,
      "epoch": 0.7746129416435094,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004948537304503221,
      "loss": 5.4947,
      "mean_token_accuracy": 0.14568108916282654,
      "num_tokens": 17099386.0,
      "step": 7805
    },
    {
      "entropy": 5.681252431869507,
      "epoch": 0.7751091703056768,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004948461930956959,
      "loss": 5.5646,
      "mean_token_accuracy": 0.1383578047156334,
      "num_tokens": 17111754.0,
      "step": 7810
    },
    {
      "entropy": 5.650686931610108,
      "epoch": 0.7756053989678444,
      "grad_norm": 1.25,
      "learning_rate": 0.0004948386502893577,
      "loss": 5.448,
      "mean_token_accuracy": 0.14562888592481613,
      "num_tokens": 17122882.0,
      "step": 7815
    },
    {
      "entropy": 5.586495399475098,
      "epoch": 0.7761016276300119,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004948311020314946,
      "loss": 5.5715,
      "mean_token_accuracy": 0.14041213020682336,
      "num_tokens": 17133986.0,
      "step": 7820
    },
    {
      "entropy": 5.62388653755188,
      "epoch": 0.7765978562921795,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004948235483222941,
      "loss": 5.5453,
      "mean_token_accuracy": 0.14663455709815026,
      "num_tokens": 17144784.0,
      "step": 7825
    },
    {
      "entropy": 5.652572917938232,
      "epoch": 0.777094084954347,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004948159891619431,
      "loss": 5.6287,
      "mean_token_accuracy": 0.13586134910583497,
      "num_tokens": 17156328.0,
      "step": 7830
    },
    {
      "entropy": 5.710602951049805,
      "epoch": 0.7775903136165145,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004948084245506293,
      "loss": 5.4642,
      "mean_token_accuracy": 0.14133559837937354,
      "num_tokens": 17168062.0,
      "step": 7835
    },
    {
      "entropy": 5.6492588996887205,
      "epoch": 0.778086542278682,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00049480085448854,
      "loss": 5.4894,
      "mean_token_accuracy": 0.14454836696386336,
      "num_tokens": 17178872.0,
      "step": 7840
    },
    {
      "entropy": 5.6525005340576175,
      "epoch": 0.7785827709408495,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004947932789758635,
      "loss": 5.5435,
      "mean_token_accuracy": 0.14006778448820115,
      "num_tokens": 17189445.0,
      "step": 7845
    },
    {
      "entropy": 5.57915768623352,
      "epoch": 0.7790789996030171,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000494785698012787,
      "loss": 5.4919,
      "mean_token_accuracy": 0.14418973475694657,
      "num_tokens": 17199454.0,
      "step": 7850
    },
    {
      "entropy": 5.599864912033081,
      "epoch": 0.7795752282651845,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004947781115994989,
      "loss": 5.549,
      "mean_token_accuracy": 0.13822765350341798,
      "num_tokens": 17210729.0,
      "step": 7855
    },
    {
      "entropy": 5.659051084518433,
      "epoch": 0.7800714569273521,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004947705197361872,
      "loss": 5.6149,
      "mean_token_accuracy": 0.13760534673929214,
      "num_tokens": 17222961.0,
      "step": 7860
    },
    {
      "entropy": 5.673983192443847,
      "epoch": 0.7805676855895196,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004947629224230401,
      "loss": 5.4925,
      "mean_token_accuracy": 0.14899449422955513,
      "num_tokens": 17234026.0,
      "step": 7865
    },
    {
      "entropy": 5.710447072982788,
      "epoch": 0.7810639142516872,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004947553196602461,
      "loss": 5.6225,
      "mean_token_accuracy": 0.14361389130353927,
      "num_tokens": 17244643.0,
      "step": 7870
    },
    {
      "entropy": 5.69153995513916,
      "epoch": 0.7815601429138547,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004947477114479939,
      "loss": 5.5433,
      "mean_token_accuracy": 0.14631589204072953,
      "num_tokens": 17255382.0,
      "step": 7875
    },
    {
      "entropy": 5.606116437911988,
      "epoch": 0.7820563715760223,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004947400977864718,
      "loss": 5.5204,
      "mean_token_accuracy": 0.1373639538884163,
      "num_tokens": 17265860.0,
      "step": 7880
    },
    {
      "entropy": 5.664776563644409,
      "epoch": 0.7825526002381897,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004947324786758688,
      "loss": 5.5999,
      "mean_token_accuracy": 0.13771443516016008,
      "num_tokens": 17277570.0,
      "step": 7885
    },
    {
      "entropy": 5.760136890411377,
      "epoch": 0.7830488289003573,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004947248541163738,
      "loss": 5.6384,
      "mean_token_accuracy": 0.13873277902603148,
      "num_tokens": 17288684.0,
      "step": 7890
    },
    {
      "entropy": 5.668456363677978,
      "epoch": 0.7835450575625248,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004947172241081759,
      "loss": 5.4993,
      "mean_token_accuracy": 0.14644264727830886,
      "num_tokens": 17299541.0,
      "step": 7895
    },
    {
      "entropy": 5.6273026943206785,
      "epoch": 0.7840412862246924,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004947095886514643,
      "loss": 5.4709,
      "mean_token_accuracy": 0.14965724796056748,
      "num_tokens": 17310243.0,
      "step": 7900
    },
    {
      "entropy": 5.63713493347168,
      "epoch": 0.7845375148868599,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004947019477464283,
      "loss": 5.57,
      "mean_token_accuracy": 0.14205240532755853,
      "num_tokens": 17321167.0,
      "step": 7905
    },
    {
      "entropy": 5.534025382995606,
      "epoch": 0.7850337435490274,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004946943013932574,
      "loss": 5.5262,
      "mean_token_accuracy": 0.14619433730840684,
      "num_tokens": 17331076.0,
      "step": 7910
    },
    {
      "entropy": 5.675403499603272,
      "epoch": 0.7855299722111949,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004946866495921413,
      "loss": 5.5196,
      "mean_token_accuracy": 0.1414807878434658,
      "num_tokens": 17341337.0,
      "step": 7915
    },
    {
      "entropy": 5.684132242202759,
      "epoch": 0.7860262008733624,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004946789923432695,
      "loss": 5.5462,
      "mean_token_accuracy": 0.14194025695323945,
      "num_tokens": 17353055.0,
      "step": 7920
    },
    {
      "entropy": 5.542935991287232,
      "epoch": 0.78652242953553,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004946713296468322,
      "loss": 5.4456,
      "mean_token_accuracy": 0.14502155482769014,
      "num_tokens": 17364442.0,
      "step": 7925
    },
    {
      "entropy": 5.645188665390014,
      "epoch": 0.7870186581976975,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004946636615030192,
      "loss": 5.5508,
      "mean_token_accuracy": 0.14122460335493087,
      "num_tokens": 17376237.0,
      "step": 7930
    },
    {
      "entropy": 5.659143877029419,
      "epoch": 0.787514886859865,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004946559879120208,
      "loss": 5.4778,
      "mean_token_accuracy": 0.14349617809057236,
      "num_tokens": 17386188.0,
      "step": 7935
    },
    {
      "entropy": 5.561950588226319,
      "epoch": 0.7880111155220325,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004946483088740273,
      "loss": 5.497,
      "mean_token_accuracy": 0.14329716339707374,
      "num_tokens": 17397058.0,
      "step": 7940
    },
    {
      "entropy": 5.699773120880127,
      "epoch": 0.7885073441842001,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004946406243892289,
      "loss": 5.5419,
      "mean_token_accuracy": 0.14394629895687103,
      "num_tokens": 17408191.0,
      "step": 7945
    },
    {
      "entropy": 5.678687524795532,
      "epoch": 0.7890035728463676,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004946329344578163,
      "loss": 5.5065,
      "mean_token_accuracy": 0.14297162592411042,
      "num_tokens": 17418472.0,
      "step": 7950
    },
    {
      "entropy": 5.611789798736572,
      "epoch": 0.7894998015085352,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004946252390799803,
      "loss": 5.4377,
      "mean_token_accuracy": 0.14308183789253234,
      "num_tokens": 17429556.0,
      "step": 7955
    },
    {
      "entropy": 5.6682045459747314,
      "epoch": 0.7899960301707026,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004946175382559116,
      "loss": 5.6096,
      "mean_token_accuracy": 0.13845320641994477,
      "num_tokens": 17440950.0,
      "step": 7960
    },
    {
      "entropy": 5.638448476791382,
      "epoch": 0.7904922588328702,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000494609831985801,
      "loss": 5.4876,
      "mean_token_accuracy": 0.1446587011218071,
      "num_tokens": 17453724.0,
      "step": 7965
    },
    {
      "entropy": 5.629575109481811,
      "epoch": 0.7909884874950377,
      "grad_norm": 1.3125,
      "learning_rate": 0.00049460212026984,
      "loss": 5.5263,
      "mean_token_accuracy": 0.144678346067667,
      "num_tokens": 17464623.0,
      "step": 7970
    },
    {
      "entropy": 5.572264909744263,
      "epoch": 0.7914847161572053,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004945944031082195,
      "loss": 5.5249,
      "mean_token_accuracy": 0.1431472770869732,
      "num_tokens": 17474919.0,
      "step": 7975
    },
    {
      "entropy": 5.683834886550903,
      "epoch": 0.7919809448193728,
      "grad_norm": 2.375,
      "learning_rate": 0.000494586680501131,
      "loss": 5.5026,
      "mean_token_accuracy": 0.14230695813894273,
      "num_tokens": 17484747.0,
      "step": 7980
    },
    {
      "entropy": 5.63348388671875,
      "epoch": 0.7924771734815403,
      "grad_norm": 1.234375,
      "learning_rate": 0.000494578952448766,
      "loss": 5.4484,
      "mean_token_accuracy": 0.14592572152614594,
      "num_tokens": 17496299.0,
      "step": 7985
    },
    {
      "entropy": 5.702383899688721,
      "epoch": 0.7929734021437078,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004945712189513161,
      "loss": 5.6066,
      "mean_token_accuracy": 0.1367204412817955,
      "num_tokens": 17507448.0,
      "step": 7990
    },
    {
      "entropy": 5.682683324813842,
      "epoch": 0.7934696308058753,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004945634800089733,
      "loss": 5.5848,
      "mean_token_accuracy": 0.1361731007695198,
      "num_tokens": 17520112.0,
      "step": 7995
    },
    {
      "entropy": 5.699263191223144,
      "epoch": 0.7939658594680429,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004945557356219291,
      "loss": 5.568,
      "mean_token_accuracy": 0.1369085967540741,
      "num_tokens": 17531368.0,
      "step": 8000
    },
    {
      "entropy": 5.583948802947998,
      "epoch": 0.7944620881302104,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004945479857903758,
      "loss": 5.5302,
      "mean_token_accuracy": 0.14669519737362863,
      "num_tokens": 17542082.0,
      "step": 8005
    },
    {
      "entropy": 5.668810558319092,
      "epoch": 0.7949583167923779,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004945402305145056,
      "loss": 5.551,
      "mean_token_accuracy": 0.13798968866467476,
      "num_tokens": 17555518.0,
      "step": 8010
    },
    {
      "entropy": 5.678214120864868,
      "epoch": 0.7954545454545454,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004945324697945107,
      "loss": 5.5246,
      "mean_token_accuracy": 0.1425155833363533,
      "num_tokens": 17566110.0,
      "step": 8015
    },
    {
      "entropy": 5.534947109222412,
      "epoch": 0.795950774116713,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004945247036305836,
      "loss": 5.4374,
      "mean_token_accuracy": 0.1458745241165161,
      "num_tokens": 17575942.0,
      "step": 8020
    },
    {
      "entropy": 5.6033344745635985,
      "epoch": 0.7964470027788805,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004945169320229169,
      "loss": 5.5202,
      "mean_token_accuracy": 0.14511556103825568,
      "num_tokens": 17586001.0,
      "step": 8025
    },
    {
      "entropy": 5.690315532684326,
      "epoch": 0.7969432314410481,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004945091549717033,
      "loss": 5.674,
      "mean_token_accuracy": 0.13399219289422035,
      "num_tokens": 17596774.0,
      "step": 8030
    },
    {
      "entropy": 5.663340473175049,
      "epoch": 0.7974394601032155,
      "grad_norm": 1.375,
      "learning_rate": 0.0004945013724771356,
      "loss": 5.517,
      "mean_token_accuracy": 0.14832287654280663,
      "num_tokens": 17606557.0,
      "step": 8035
    },
    {
      "entropy": 5.592705154418946,
      "epoch": 0.7979356887653831,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004944935845394069,
      "loss": 5.3979,
      "mean_token_accuracy": 0.14695664644241332,
      "num_tokens": 17616465.0,
      "step": 8040
    },
    {
      "entropy": 5.635967969894409,
      "epoch": 0.7984319174275506,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004944857911587103,
      "loss": 5.572,
      "mean_token_accuracy": 0.13814229220151902,
      "num_tokens": 17626907.0,
      "step": 8045
    },
    {
      "entropy": 5.737808418273926,
      "epoch": 0.7989281460897182,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004944779923352389,
      "loss": 5.5664,
      "mean_token_accuracy": 0.1433405876159668,
      "num_tokens": 17638660.0,
      "step": 8050
    },
    {
      "entropy": 5.58116660118103,
      "epoch": 0.7994243747518857,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004944701880691864,
      "loss": 5.4767,
      "mean_token_accuracy": 0.14689321219921112,
      "num_tokens": 17650135.0,
      "step": 8055
    },
    {
      "entropy": 5.551852083206176,
      "epoch": 0.7999206034140532,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000494462378360746,
      "loss": 5.5138,
      "mean_token_accuracy": 0.14458329528570174,
      "num_tokens": 17660987.0,
      "step": 8060
    },
    {
      "entropy": 5.717930603027344,
      "epoch": 0.8004168320762207,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004944545632101117,
      "loss": 5.4468,
      "mean_token_accuracy": 0.14894245564937592,
      "num_tokens": 17672520.0,
      "step": 8065
    },
    {
      "entropy": 5.6782369136810305,
      "epoch": 0.8009130607383882,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004944467426174769,
      "loss": 5.5147,
      "mean_token_accuracy": 0.14444962739944459,
      "num_tokens": 17683688.0,
      "step": 8070
    },
    {
      "entropy": 5.54812068939209,
      "epoch": 0.8014092894005558,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000494438916583036,
      "loss": 5.4443,
      "mean_token_accuracy": 0.1481657952070236,
      "num_tokens": 17695213.0,
      "step": 8075
    },
    {
      "entropy": 5.614107990264893,
      "epoch": 0.8019055180627233,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004944310851069826,
      "loss": 5.5266,
      "mean_token_accuracy": 0.14110470414161683,
      "num_tokens": 17705545.0,
      "step": 8080
    },
    {
      "entropy": 5.672084760665894,
      "epoch": 0.8024017467248908,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004944232481895112,
      "loss": 5.5122,
      "mean_token_accuracy": 0.1496380865573883,
      "num_tokens": 17715855.0,
      "step": 8085
    },
    {
      "entropy": 5.663711261749268,
      "epoch": 0.8028979753870583,
      "grad_norm": 1.265625,
      "learning_rate": 0.000494415405830816,
      "loss": 5.5836,
      "mean_token_accuracy": 0.14323716685175897,
      "num_tokens": 17727764.0,
      "step": 8090
    },
    {
      "entropy": 5.638366651535034,
      "epoch": 0.8033942040492259,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004944075580310916,
      "loss": 5.5596,
      "mean_token_accuracy": 0.14378135725855828,
      "num_tokens": 17738464.0,
      "step": 8095
    },
    {
      "entropy": 5.655952787399292,
      "epoch": 0.8038904327113934,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004943997047905325,
      "loss": 5.5372,
      "mean_token_accuracy": 0.1485045626759529,
      "num_tokens": 17747502.0,
      "step": 8100
    },
    {
      "entropy": 5.606498050689697,
      "epoch": 0.804386661373561,
      "grad_norm": 3.421875,
      "learning_rate": 0.0004943918461093335,
      "loss": 5.5504,
      "mean_token_accuracy": 0.14124829024076463,
      "num_tokens": 17759520.0,
      "step": 8105
    },
    {
      "entropy": 5.616118383407593,
      "epoch": 0.8048828900357284,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004943839819876894,
      "loss": 5.5198,
      "mean_token_accuracy": 0.14138075187802315,
      "num_tokens": 17770631.0,
      "step": 8110
    },
    {
      "entropy": 5.67958517074585,
      "epoch": 0.805379118697896,
      "grad_norm": 2.828125,
      "learning_rate": 0.0004943761124257954,
      "loss": 5.4741,
      "mean_token_accuracy": 0.1408017747104168,
      "num_tokens": 17781320.0,
      "step": 8115
    },
    {
      "entropy": 5.649888277053833,
      "epoch": 0.8058753473600635,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004943682374238464,
      "loss": 5.5256,
      "mean_token_accuracy": 0.14489989951252938,
      "num_tokens": 17792627.0,
      "step": 8120
    },
    {
      "entropy": 5.602329206466675,
      "epoch": 0.8063715760222311,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000494360356982038,
      "loss": 5.5661,
      "mean_token_accuracy": 0.139420023560524,
      "num_tokens": 17803286.0,
      "step": 8125
    },
    {
      "entropy": 5.559131622314453,
      "epoch": 0.8068678046843986,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004943524711005653,
      "loss": 5.3796,
      "mean_token_accuracy": 0.1471578486263752,
      "num_tokens": 17813631.0,
      "step": 8130
    },
    {
      "entropy": 5.592371654510498,
      "epoch": 0.8073640333465661,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000494344579779624,
      "loss": 5.4917,
      "mean_token_accuracy": 0.1454719200730324,
      "num_tokens": 17823467.0,
      "step": 8135
    },
    {
      "entropy": 5.628396701812744,
      "epoch": 0.8078602620087336,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004943366830194098,
      "loss": 5.4842,
      "mean_token_accuracy": 0.14438703209161757,
      "num_tokens": 17833600.0,
      "step": 8140
    },
    {
      "entropy": 5.620648908615112,
      "epoch": 0.8083564906709012,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004943287808201183,
      "loss": 5.4779,
      "mean_token_accuracy": 0.13972817435860635,
      "num_tokens": 17843112.0,
      "step": 8145
    },
    {
      "entropy": 5.6279061794281,
      "epoch": 0.8088527193330687,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004943208731819459,
      "loss": 5.4867,
      "mean_token_accuracy": 0.14555032774806023,
      "num_tokens": 17853678.0,
      "step": 8150
    },
    {
      "entropy": 5.597820854187011,
      "epoch": 0.8093489479952362,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004943129601050883,
      "loss": 5.4452,
      "mean_token_accuracy": 0.15026480257511138,
      "num_tokens": 17863537.0,
      "step": 8155
    },
    {
      "entropy": 5.585662412643432,
      "epoch": 0.8098451766574037,
      "grad_norm": 4.96875,
      "learning_rate": 0.000494305041589742,
      "loss": 5.5448,
      "mean_token_accuracy": 0.13825273960828782,
      "num_tokens": 17873511.0,
      "step": 8160
    },
    {
      "entropy": 5.603287220001221,
      "epoch": 0.8103414053195712,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004942971176361031,
      "loss": 5.4721,
      "mean_token_accuracy": 0.14937687814235687,
      "num_tokens": 17884546.0,
      "step": 8165
    },
    {
      "entropy": 5.6172081470489506,
      "epoch": 0.8108376339817388,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004942891882443682,
      "loss": 5.4268,
      "mean_token_accuracy": 0.14770203977823257,
      "num_tokens": 17895331.0,
      "step": 8170
    },
    {
      "entropy": 5.612709903717041,
      "epoch": 0.8113338626439063,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000494281253414734,
      "loss": 5.5933,
      "mean_token_accuracy": 0.14232611358165742,
      "num_tokens": 17906465.0,
      "step": 8175
    },
    {
      "entropy": 5.611964845657349,
      "epoch": 0.8118300913060739,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004942733131473973,
      "loss": 5.5183,
      "mean_token_accuracy": 0.1408399298787117,
      "num_tokens": 17916936.0,
      "step": 8180
    },
    {
      "entropy": 5.741065454483032,
      "epoch": 0.8123263199682413,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004942653674425548,
      "loss": 5.5502,
      "mean_token_accuracy": 0.1387503318488598,
      "num_tokens": 17927522.0,
      "step": 8185
    },
    {
      "entropy": 5.65732707977295,
      "epoch": 0.8128225486304089,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004942574163004038,
      "loss": 5.5931,
      "mean_token_accuracy": 0.142314213514328,
      "num_tokens": 17938216.0,
      "step": 8190
    },
    {
      "entropy": 5.613345384597778,
      "epoch": 0.8133187772925764,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004942494597211412,
      "loss": 5.4795,
      "mean_token_accuracy": 0.14077685475349427,
      "num_tokens": 17949333.0,
      "step": 8195
    },
    {
      "entropy": 5.696963453292847,
      "epoch": 0.813815005954744,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004942414977049646,
      "loss": 5.5443,
      "mean_token_accuracy": 0.1364717774093151,
      "num_tokens": 17960178.0,
      "step": 8200
    },
    {
      "entropy": 5.640322017669678,
      "epoch": 0.8143112346169115,
      "grad_norm": 1.171875,
      "learning_rate": 0.000494233530252071,
      "loss": 5.5153,
      "mean_token_accuracy": 0.14458856731653214,
      "num_tokens": 17971180.0,
      "step": 8205
    },
    {
      "entropy": 5.598449087142944,
      "epoch": 0.814807463279079,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004942255573626585,
      "loss": 5.5316,
      "mean_token_accuracy": 0.1446111559867859,
      "num_tokens": 17982572.0,
      "step": 8210
    },
    {
      "entropy": 5.640258932113648,
      "epoch": 0.8153036919412465,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004942175790369244,
      "loss": 5.4548,
      "mean_token_accuracy": 0.15170608162879945,
      "num_tokens": 17993236.0,
      "step": 8215
    },
    {
      "entropy": 5.653780841827393,
      "epoch": 0.8157999206034141,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004942095952750669,
      "loss": 5.5893,
      "mean_token_accuracy": 0.14500430822372437,
      "num_tokens": 18004215.0,
      "step": 8220
    },
    {
      "entropy": 5.592516231536865,
      "epoch": 0.8162961492655816,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004942016060772837,
      "loss": 5.4782,
      "mean_token_accuracy": 0.148300664126873,
      "num_tokens": 18013880.0,
      "step": 8225
    },
    {
      "entropy": 5.616851091384888,
      "epoch": 0.8167923779277491,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004941936114437731,
      "loss": 5.5484,
      "mean_token_accuracy": 0.14408496618270875,
      "num_tokens": 18024928.0,
      "step": 8230
    },
    {
      "entropy": 5.667749357223511,
      "epoch": 0.8172886065899166,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004941856113747331,
      "loss": 5.4914,
      "mean_token_accuracy": 0.1415318541228771,
      "num_tokens": 18035857.0,
      "step": 8235
    },
    {
      "entropy": 5.696399688720703,
      "epoch": 0.8177848352520841,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004941776058703623,
      "loss": 5.6197,
      "mean_token_accuracy": 0.14222904294729233,
      "num_tokens": 18046776.0,
      "step": 8240
    },
    {
      "entropy": 5.633800220489502,
      "epoch": 0.8182810639142517,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004941695949308593,
      "loss": 5.5551,
      "mean_token_accuracy": 0.14232334569096566,
      "num_tokens": 18057167.0,
      "step": 8245
    },
    {
      "entropy": 5.673133373260498,
      "epoch": 0.8187772925764192,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004941615785564224,
      "loss": 5.579,
      "mean_token_accuracy": 0.13846599608659743,
      "num_tokens": 18066953.0,
      "step": 8250
    },
    {
      "entropy": 5.634580898284912,
      "epoch": 0.8192735212385868,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004941535567472506,
      "loss": 5.4207,
      "mean_token_accuracy": 0.1431841015815735,
      "num_tokens": 18077977.0,
      "step": 8255
    },
    {
      "entropy": 5.620675802230835,
      "epoch": 0.8197697499007542,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000494145529503543,
      "loss": 5.5063,
      "mean_token_accuracy": 0.14469720274209977,
      "num_tokens": 18087581.0,
      "step": 8260
    },
    {
      "entropy": 5.565089273452759,
      "epoch": 0.8202659785629218,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004941374968254985,
      "loss": 5.4146,
      "mean_token_accuracy": 0.14956056848168373,
      "num_tokens": 18098193.0,
      "step": 8265
    },
    {
      "entropy": 5.581245994567871,
      "epoch": 0.8207622072250893,
      "grad_norm": 1.25,
      "learning_rate": 0.0004941294587133161,
      "loss": 5.539,
      "mean_token_accuracy": 0.14510907977819443,
      "num_tokens": 18110830.0,
      "step": 8270
    },
    {
      "entropy": 5.652122926712036,
      "epoch": 0.8212584358872569,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004941214151671954,
      "loss": 5.5538,
      "mean_token_accuracy": 0.13760428726673127,
      "num_tokens": 18122073.0,
      "step": 8275
    },
    {
      "entropy": 5.636314392089844,
      "epoch": 0.8217546645494244,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004941133661873357,
      "loss": 5.5316,
      "mean_token_accuracy": 0.14411617666482926,
      "num_tokens": 18132777.0,
      "step": 8280
    },
    {
      "entropy": 5.597688293457031,
      "epoch": 0.822250893211592,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004941053117739367,
      "loss": 5.443,
      "mean_token_accuracy": 0.1462380602955818,
      "num_tokens": 18143232.0,
      "step": 8285
    },
    {
      "entropy": 5.636086177825928,
      "epoch": 0.8227471218737594,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004940972519271982,
      "loss": 5.6315,
      "mean_token_accuracy": 0.1389746017754078,
      "num_tokens": 18155225.0,
      "step": 8290
    },
    {
      "entropy": 5.706859016418457,
      "epoch": 0.823243350535927,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004940891866473199,
      "loss": 5.556,
      "mean_token_accuracy": 0.14077101573348044,
      "num_tokens": 18165945.0,
      "step": 8295
    },
    {
      "entropy": 5.707482051849365,
      "epoch": 0.8237395791980945,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004940811159345017,
      "loss": 5.5867,
      "mean_token_accuracy": 0.13941670283675195,
      "num_tokens": 18177271.0,
      "step": 8300
    },
    {
      "entropy": 5.705527067184448,
      "epoch": 0.824235807860262,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004940730397889441,
      "loss": 5.5402,
      "mean_token_accuracy": 0.14304688423871995,
      "num_tokens": 18189732.0,
      "step": 8305
    },
    {
      "entropy": 5.684334135055542,
      "epoch": 0.8247320365224295,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004940649582108472,
      "loss": 5.6286,
      "mean_token_accuracy": 0.13693071380257607,
      "num_tokens": 18200731.0,
      "step": 8310
    },
    {
      "entropy": 5.678423595428467,
      "epoch": 0.825228265184597,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004940568712004113,
      "loss": 5.5395,
      "mean_token_accuracy": 0.14608079940080643,
      "num_tokens": 18210873.0,
      "step": 8315
    },
    {
      "entropy": 5.712420892715454,
      "epoch": 0.8257244938467646,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004940487787578369,
      "loss": 5.5368,
      "mean_token_accuracy": 0.13934941068291665,
      "num_tokens": 18222072.0,
      "step": 8320
    },
    {
      "entropy": 5.544324493408203,
      "epoch": 0.8262207225089321,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000494040680883325,
      "loss": 5.4479,
      "mean_token_accuracy": 0.13859313130378723,
      "num_tokens": 18232522.0,
      "step": 8325
    },
    {
      "entropy": 5.650290584564209,
      "epoch": 0.8267169511710997,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004940325775770762,
      "loss": 5.5405,
      "mean_token_accuracy": 0.13889424577355386,
      "num_tokens": 18243587.0,
      "step": 8330
    },
    {
      "entropy": 5.609315586090088,
      "epoch": 0.8272131798332671,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004940244688392913,
      "loss": 5.4763,
      "mean_token_accuracy": 0.14655772149562835,
      "num_tokens": 18254965.0,
      "step": 8335
    },
    {
      "entropy": 5.600151443481446,
      "epoch": 0.8277094084954347,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004940163546701716,
      "loss": 5.4705,
      "mean_token_accuracy": 0.14697985202074051,
      "num_tokens": 18265892.0,
      "step": 8340
    },
    {
      "entropy": 5.630267667770386,
      "epoch": 0.8282056371576022,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004940082350699183,
      "loss": 5.4803,
      "mean_token_accuracy": 0.14304709956049919,
      "num_tokens": 18276713.0,
      "step": 8345
    },
    {
      "entropy": 5.64806718826294,
      "epoch": 0.8287018658197698,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004940001100387325,
      "loss": 5.5345,
      "mean_token_accuracy": 0.1406284585595131,
      "num_tokens": 18288846.0,
      "step": 8350
    },
    {
      "entropy": 5.6618469715118405,
      "epoch": 0.8291980944819373,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004939919795768161,
      "loss": 5.54,
      "mean_token_accuracy": 0.14647294580936432,
      "num_tokens": 18298750.0,
      "step": 8355
    },
    {
      "entropy": 5.638445425033569,
      "epoch": 0.8296943231441049,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004939838436843704,
      "loss": 5.5235,
      "mean_token_accuracy": 0.14446348547935486,
      "num_tokens": 18310065.0,
      "step": 8360
    },
    {
      "entropy": 5.593844938278198,
      "epoch": 0.8301905518062723,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004939757023615973,
      "loss": 5.4709,
      "mean_token_accuracy": 0.14464512020349501,
      "num_tokens": 18319498.0,
      "step": 8365
    },
    {
      "entropy": 5.556368446350097,
      "epoch": 0.8306867804684399,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004939675556086986,
      "loss": 5.4477,
      "mean_token_accuracy": 0.14987604841589927,
      "num_tokens": 18329718.0,
      "step": 8370
    },
    {
      "entropy": 5.579064178466797,
      "epoch": 0.8311830091306074,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004939594034258763,
      "loss": 5.4667,
      "mean_token_accuracy": 0.14808508902788162,
      "num_tokens": 18341628.0,
      "step": 8375
    },
    {
      "entropy": 5.612956619262695,
      "epoch": 0.8316792377927749,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004939512458133326,
      "loss": 5.501,
      "mean_token_accuracy": 0.14876915365457535,
      "num_tokens": 18352869.0,
      "step": 8380
    },
    {
      "entropy": 5.617215967178344,
      "epoch": 0.8321754664549424,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004939430827712698,
      "loss": 5.5128,
      "mean_token_accuracy": 0.14319611415266992,
      "num_tokens": 18363664.0,
      "step": 8385
    },
    {
      "entropy": 5.673775148391724,
      "epoch": 0.8326716951171099,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004939349142998905,
      "loss": 5.4934,
      "mean_token_accuracy": 0.1411449357867241,
      "num_tokens": 18374325.0,
      "step": 8390
    },
    {
      "entropy": 5.601881647109986,
      "epoch": 0.8331679237792775,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004939267403993968,
      "loss": 5.4886,
      "mean_token_accuracy": 0.14374437406659127,
      "num_tokens": 18384547.0,
      "step": 8395
    },
    {
      "entropy": 5.607846355438232,
      "epoch": 0.833664152441445,
      "grad_norm": 2.28125,
      "learning_rate": 0.000493918561069992,
      "loss": 5.4525,
      "mean_token_accuracy": 0.1440906420350075,
      "num_tokens": 18395141.0,
      "step": 8400
    },
    {
      "entropy": 5.576317214965821,
      "epoch": 0.8341603811036126,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004939103763118784,
      "loss": 5.4793,
      "mean_token_accuracy": 0.14253750666975976,
      "num_tokens": 18406326.0,
      "step": 8405
    },
    {
      "entropy": 5.638917636871338,
      "epoch": 0.83465660976578,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004939021861252593,
      "loss": 5.5472,
      "mean_token_accuracy": 0.14200533032417298,
      "num_tokens": 18417521.0,
      "step": 8410
    },
    {
      "entropy": 5.659324312210083,
      "epoch": 0.8351528384279476,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004938939905103376,
      "loss": 5.4584,
      "mean_token_accuracy": 0.15007658153772355,
      "num_tokens": 18426813.0,
      "step": 8415
    },
    {
      "entropy": 5.574454021453858,
      "epoch": 0.8356490670901151,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004938857894673166,
      "loss": 5.419,
      "mean_token_accuracy": 0.14627117812633514,
      "num_tokens": 18438227.0,
      "step": 8420
    },
    {
      "entropy": 5.6082353591918945,
      "epoch": 0.8361452957522827,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004938775829963997,
      "loss": 5.5611,
      "mean_token_accuracy": 0.14154002815485,
      "num_tokens": 18448915.0,
      "step": 8425
    },
    {
      "entropy": 5.632203149795532,
      "epoch": 0.8366415244144502,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004938693710977904,
      "loss": 5.4437,
      "mean_token_accuracy": 0.1465119332075119,
      "num_tokens": 18459330.0,
      "step": 8430
    },
    {
      "entropy": 5.611264181137085,
      "epoch": 0.8371377530766178,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004938611537716922,
      "loss": 5.418,
      "mean_token_accuracy": 0.1501227267086506,
      "num_tokens": 18469813.0,
      "step": 8435
    },
    {
      "entropy": 5.5994469165802006,
      "epoch": 0.8376339817387852,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004938529310183091,
      "loss": 5.4724,
      "mean_token_accuracy": 0.14105386286973953,
      "num_tokens": 18479783.0,
      "step": 8440
    },
    {
      "entropy": 5.58183536529541,
      "epoch": 0.8381302104009528,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000493844702837845,
      "loss": 5.4856,
      "mean_token_accuracy": 0.14323654025793076,
      "num_tokens": 18489640.0,
      "step": 8445
    },
    {
      "entropy": 5.663530254364014,
      "epoch": 0.8386264390631203,
      "grad_norm": 1.375,
      "learning_rate": 0.0004938364692305036,
      "loss": 5.5607,
      "mean_token_accuracy": 0.14894553273916245,
      "num_tokens": 18500744.0,
      "step": 8450
    },
    {
      "entropy": 5.6642372608184814,
      "epoch": 0.8391226677252878,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004938282301964896,
      "loss": 5.4642,
      "mean_token_accuracy": 0.142983091622591,
      "num_tokens": 18513286.0,
      "step": 8455
    },
    {
      "entropy": 5.653742408752441,
      "epoch": 0.8396188963874553,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004938199857360068,
      "loss": 5.4282,
      "mean_token_accuracy": 0.1494394324719906,
      "num_tokens": 18522934.0,
      "step": 8460
    },
    {
      "entropy": 5.577210330963135,
      "epoch": 0.8401151250496228,
      "grad_norm": 1.515625,
      "learning_rate": 0.00049381173584926,
      "loss": 5.5111,
      "mean_token_accuracy": 0.14622715413570403,
      "num_tokens": 18534828.0,
      "step": 8465
    },
    {
      "entropy": 5.544417047500611,
      "epoch": 0.8406113537117904,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004938034805364535,
      "loss": 5.4676,
      "mean_token_accuracy": 0.14070868194103242,
      "num_tokens": 18544733.0,
      "step": 8470
    },
    {
      "entropy": 5.584074687957764,
      "epoch": 0.8411075823739579,
      "grad_norm": 1.625,
      "learning_rate": 0.0004937952197977923,
      "loss": 5.5014,
      "mean_token_accuracy": 0.1424806982278824,
      "num_tokens": 18554583.0,
      "step": 8475
    },
    {
      "entropy": 5.737908124923706,
      "epoch": 0.8416038110361255,
      "grad_norm": 1.265625,
      "learning_rate": 0.000493786953633481,
      "loss": 5.561,
      "mean_token_accuracy": 0.1397113934159279,
      "num_tokens": 18566503.0,
      "step": 8480
    },
    {
      "entropy": 5.65299620628357,
      "epoch": 0.842100039698293,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004937786820437248,
      "loss": 5.5041,
      "mean_token_accuracy": 0.14515818953514098,
      "num_tokens": 18578161.0,
      "step": 8485
    },
    {
      "entropy": 5.633429098129272,
      "epoch": 0.8425962683604605,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004937704050287286,
      "loss": 5.5747,
      "mean_token_accuracy": 0.14149901792407035,
      "num_tokens": 18588212.0,
      "step": 8490
    },
    {
      "entropy": 5.754056596755982,
      "epoch": 0.843092497022628,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004937621225886978,
      "loss": 5.5792,
      "mean_token_accuracy": 0.1345731757581234,
      "num_tokens": 18601293.0,
      "step": 8495
    },
    {
      "entropy": 5.675776958465576,
      "epoch": 0.8435887256847956,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004937538347238378,
      "loss": 5.614,
      "mean_token_accuracy": 0.1380243681371212,
      "num_tokens": 18613316.0,
      "step": 8500
    },
    {
      "entropy": 5.642468690872192,
      "epoch": 0.8440849543469631,
      "grad_norm": 1.40625,
      "learning_rate": 0.000493745541434354,
      "loss": 5.5437,
      "mean_token_accuracy": 0.1452810510993004,
      "num_tokens": 18624461.0,
      "step": 8505
    },
    {
      "entropy": 5.6171791553497314,
      "epoch": 0.8445811830091307,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004937372427204522,
      "loss": 5.4363,
      "mean_token_accuracy": 0.1436253473162651,
      "num_tokens": 18635165.0,
      "step": 8510
    },
    {
      "entropy": 5.645295572280884,
      "epoch": 0.8450774116712981,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004937289385823382,
      "loss": 5.5054,
      "mean_token_accuracy": 0.1488940477371216,
      "num_tokens": 18646649.0,
      "step": 8515
    },
    {
      "entropy": 5.5825573921203615,
      "epoch": 0.8455736403334657,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004937206290202176,
      "loss": 5.4748,
      "mean_token_accuracy": 0.14158470779657364,
      "num_tokens": 18658527.0,
      "step": 8520
    },
    {
      "entropy": 5.6378288745880125,
      "epoch": 0.8460698689956332,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004937123140342969,
      "loss": 5.4934,
      "mean_token_accuracy": 0.14320269525051116,
      "num_tokens": 18669463.0,
      "step": 8525
    },
    {
      "entropy": 5.669389724731445,
      "epoch": 0.8465660976578007,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000493703993624782,
      "loss": 5.5291,
      "mean_token_accuracy": 0.14309728443622588,
      "num_tokens": 18680806.0,
      "step": 8530
    },
    {
      "entropy": 5.5877429962158205,
      "epoch": 0.8470623263199683,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004936956677918794,
      "loss": 5.5252,
      "mean_token_accuracy": 0.14610027223825456,
      "num_tokens": 18692124.0,
      "step": 8535
    },
    {
      "entropy": 5.674967622756958,
      "epoch": 0.8475585549821357,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004936873365357955,
      "loss": 5.5297,
      "mean_token_accuracy": 0.1436805747449398,
      "num_tokens": 18702017.0,
      "step": 8540
    },
    {
      "entropy": 5.580541467666626,
      "epoch": 0.8480547836443033,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004936789998567368,
      "loss": 5.4368,
      "mean_token_accuracy": 0.14957460016012192,
      "num_tokens": 18713205.0,
      "step": 8545
    },
    {
      "entropy": 5.6030243873596195,
      "epoch": 0.8485510123064708,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004936706577549101,
      "loss": 5.4624,
      "mean_token_accuracy": 0.1454247087240219,
      "num_tokens": 18724165.0,
      "step": 8550
    },
    {
      "entropy": 5.621714115142822,
      "epoch": 0.8490472409686384,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004936623102305223,
      "loss": 5.5262,
      "mean_token_accuracy": 0.14346329271793365,
      "num_tokens": 18735155.0,
      "step": 8555
    },
    {
      "entropy": 5.622293996810913,
      "epoch": 0.8495434696308058,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004936539572837804,
      "loss": 5.5951,
      "mean_token_accuracy": 0.14516284465789794,
      "num_tokens": 18746228.0,
      "step": 8560
    },
    {
      "entropy": 5.615607643127442,
      "epoch": 0.8500396982929734,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004936455989148916,
      "loss": 5.5554,
      "mean_token_accuracy": 0.13957604244351388,
      "num_tokens": 18756638.0,
      "step": 8565
    },
    {
      "entropy": 5.693037223815918,
      "epoch": 0.8505359269551409,
      "grad_norm": 1.125,
      "learning_rate": 0.0004936372351240629,
      "loss": 5.5252,
      "mean_token_accuracy": 0.143225060403347,
      "num_tokens": 18768651.0,
      "step": 8570
    },
    {
      "entropy": 5.704475593566895,
      "epoch": 0.8510321556173085,
      "grad_norm": 1.515625,
      "learning_rate": 0.000493628865911502,
      "loss": 5.6206,
      "mean_token_accuracy": 0.14038438424468042,
      "num_tokens": 18780399.0,
      "step": 8575
    },
    {
      "entropy": 5.613246154785156,
      "epoch": 0.851528384279476,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004936204912774163,
      "loss": 5.497,
      "mean_token_accuracy": 0.1432039074599743,
      "num_tokens": 18791927.0,
      "step": 8580
    },
    {
      "entropy": 5.563634300231934,
      "epoch": 0.8520246129416436,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004936121112220135,
      "loss": 5.4629,
      "mean_token_accuracy": 0.14175041019916534,
      "num_tokens": 18802779.0,
      "step": 8585
    },
    {
      "entropy": 5.598039770126343,
      "epoch": 0.852520841603811,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004936037257455015,
      "loss": 5.4528,
      "mean_token_accuracy": 0.1468744769692421,
      "num_tokens": 18813123.0,
      "step": 8590
    },
    {
      "entropy": 5.59857063293457,
      "epoch": 0.8530170702659786,
      "grad_norm": 1.328125,
      "learning_rate": 0.000493595334848088,
      "loss": 5.4628,
      "mean_token_accuracy": 0.15100581273436547,
      "num_tokens": 18823699.0,
      "step": 8595
    },
    {
      "entropy": 5.695687341690063,
      "epoch": 0.8535132989281461,
      "grad_norm": 1.375,
      "learning_rate": 0.0004935869385299813,
      "loss": 5.6293,
      "mean_token_accuracy": 0.14018359407782555,
      "num_tokens": 18835450.0,
      "step": 8600
    },
    {
      "entropy": 5.62524642944336,
      "epoch": 0.8540095275903136,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004935785367913896,
      "loss": 5.463,
      "mean_token_accuracy": 0.1412882089614868,
      "num_tokens": 18845919.0,
      "step": 8605
    },
    {
      "entropy": 5.66269211769104,
      "epoch": 0.8545057562524812,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004935701296325212,
      "loss": 5.5924,
      "mean_token_accuracy": 0.13443027883768083,
      "num_tokens": 18858376.0,
      "step": 8610
    },
    {
      "entropy": 5.6874439239501955,
      "epoch": 0.8550019849146486,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004935617170535845,
      "loss": 5.5402,
      "mean_token_accuracy": 0.145059797167778,
      "num_tokens": 18870456.0,
      "step": 8615
    },
    {
      "entropy": 5.649290323257446,
      "epoch": 0.8554982135768162,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004935532990547882,
      "loss": 5.55,
      "mean_token_accuracy": 0.14433663934469224,
      "num_tokens": 18882174.0,
      "step": 8620
    },
    {
      "entropy": 5.647044992446899,
      "epoch": 0.8559944422389837,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004935448756363409,
      "loss": 5.6804,
      "mean_token_accuracy": 0.14336168467998506,
      "num_tokens": 18892982.0,
      "step": 8625
    },
    {
      "entropy": 5.7488080024719235,
      "epoch": 0.8564906709011513,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004935364467984518,
      "loss": 5.4984,
      "mean_token_accuracy": 0.14699309170246125,
      "num_tokens": 18903557.0,
      "step": 8630
    },
    {
      "entropy": 5.654719829559326,
      "epoch": 0.8569868995633187,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004935280125413295,
      "loss": 5.6399,
      "mean_token_accuracy": 0.14200191497802733,
      "num_tokens": 18914912.0,
      "step": 8635
    },
    {
      "entropy": 5.6892798900604244,
      "epoch": 0.8574831282254863,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004935195728651836,
      "loss": 5.575,
      "mean_token_accuracy": 0.13965004906058312,
      "num_tokens": 18927773.0,
      "step": 8640
    },
    {
      "entropy": 5.654879188537597,
      "epoch": 0.8579793568876538,
      "grad_norm": 1.453125,
      "learning_rate": 0.000493511127770223,
      "loss": 5.5138,
      "mean_token_accuracy": 0.14328212216496466,
      "num_tokens": 18938313.0,
      "step": 8645
    },
    {
      "entropy": 5.586103391647339,
      "epoch": 0.8584755855498214,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004935026772566574,
      "loss": 5.4562,
      "mean_token_accuracy": 0.1505862683057785,
      "num_tokens": 18948600.0,
      "step": 8650
    },
    {
      "entropy": 5.66076512336731,
      "epoch": 0.8589718142119889,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004934942213246961,
      "loss": 5.543,
      "mean_token_accuracy": 0.1423726499080658,
      "num_tokens": 18958927.0,
      "step": 8655
    },
    {
      "entropy": 5.609540367126465,
      "epoch": 0.8594680428741565,
      "grad_norm": 1.296875,
      "learning_rate": 0.000493485759974549,
      "loss": 5.489,
      "mean_token_accuracy": 0.1372593894600868,
      "num_tokens": 18968625.0,
      "step": 8660
    },
    {
      "entropy": 5.573156881332397,
      "epoch": 0.8599642715363239,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004934772932064259,
      "loss": 5.433,
      "mean_token_accuracy": 0.1469096153974533,
      "num_tokens": 18979772.0,
      "step": 8665
    },
    {
      "entropy": 5.66607813835144,
      "epoch": 0.8604605001984915,
      "grad_norm": 1.5,
      "learning_rate": 0.0004934688210205367,
      "loss": 5.5922,
      "mean_token_accuracy": 0.14221949577331544,
      "num_tokens": 18991522.0,
      "step": 8670
    },
    {
      "entropy": 5.662872266769409,
      "epoch": 0.860956728860659,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004934603434170913,
      "loss": 5.5298,
      "mean_token_accuracy": 0.14501567035913468,
      "num_tokens": 19002479.0,
      "step": 8675
    },
    {
      "entropy": 5.637520885467529,
      "epoch": 0.8614529575228265,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004934518603963003,
      "loss": 5.6139,
      "mean_token_accuracy": 0.13431279808282853,
      "num_tokens": 19013254.0,
      "step": 8680
    },
    {
      "entropy": 5.608312320709229,
      "epoch": 0.861949186184994,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004934433719583739,
      "loss": 5.433,
      "mean_token_accuracy": 0.14385922849178315,
      "num_tokens": 19023943.0,
      "step": 8685
    },
    {
      "entropy": 5.71530065536499,
      "epoch": 0.8624454148471615,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004934348781035226,
      "loss": 5.5636,
      "mean_token_accuracy": 0.14278231859207152,
      "num_tokens": 19034672.0,
      "step": 8690
    },
    {
      "entropy": 5.698452138900757,
      "epoch": 0.8629416435093291,
      "grad_norm": 1.3125,
      "learning_rate": 0.000493426378831957,
      "loss": 5.5437,
      "mean_token_accuracy": 0.14793606474995613,
      "num_tokens": 19045758.0,
      "step": 8695
    },
    {
      "entropy": 5.618188762664795,
      "epoch": 0.8634378721714966,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004934178741438877,
      "loss": 5.5495,
      "mean_token_accuracy": 0.14717491567134858,
      "num_tokens": 19057526.0,
      "step": 8700
    },
    {
      "entropy": 5.536909198760986,
      "epoch": 0.8639341008336642,
      "grad_norm": 1.46875,
      "learning_rate": 0.000493409364039526,
      "loss": 5.4088,
      "mean_token_accuracy": 0.14602681025862693,
      "num_tokens": 19069657.0,
      "step": 8705
    },
    {
      "entropy": 5.587775325775146,
      "epoch": 0.8644303294958317,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004934008485190825,
      "loss": 5.4589,
      "mean_token_accuracy": 0.1469672739505768,
      "num_tokens": 19080699.0,
      "step": 8710
    },
    {
      "entropy": 5.583184862136841,
      "epoch": 0.8649265581579992,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004933923275827687,
      "loss": 5.4954,
      "mean_token_accuracy": 0.14155903458595276,
      "num_tokens": 19093041.0,
      "step": 8715
    },
    {
      "entropy": 5.7123151302337645,
      "epoch": 0.8654227868201667,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004933838012307957,
      "loss": 5.5876,
      "mean_token_accuracy": 0.13913892731070518,
      "num_tokens": 19104249.0,
      "step": 8720
    },
    {
      "entropy": 5.67575511932373,
      "epoch": 0.8659190154823343,
      "grad_norm": 1.328125,
      "learning_rate": 0.000493375269463375,
      "loss": 5.5214,
      "mean_token_accuracy": 0.13837366476655005,
      "num_tokens": 19114434.0,
      "step": 8725
    },
    {
      "entropy": 5.5801482677459715,
      "epoch": 0.8664152441445018,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004933667322807182,
      "loss": 5.5231,
      "mean_token_accuracy": 0.13905722573399543,
      "num_tokens": 19124814.0,
      "step": 8730
    },
    {
      "entropy": 5.587121868133545,
      "epoch": 0.8669114728066694,
      "grad_norm": 1.359375,
      "learning_rate": 0.000493358189683037,
      "loss": 5.5628,
      "mean_token_accuracy": 0.13832299858331681,
      "num_tokens": 19135205.0,
      "step": 8735
    },
    {
      "entropy": 5.603911018371582,
      "epoch": 0.8674077014688368,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004933496416705432,
      "loss": 5.4467,
      "mean_token_accuracy": 0.14453877732157708,
      "num_tokens": 19147117.0,
      "step": 8740
    },
    {
      "entropy": 5.613752746582032,
      "epoch": 0.8679039301310044,
      "grad_norm": 1.25,
      "learning_rate": 0.0004933410882434488,
      "loss": 5.5226,
      "mean_token_accuracy": 0.1434659019112587,
      "num_tokens": 19158131.0,
      "step": 8745
    },
    {
      "entropy": 5.585004281997681,
      "epoch": 0.8684001587931719,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004933325294019659,
      "loss": 5.3969,
      "mean_token_accuracy": 0.1479727417230606,
      "num_tokens": 19168820.0,
      "step": 8750
    },
    {
      "entropy": 5.663940334320069,
      "epoch": 0.8688963874553394,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004933239651463068,
      "loss": 5.5063,
      "mean_token_accuracy": 0.1495026595890522,
      "num_tokens": 19178845.0,
      "step": 8755
    },
    {
      "entropy": 5.642027139663696,
      "epoch": 0.869392616117507,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004933153954766837,
      "loss": 5.4859,
      "mean_token_accuracy": 0.1454274497926235,
      "num_tokens": 19190830.0,
      "step": 8760
    },
    {
      "entropy": 5.482628536224365,
      "epoch": 0.8698888447796744,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004933068203933092,
      "loss": 5.3334,
      "mean_token_accuracy": 0.1463322803378105,
      "num_tokens": 19202056.0,
      "step": 8765
    },
    {
      "entropy": 5.5951841354370115,
      "epoch": 0.870385073441842,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000493298239896396,
      "loss": 5.5301,
      "mean_token_accuracy": 0.1413949705660343,
      "num_tokens": 19213497.0,
      "step": 8770
    },
    {
      "entropy": 5.6953788757324215,
      "epoch": 0.8708813021040095,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004932896539861568,
      "loss": 5.5793,
      "mean_token_accuracy": 0.14131228625774384,
      "num_tokens": 19226103.0,
      "step": 8775
    },
    {
      "entropy": 5.659559726715088,
      "epoch": 0.8713775307661771,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004932810626628047,
      "loss": 5.5874,
      "mean_token_accuracy": 0.14216968417167664,
      "num_tokens": 19237297.0,
      "step": 8780
    },
    {
      "entropy": 5.612377071380616,
      "epoch": 0.8718737594283446,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004932724659265526,
      "loss": 5.4996,
      "mean_token_accuracy": 0.14469050243496895,
      "num_tokens": 19247537.0,
      "step": 8785
    },
    {
      "entropy": 5.59554615020752,
      "epoch": 0.8723699880905121,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004932638637776134,
      "loss": 5.4469,
      "mean_token_accuracy": 0.13974902555346488,
      "num_tokens": 19260465.0,
      "step": 8790
    },
    {
      "entropy": 5.659331893920898,
      "epoch": 0.8728662167526796,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000493255256216201,
      "loss": 5.5147,
      "mean_token_accuracy": 0.14172475710511206,
      "num_tokens": 19270534.0,
      "step": 8795
    },
    {
      "entropy": 5.603011894226074,
      "epoch": 0.8733624454148472,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004932466432425284,
      "loss": 5.529,
      "mean_token_accuracy": 0.14091283306479455,
      "num_tokens": 19281551.0,
      "step": 8800
    },
    {
      "entropy": 5.600729990005493,
      "epoch": 0.8738586740770147,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004932380248568094,
      "loss": 5.4662,
      "mean_token_accuracy": 0.15053176134824753,
      "num_tokens": 19293300.0,
      "step": 8805
    },
    {
      "entropy": 5.63754825592041,
      "epoch": 0.8743549027391823,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004932294010592575,
      "loss": 5.3988,
      "mean_token_accuracy": 0.15093474090099335,
      "num_tokens": 19302701.0,
      "step": 8810
    },
    {
      "entropy": 5.64299464225769,
      "epoch": 0.8748511314013497,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004932207718500868,
      "loss": 5.4879,
      "mean_token_accuracy": 0.13831244856119157,
      "num_tokens": 19313024.0,
      "step": 8815
    },
    {
      "entropy": 5.628232955932617,
      "epoch": 0.8753473600635173,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004932121372295112,
      "loss": 5.5057,
      "mean_token_accuracy": 0.14094580709934235,
      "num_tokens": 19324014.0,
      "step": 8820
    },
    {
      "entropy": 5.575881481170654,
      "epoch": 0.8758435887256848,
      "grad_norm": 1.375,
      "learning_rate": 0.0004932034971977448,
      "loss": 5.5236,
      "mean_token_accuracy": 0.14353002235293388,
      "num_tokens": 19334226.0,
      "step": 8825
    },
    {
      "entropy": 5.708151388168335,
      "epoch": 0.8763398173878523,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004931948517550019,
      "loss": 5.5709,
      "mean_token_accuracy": 0.13959897309541702,
      "num_tokens": 19344363.0,
      "step": 8830
    },
    {
      "entropy": 5.671544361114502,
      "epoch": 0.8768360460500199,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004931862009014968,
      "loss": 5.5667,
      "mean_token_accuracy": 0.13856880664825438,
      "num_tokens": 19354915.0,
      "step": 8835
    },
    {
      "entropy": 5.680130243301392,
      "epoch": 0.8773322747121873,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000493177544637444,
      "loss": 5.5754,
      "mean_token_accuracy": 0.1429696723818779,
      "num_tokens": 19365570.0,
      "step": 8840
    },
    {
      "entropy": 5.695643377304077,
      "epoch": 0.8778285033743549,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004931688829630582,
      "loss": 5.5037,
      "mean_token_accuracy": 0.14831582456827164,
      "num_tokens": 19375731.0,
      "step": 8845
    },
    {
      "entropy": 5.59115891456604,
      "epoch": 0.8783247320365224,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004931602158785544,
      "loss": 5.452,
      "mean_token_accuracy": 0.14365211874246597,
      "num_tokens": 19386248.0,
      "step": 8850
    },
    {
      "entropy": 5.600852632522583,
      "epoch": 0.87882096069869,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004931515433841472,
      "loss": 5.4575,
      "mean_token_accuracy": 0.15065710246562958,
      "num_tokens": 19397593.0,
      "step": 8855
    },
    {
      "entropy": 5.665887403488159,
      "epoch": 0.8793171893608575,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004931428654800518,
      "loss": 5.5542,
      "mean_token_accuracy": 0.13736669346690178,
      "num_tokens": 19409701.0,
      "step": 8860
    },
    {
      "entropy": 5.6141908168792725,
      "epoch": 0.879813418023025,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004931341821664834,
      "loss": 5.5182,
      "mean_token_accuracy": 0.13888438194990158,
      "num_tokens": 19420280.0,
      "step": 8865
    },
    {
      "entropy": 5.61778712272644,
      "epoch": 0.8803096466851925,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004931254934436573,
      "loss": 5.5016,
      "mean_token_accuracy": 0.14510925859212875,
      "num_tokens": 19432208.0,
      "step": 8870
    },
    {
      "entropy": 5.746235275268555,
      "epoch": 0.8808058753473601,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004931167993117891,
      "loss": 5.6305,
      "mean_token_accuracy": 0.13860925137996674,
      "num_tokens": 19443333.0,
      "step": 8875
    },
    {
      "entropy": 5.587093782424927,
      "epoch": 0.8813021040095276,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004931080997710942,
      "loss": 5.37,
      "mean_token_accuracy": 0.14729210063815118,
      "num_tokens": 19452867.0,
      "step": 8880
    },
    {
      "entropy": 5.528789329528808,
      "epoch": 0.8817983326716952,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004930993948217884,
      "loss": 5.4102,
      "mean_token_accuracy": 0.15188175737857817,
      "num_tokens": 19463021.0,
      "step": 8885
    },
    {
      "entropy": 5.6156854152679445,
      "epoch": 0.8822945613338626,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004930906844640876,
      "loss": 5.5557,
      "mean_token_accuracy": 0.13963332325220107,
      "num_tokens": 19474092.0,
      "step": 8890
    },
    {
      "entropy": 5.638601016998291,
      "epoch": 0.8827907899960302,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004930819686982077,
      "loss": 5.5216,
      "mean_token_accuracy": 0.13767966106534005,
      "num_tokens": 19486035.0,
      "step": 8895
    },
    {
      "entropy": 5.605260944366455,
      "epoch": 0.8832870186581977,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000493073247524365,
      "loss": 5.4668,
      "mean_token_accuracy": 0.14679210931062697,
      "num_tokens": 19497484.0,
      "step": 8900
    },
    {
      "entropy": 5.6260233402252195,
      "epoch": 0.8837832473203652,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004930645209427756,
      "loss": 5.5689,
      "mean_token_accuracy": 0.14138772189617158,
      "num_tokens": 19508463.0,
      "step": 8905
    },
    {
      "entropy": 5.649750661849976,
      "epoch": 0.8842794759825328,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000493055788953656,
      "loss": 5.4799,
      "mean_token_accuracy": 0.1453041225671768,
      "num_tokens": 19520260.0,
      "step": 8910
    },
    {
      "entropy": 5.66017632484436,
      "epoch": 0.8847757046447002,
      "grad_norm": 1.625,
      "learning_rate": 0.0004930470515572227,
      "loss": 5.4568,
      "mean_token_accuracy": 0.14568835496902466,
      "num_tokens": 19531621.0,
      "step": 8915
    },
    {
      "entropy": 5.638175344467163,
      "epoch": 0.8852719333068678,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004930383087536923,
      "loss": 5.438,
      "mean_token_accuracy": 0.14532916396856307,
      "num_tokens": 19543271.0,
      "step": 8920
    },
    {
      "entropy": 5.685996627807617,
      "epoch": 0.8857681619690353,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004930295605432818,
      "loss": 5.5952,
      "mean_token_accuracy": 0.14887821227312087,
      "num_tokens": 19555178.0,
      "step": 8925
    },
    {
      "entropy": 5.64556999206543,
      "epoch": 0.8862643906312029,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004930208069262079,
      "loss": 5.5237,
      "mean_token_accuracy": 0.1415979266166687,
      "num_tokens": 19565153.0,
      "step": 8930
    },
    {
      "entropy": 5.612462949752808,
      "epoch": 0.8867606192933704,
      "grad_norm": 1.5,
      "learning_rate": 0.0004930120479026878,
      "loss": 5.459,
      "mean_token_accuracy": 0.14630941599607467,
      "num_tokens": 19575608.0,
      "step": 8935
    },
    {
      "entropy": 5.6398132801055905,
      "epoch": 0.8872568479555379,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004930032834729388,
      "loss": 5.4212,
      "mean_token_accuracy": 0.15149863809347153,
      "num_tokens": 19587556.0,
      "step": 8940
    },
    {
      "entropy": 5.572930002212525,
      "epoch": 0.8877530766177054,
      "grad_norm": 1.46875,
      "learning_rate": 0.000492994513637178,
      "loss": 5.459,
      "mean_token_accuracy": 0.1443370074033737,
      "num_tokens": 19597341.0,
      "step": 8945
    },
    {
      "entropy": 5.617953205108643,
      "epoch": 0.888249305279873,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004929857383956229,
      "loss": 5.5015,
      "mean_token_accuracy": 0.1384238876402378,
      "num_tokens": 19608431.0,
      "step": 8950
    },
    {
      "entropy": 5.641038131713867,
      "epoch": 0.8887455339420405,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004929769577484914,
      "loss": 5.4789,
      "mean_token_accuracy": 0.15090042501688003,
      "num_tokens": 19619645.0,
      "step": 8955
    },
    {
      "entropy": 5.562280654907227,
      "epoch": 0.8892417626042081,
      "grad_norm": 1.296875,
      "learning_rate": 0.000492968171696001,
      "loss": 5.4466,
      "mean_token_accuracy": 0.15213699340820314,
      "num_tokens": 19630412.0,
      "step": 8960
    },
    {
      "entropy": 5.6368982791900635,
      "epoch": 0.8897379912663755,
      "grad_norm": 1.375,
      "learning_rate": 0.0004929593802383696,
      "loss": 5.5018,
      "mean_token_accuracy": 0.14288419634103774,
      "num_tokens": 19640055.0,
      "step": 8965
    },
    {
      "entropy": 5.6397336483001705,
      "epoch": 0.8902342199285431,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004929505833758152,
      "loss": 5.4987,
      "mean_token_accuracy": 0.13700045347213746,
      "num_tokens": 19650814.0,
      "step": 8970
    },
    {
      "entropy": 5.609755945205689,
      "epoch": 0.8907304485907106,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004929417811085561,
      "loss": 5.4958,
      "mean_token_accuracy": 0.14362689331173897,
      "num_tokens": 19663669.0,
      "step": 8975
    },
    {
      "entropy": 5.700658893585205,
      "epoch": 0.8912266772528781,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004929329734368103,
      "loss": 5.5651,
      "mean_token_accuracy": 0.14093375727534294,
      "num_tokens": 19675851.0,
      "step": 8980
    },
    {
      "entropy": 5.689395999908447,
      "epoch": 0.8917229059150457,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004929241603607964,
      "loss": 5.4528,
      "mean_token_accuracy": 0.14935582354664803,
      "num_tokens": 19686613.0,
      "step": 8985
    },
    {
      "entropy": 5.5647519588470455,
      "epoch": 0.8922191345772131,
      "grad_norm": 1.3125,
      "learning_rate": 0.000492915341880733,
      "loss": 5.3846,
      "mean_token_accuracy": 0.1500229209661484,
      "num_tokens": 19696419.0,
      "step": 8990
    },
    {
      "entropy": 5.58155632019043,
      "epoch": 0.8927153632393807,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004929065179968387,
      "loss": 5.4885,
      "mean_token_accuracy": 0.14050796031951904,
      "num_tokens": 19707269.0,
      "step": 8995
    },
    {
      "entropy": 5.7278733253479,
      "epoch": 0.8932115919015482,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004928976887093321,
      "loss": 5.54,
      "mean_token_accuracy": 0.14154871702194213,
      "num_tokens": 19716099.0,
      "step": 9000
    },
    {
      "epoch": 0.8932115919015482,
      "eval_entropy": 5.5592718176503,
      "eval_loss": 5.572774887084961,
      "eval_mean_token_accuracy": 0.14549926674158536,
      "eval_num_tokens": 19716099.0,
      "eval_runtime": 26.7335,
      "eval_samples_per_second": 1259.433,
      "eval_steps_per_second": 157.443,
      "step": 9000
    },
    {
      "entropy": 5.6516022205352785,
      "epoch": 0.8937078205637158,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004928888540184325,
      "loss": 5.5203,
      "mean_token_accuracy": 0.14577429816126825,
      "num_tokens": 19726632.0,
      "step": 9005
    },
    {
      "entropy": 5.560913753509522,
      "epoch": 0.8942040492258833,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004928800139243587,
      "loss": 5.4084,
      "mean_token_accuracy": 0.14791175127029418,
      "num_tokens": 19737929.0,
      "step": 9010
    },
    {
      "entropy": 5.583985137939453,
      "epoch": 0.8947002778880508,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004928711684273302,
      "loss": 5.4424,
      "mean_token_accuracy": 0.14828080236911773,
      "num_tokens": 19748191.0,
      "step": 9015
    },
    {
      "entropy": 5.588948583602905,
      "epoch": 0.8951965065502183,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004928623175275661,
      "loss": 5.4996,
      "mean_token_accuracy": 0.14654425531625748,
      "num_tokens": 19758892.0,
      "step": 9020
    },
    {
      "entropy": 5.578219747543335,
      "epoch": 0.8956927352123859,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004928534612252862,
      "loss": 5.3894,
      "mean_token_accuracy": 0.14915963858366013,
      "num_tokens": 19769120.0,
      "step": 9025
    },
    {
      "entropy": 5.603693866729737,
      "epoch": 0.8961889638745534,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004928445995207095,
      "loss": 5.4638,
      "mean_token_accuracy": 0.14898087680339814,
      "num_tokens": 19779484.0,
      "step": 9030
    },
    {
      "entropy": 5.610855674743652,
      "epoch": 0.896685192536721,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004928357324140564,
      "loss": 5.471,
      "mean_token_accuracy": 0.1460346832871437,
      "num_tokens": 19790438.0,
      "step": 9035
    },
    {
      "entropy": 5.658837890625,
      "epoch": 0.8971814211988884,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004928268599055465,
      "loss": 5.5799,
      "mean_token_accuracy": 0.14448008686304092,
      "num_tokens": 19801646.0,
      "step": 9040
    },
    {
      "entropy": 5.622379302978516,
      "epoch": 0.897677649861056,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004928179819953998,
      "loss": 5.4625,
      "mean_token_accuracy": 0.15114278495311737,
      "num_tokens": 19812887.0,
      "step": 9045
    },
    {
      "entropy": 5.584934949874878,
      "epoch": 0.8981738785232235,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004928090986838366,
      "loss": 5.5132,
      "mean_token_accuracy": 0.14292742013931276,
      "num_tokens": 19822938.0,
      "step": 9050
    },
    {
      "entropy": 5.586890840530396,
      "epoch": 0.898670107185391,
      "grad_norm": 1.359375,
      "learning_rate": 0.000492800209971077,
      "loss": 5.4361,
      "mean_token_accuracy": 0.15070079863071442,
      "num_tokens": 19833988.0,
      "step": 9055
    },
    {
      "entropy": 5.69871244430542,
      "epoch": 0.8991663358475586,
      "grad_norm": 1.25,
      "learning_rate": 0.0004927913158573416,
      "loss": 5.522,
      "mean_token_accuracy": 0.1457460805773735,
      "num_tokens": 19845112.0,
      "step": 9060
    },
    {
      "entropy": 5.605101251602173,
      "epoch": 0.899662564509726,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004927824163428508,
      "loss": 5.4819,
      "mean_token_accuracy": 0.14132517203688622,
      "num_tokens": 19856304.0,
      "step": 9065
    },
    {
      "entropy": 5.565920734405518,
      "epoch": 0.9001587931718936,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004927735114278254,
      "loss": 5.4883,
      "mean_token_accuracy": 0.1474359929561615,
      "num_tokens": 19866776.0,
      "step": 9070
    },
    {
      "entropy": 5.627792644500732,
      "epoch": 0.9006550218340611,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004927646011124861,
      "loss": 5.4515,
      "mean_token_accuracy": 0.14474983215332032,
      "num_tokens": 19876392.0,
      "step": 9075
    },
    {
      "entropy": 5.609489631652832,
      "epoch": 0.9011512504962287,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004927556853970539,
      "loss": 5.4845,
      "mean_token_accuracy": 0.1408396489918232,
      "num_tokens": 19887362.0,
      "step": 9080
    },
    {
      "entropy": 5.607910251617431,
      "epoch": 0.9016474791583962,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004927467642817499,
      "loss": 5.4657,
      "mean_token_accuracy": 0.14311303496360778,
      "num_tokens": 19898989.0,
      "step": 9085
    },
    {
      "entropy": 5.652491855621338,
      "epoch": 0.9021437078205637,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004927378377667955,
      "loss": 5.4841,
      "mean_token_accuracy": 0.13832548409700393,
      "num_tokens": 19910112.0,
      "step": 9090
    },
    {
      "entropy": 5.690199756622315,
      "epoch": 0.9026399364827312,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000492728905852412,
      "loss": 5.6025,
      "mean_token_accuracy": 0.14233850836753845,
      "num_tokens": 19922601.0,
      "step": 9095
    },
    {
      "entropy": 5.589430522918701,
      "epoch": 0.9031361651448988,
      "grad_norm": 2.265625,
      "learning_rate": 0.0004927199685388205,
      "loss": 5.5063,
      "mean_token_accuracy": 0.14448769390583038,
      "num_tokens": 19934153.0,
      "step": 9100
    },
    {
      "entropy": 5.5954100608825685,
      "epoch": 0.9036323938070663,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000492711025826243,
      "loss": 5.4576,
      "mean_token_accuracy": 0.14790784940123558,
      "num_tokens": 19944572.0,
      "step": 9105
    },
    {
      "entropy": 5.657532167434693,
      "epoch": 0.9041286224692339,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004927020777149014,
      "loss": 5.5774,
      "mean_token_accuracy": 0.13514921218156814,
      "num_tokens": 19956431.0,
      "step": 9110
    },
    {
      "entropy": 5.700297832489014,
      "epoch": 0.9046248511314013,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004926931242050171,
      "loss": 5.4823,
      "mean_token_accuracy": 0.13819015324115752,
      "num_tokens": 19968118.0,
      "step": 9115
    },
    {
      "entropy": 5.644583463668823,
      "epoch": 0.9051210797935689,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004926841652968125,
      "loss": 5.5678,
      "mean_token_accuracy": 0.14897729232907295,
      "num_tokens": 19979545.0,
      "step": 9120
    },
    {
      "entropy": 5.616912031173706,
      "epoch": 0.9056173084557364,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004926752009905099,
      "loss": 5.5203,
      "mean_token_accuracy": 0.1344015933573246,
      "num_tokens": 19989832.0,
      "step": 9125
    },
    {
      "entropy": 5.632850551605225,
      "epoch": 0.9061135371179039,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004926662312863311,
      "loss": 5.5267,
      "mean_token_accuracy": 0.14392296597361565,
      "num_tokens": 20000935.0,
      "step": 9130
    },
    {
      "entropy": 5.6061439037323,
      "epoch": 0.9066097657800715,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004926572561844988,
      "loss": 5.531,
      "mean_token_accuracy": 0.13784194365143776,
      "num_tokens": 20011485.0,
      "step": 9135
    },
    {
      "entropy": 5.619773912429809,
      "epoch": 0.9071059944422389,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004926482756852356,
      "loss": 5.3874,
      "mean_token_accuracy": 0.14934873431921006,
      "num_tokens": 20023043.0,
      "step": 9140
    },
    {
      "entropy": 5.655905771255493,
      "epoch": 0.9076022231044065,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004926392897887642,
      "loss": 5.4953,
      "mean_token_accuracy": 0.15270624458789825,
      "num_tokens": 20032437.0,
      "step": 9145
    },
    {
      "entropy": 5.626210355758667,
      "epoch": 0.908098451766574,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004926302984953073,
      "loss": 5.5321,
      "mean_token_accuracy": 0.14077846258878707,
      "num_tokens": 20044662.0,
      "step": 9150
    },
    {
      "entropy": 5.6496881484985355,
      "epoch": 0.9085946804287416,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004926213018050881,
      "loss": 5.5342,
      "mean_token_accuracy": 0.14376182556152345,
      "num_tokens": 20055924.0,
      "step": 9155
    },
    {
      "entropy": 5.617496299743652,
      "epoch": 0.9090909090909091,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004926122997183295,
      "loss": 5.4899,
      "mean_token_accuracy": 0.13914088681340217,
      "num_tokens": 20067909.0,
      "step": 9160
    },
    {
      "entropy": 5.619670438766479,
      "epoch": 0.9095871377530766,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004926032922352546,
      "loss": 5.4553,
      "mean_token_accuracy": 0.15400660037994385,
      "num_tokens": 20079333.0,
      "step": 9165
    },
    {
      "entropy": 5.6066210746765135,
      "epoch": 0.9100833664152441,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004925942793560872,
      "loss": 5.4368,
      "mean_token_accuracy": 0.14459981843829156,
      "num_tokens": 20089741.0,
      "step": 9170
    },
    {
      "entropy": 5.5871826171875,
      "epoch": 0.9105795950774117,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004925852610810504,
      "loss": 5.4543,
      "mean_token_accuracy": 0.14535497650504112,
      "num_tokens": 20100267.0,
      "step": 9175
    },
    {
      "entropy": 5.57427225112915,
      "epoch": 0.9110758237395792,
      "grad_norm": 1.484375,
      "learning_rate": 0.000492576237410368,
      "loss": 5.438,
      "mean_token_accuracy": 0.14965663105249405,
      "num_tokens": 20111964.0,
      "step": 9180
    },
    {
      "entropy": 5.643334865570068,
      "epoch": 0.9115720524017468,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004925672083442638,
      "loss": 5.4923,
      "mean_token_accuracy": 0.13781903982162474,
      "num_tokens": 20123132.0,
      "step": 9185
    },
    {
      "entropy": 5.559805679321289,
      "epoch": 0.9120682810639142,
      "grad_norm": 1.375,
      "learning_rate": 0.0004925581738829616,
      "loss": 5.4138,
      "mean_token_accuracy": 0.14788240641355516,
      "num_tokens": 20134050.0,
      "step": 9190
    },
    {
      "entropy": 5.607632827758789,
      "epoch": 0.9125645097260818,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004925491340266855,
      "loss": 5.4602,
      "mean_token_accuracy": 0.14647637456655502,
      "num_tokens": 20143814.0,
      "step": 9195
    },
    {
      "entropy": 5.693362903594971,
      "epoch": 0.9130607383882493,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004925400887756596,
      "loss": 5.5606,
      "mean_token_accuracy": 0.14718564450740815,
      "num_tokens": 20157647.0,
      "step": 9200
    },
    {
      "entropy": 5.612559032440186,
      "epoch": 0.9135569670504168,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004925310381301083,
      "loss": 5.4997,
      "mean_token_accuracy": 0.14863042831420897,
      "num_tokens": 20168366.0,
      "step": 9205
    },
    {
      "entropy": 5.622346544265747,
      "epoch": 0.9140531957125844,
      "grad_norm": 1.96875,
      "learning_rate": 0.000492521982090256,
      "loss": 5.5733,
      "mean_token_accuracy": 0.14517005831003188,
      "num_tokens": 20179362.0,
      "step": 9210
    },
    {
      "entropy": 5.635663032531738,
      "epoch": 0.9145494243747518,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004925129206563273,
      "loss": 5.4634,
      "mean_token_accuracy": 0.14707203209400177,
      "num_tokens": 20190320.0,
      "step": 9215
    },
    {
      "entropy": 5.585685110092163,
      "epoch": 0.9150456530369194,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004925038538285467,
      "loss": 5.4791,
      "mean_token_accuracy": 0.14481002911925317,
      "num_tokens": 20201114.0,
      "step": 9220
    },
    {
      "entropy": 5.636270570755005,
      "epoch": 0.9155418816990869,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004924947816071394,
      "loss": 5.4984,
      "mean_token_accuracy": 0.14542139768600465,
      "num_tokens": 20213351.0,
      "step": 9225
    },
    {
      "entropy": 5.573896884918213,
      "epoch": 0.9160381103612545,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00049248570399233,
      "loss": 5.4087,
      "mean_token_accuracy": 0.15508366972208024,
      "num_tokens": 20223376.0,
      "step": 9230
    },
    {
      "entropy": 5.637418222427368,
      "epoch": 0.916534339023422,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000492476620984344,
      "loss": 5.5196,
      "mean_token_accuracy": 0.14469873905181885,
      "num_tokens": 20234248.0,
      "step": 9235
    },
    {
      "entropy": 5.616708612442016,
      "epoch": 0.9170305676855895,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004924675325834063,
      "loss": 5.4716,
      "mean_token_accuracy": 0.14426101073622705,
      "num_tokens": 20245823.0,
      "step": 9240
    },
    {
      "entropy": 5.682889366149903,
      "epoch": 0.917526796347757,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004924584387897424,
      "loss": 5.524,
      "mean_token_accuracy": 0.14275336265563965,
      "num_tokens": 20257003.0,
      "step": 9245
    },
    {
      "entropy": 5.691528272628784,
      "epoch": 0.9180230250099246,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004924493396035777,
      "loss": 5.6524,
      "mean_token_accuracy": 0.13683875873684884,
      "num_tokens": 20268500.0,
      "step": 9250
    },
    {
      "entropy": 5.677854585647583,
      "epoch": 0.9185192536720921,
      "grad_norm": 1.296875,
      "learning_rate": 0.000492440235025138,
      "loss": 5.5197,
      "mean_token_accuracy": 0.14637840092182158,
      "num_tokens": 20279301.0,
      "step": 9255
    },
    {
      "entropy": 5.738165187835693,
      "epoch": 0.9190154823342597,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004924311250546491,
      "loss": 5.6225,
      "mean_token_accuracy": 0.13661914467811584,
      "num_tokens": 20291263.0,
      "step": 9260
    },
    {
      "entropy": 5.631165504455566,
      "epoch": 0.9195117109964271,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004924220096923367,
      "loss": 5.4998,
      "mean_token_accuracy": 0.14364999532699585,
      "num_tokens": 20301554.0,
      "step": 9265
    },
    {
      "entropy": 5.693079948425293,
      "epoch": 0.9200079396585947,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004924128889384271,
      "loss": 5.5285,
      "mean_token_accuracy": 0.13806682974100112,
      "num_tokens": 20314299.0,
      "step": 9270
    },
    {
      "entropy": 5.584093999862671,
      "epoch": 0.9205041683207622,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004924037627931463,
      "loss": 5.4854,
      "mean_token_accuracy": 0.14932286888360977,
      "num_tokens": 20324595.0,
      "step": 9275
    },
    {
      "entropy": 5.634900283813477,
      "epoch": 0.9210003969829297,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004923946312567206,
      "loss": 5.4699,
      "mean_token_accuracy": 0.14457962214946746,
      "num_tokens": 20335744.0,
      "step": 9280
    },
    {
      "entropy": 5.568596601486206,
      "epoch": 0.9214966256450973,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004923854943293765,
      "loss": 5.3739,
      "mean_token_accuracy": 0.14239849373698235,
      "num_tokens": 20346072.0,
      "step": 9285
    },
    {
      "entropy": 5.6049731254577635,
      "epoch": 0.9219928543072647,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004923763520113406,
      "loss": 5.4527,
      "mean_token_accuracy": 0.1572947159409523,
      "num_tokens": 20357747.0,
      "step": 9290
    },
    {
      "entropy": 5.667739772796631,
      "epoch": 0.9224890829694323,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004923672043028395,
      "loss": 5.4648,
      "mean_token_accuracy": 0.14514413923025132,
      "num_tokens": 20368838.0,
      "step": 9295
    },
    {
      "entropy": 5.627686595916748,
      "epoch": 0.9229853116315998,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004923580512041001,
      "loss": 5.5216,
      "mean_token_accuracy": 0.1438218206167221,
      "num_tokens": 20380844.0,
      "step": 9300
    },
    {
      "entropy": 5.647341775894165,
      "epoch": 0.9234815402937674,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004923488927153495,
      "loss": 5.5245,
      "mean_token_accuracy": 0.1451920486986637,
      "num_tokens": 20391739.0,
      "step": 9305
    },
    {
      "entropy": 5.655885314941406,
      "epoch": 0.9239777689559349,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004923397288368146,
      "loss": 5.4303,
      "mean_token_accuracy": 0.15129459649324417,
      "num_tokens": 20402571.0,
      "step": 9310
    },
    {
      "entropy": 5.6016197681427,
      "epoch": 0.9244739976181024,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004923305595687229,
      "loss": 5.4612,
      "mean_token_accuracy": 0.13848201781511307,
      "num_tokens": 20414766.0,
      "step": 9315
    },
    {
      "entropy": 5.55289134979248,
      "epoch": 0.9249702262802699,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004923213849113015,
      "loss": 5.4284,
      "mean_token_accuracy": 0.14882013499736785,
      "num_tokens": 20425508.0,
      "step": 9320
    },
    {
      "entropy": 5.581551456451416,
      "epoch": 0.9254664549424375,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004923122048647781,
      "loss": 5.4206,
      "mean_token_accuracy": 0.14649079740047455,
      "num_tokens": 20436128.0,
      "step": 9325
    },
    {
      "entropy": 5.62555422782898,
      "epoch": 0.925962683604605,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004923030194293802,
      "loss": 5.5165,
      "mean_token_accuracy": 0.14568652510643004,
      "num_tokens": 20447212.0,
      "step": 9330
    },
    {
      "entropy": 5.579163217544556,
      "epoch": 0.9264589122667726,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004922938286053358,
      "loss": 5.4266,
      "mean_token_accuracy": 0.1495523899793625,
      "num_tokens": 20457770.0,
      "step": 9335
    },
    {
      "entropy": 5.623593425750732,
      "epoch": 0.92695514092894,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004922846323928726,
      "loss": 5.5173,
      "mean_token_accuracy": 0.1445883110165596,
      "num_tokens": 20468172.0,
      "step": 9340
    },
    {
      "entropy": 5.609854412078858,
      "epoch": 0.9274513695911076,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004922754307922188,
      "loss": 5.439,
      "mean_token_accuracy": 0.1448766238987446,
      "num_tokens": 20479429.0,
      "step": 9345
    },
    {
      "entropy": 5.576745653152466,
      "epoch": 0.9279475982532751,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004922662238036024,
      "loss": 5.4098,
      "mean_token_accuracy": 0.14659286737442018,
      "num_tokens": 20490373.0,
      "step": 9350
    },
    {
      "entropy": 5.634701108932495,
      "epoch": 0.9284438269154426,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004922570114272518,
      "loss": 5.5277,
      "mean_token_accuracy": 0.1420856796205044,
      "num_tokens": 20500002.0,
      "step": 9355
    },
    {
      "entropy": 5.677306318283081,
      "epoch": 0.9289400555776102,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004922477936633955,
      "loss": 5.5887,
      "mean_token_accuracy": 0.13643433451652526,
      "num_tokens": 20512990.0,
      "step": 9360
    },
    {
      "entropy": 5.606054830551147,
      "epoch": 0.9294362842397776,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004922385705122621,
      "loss": 5.4968,
      "mean_token_accuracy": 0.14656497314572334,
      "num_tokens": 20522979.0,
      "step": 9365
    },
    {
      "entropy": 5.574995756149292,
      "epoch": 0.9299325129019452,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004922293419740802,
      "loss": 5.3757,
      "mean_token_accuracy": 0.15563810095191002,
      "num_tokens": 20534620.0,
      "step": 9370
    },
    {
      "entropy": 5.633768844604492,
      "epoch": 0.9304287415641127,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004922201080490787,
      "loss": 5.483,
      "mean_token_accuracy": 0.1407161258161068,
      "num_tokens": 20545526.0,
      "step": 9375
    },
    {
      "entropy": 5.661541748046875,
      "epoch": 0.9309249702262803,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004922108687374867,
      "loss": 5.6346,
      "mean_token_accuracy": 0.13962297067046164,
      "num_tokens": 20556665.0,
      "step": 9380
    },
    {
      "entropy": 5.631024408340454,
      "epoch": 0.9314211988884478,
      "grad_norm": 1.296875,
      "learning_rate": 0.000492201624039533,
      "loss": 5.4664,
      "mean_token_accuracy": 0.1432509981095791,
      "num_tokens": 20566054.0,
      "step": 9385
    },
    {
      "entropy": 5.717415618896484,
      "epoch": 0.9319174275506154,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004921923739554472,
      "loss": 5.5585,
      "mean_token_accuracy": 0.14087095260620117,
      "num_tokens": 20577922.0,
      "step": 9390
    },
    {
      "entropy": 5.585472249984742,
      "epoch": 0.9324136562127828,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004921831184854585,
      "loss": 5.3567,
      "mean_token_accuracy": 0.15475076138973237,
      "num_tokens": 20588473.0,
      "step": 9395
    },
    {
      "entropy": 5.572597455978394,
      "epoch": 0.9329098848749504,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004921738576297965,
      "loss": 5.5589,
      "mean_token_accuracy": 0.14509392380714417,
      "num_tokens": 20599371.0,
      "step": 9400
    },
    {
      "entropy": 5.566518020629883,
      "epoch": 0.9334061135371179,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004921645913886907,
      "loss": 5.4348,
      "mean_token_accuracy": 0.14319652318954468,
      "num_tokens": 20610066.0,
      "step": 9405
    },
    {
      "entropy": 5.655844020843506,
      "epoch": 0.9339023421992855,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004921553197623711,
      "loss": 5.5287,
      "mean_token_accuracy": 0.14110815972089769,
      "num_tokens": 20621804.0,
      "step": 9410
    },
    {
      "entropy": 5.6166425228118895,
      "epoch": 0.934398570861453,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004921460427510674,
      "loss": 5.4257,
      "mean_token_accuracy": 0.1534297823905945,
      "num_tokens": 20632706.0,
      "step": 9415
    },
    {
      "entropy": 5.577079582214355,
      "epoch": 0.9348947995236205,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004921367603550098,
      "loss": 5.5194,
      "mean_token_accuracy": 0.1397561699151993,
      "num_tokens": 20643303.0,
      "step": 9420
    },
    {
      "entropy": 5.639082574844361,
      "epoch": 0.935391028185788,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004921274725744283,
      "loss": 5.5048,
      "mean_token_accuracy": 0.1498285673558712,
      "num_tokens": 20653681.0,
      "step": 9425
    },
    {
      "entropy": 5.707145690917969,
      "epoch": 0.9358872568479555,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004921181794095536,
      "loss": 5.5004,
      "mean_token_accuracy": 0.144416993111372,
      "num_tokens": 20665599.0,
      "step": 9430
    },
    {
      "entropy": 5.529110956192016,
      "epoch": 0.9363834855101231,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004921088808606157,
      "loss": 5.3831,
      "mean_token_accuracy": 0.14773372560739517,
      "num_tokens": 20675726.0,
      "step": 9435
    },
    {
      "entropy": 5.623796558380127,
      "epoch": 0.9368797141722905,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004920995769278454,
      "loss": 5.5697,
      "mean_token_accuracy": 0.13812774270772935,
      "num_tokens": 20688083.0,
      "step": 9440
    },
    {
      "entropy": 5.667687940597534,
      "epoch": 0.9373759428344581,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004920902676114734,
      "loss": 5.4393,
      "mean_token_accuracy": 0.14736258238554,
      "num_tokens": 20697642.0,
      "step": 9445
    },
    {
      "entropy": 5.669509553909302,
      "epoch": 0.9378721714966256,
      "grad_norm": 3.203125,
      "learning_rate": 0.0004920809529117306,
      "loss": 5.545,
      "mean_token_accuracy": 0.14001129418611527,
      "num_tokens": 20709172.0,
      "step": 9450
    },
    {
      "entropy": 5.689115905761719,
      "epoch": 0.9383684001587932,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004920716328288479,
      "loss": 5.5741,
      "mean_token_accuracy": 0.13982260152697562,
      "num_tokens": 20720563.0,
      "step": 9455
    },
    {
      "entropy": 5.637839317321777,
      "epoch": 0.9388646288209607,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004920623073630566,
      "loss": 5.4493,
      "mean_token_accuracy": 0.15048351585865022,
      "num_tokens": 20730943.0,
      "step": 9460
    },
    {
      "entropy": 5.600427532196045,
      "epoch": 0.9393608574831283,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004920529765145876,
      "loss": 5.4272,
      "mean_token_accuracy": 0.14764460027217866,
      "num_tokens": 20740903.0,
      "step": 9465
    },
    {
      "entropy": 5.624932861328125,
      "epoch": 0.9398570861452957,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004920436402836726,
      "loss": 5.5348,
      "mean_token_accuracy": 0.14325265362858772,
      "num_tokens": 20750580.0,
      "step": 9470
    },
    {
      "entropy": 5.626466369628906,
      "epoch": 0.9403533148074633,
      "grad_norm": 1.421875,
      "learning_rate": 0.000492034298670543,
      "loss": 5.4497,
      "mean_token_accuracy": 0.14759498685598374,
      "num_tokens": 20761357.0,
      "step": 9475
    },
    {
      "entropy": 5.730616569519043,
      "epoch": 0.9408495434696308,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004920249516754304,
      "loss": 5.5984,
      "mean_token_accuracy": 0.13724254295229912,
      "num_tokens": 20773273.0,
      "step": 9480
    },
    {
      "entropy": 5.639264059066773,
      "epoch": 0.9413457721317984,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004920155992985667,
      "loss": 5.4487,
      "mean_token_accuracy": 0.14832991659641265,
      "num_tokens": 20784290.0,
      "step": 9485
    },
    {
      "entropy": 5.624183559417725,
      "epoch": 0.9418420007939659,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004920062415401837,
      "loss": 5.474,
      "mean_token_accuracy": 0.14461207538843154,
      "num_tokens": 20794764.0,
      "step": 9490
    },
    {
      "entropy": 5.616425085067749,
      "epoch": 0.9423382294561334,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004919968784005136,
      "loss": 5.5158,
      "mean_token_accuracy": 0.14104585647583007,
      "num_tokens": 20804863.0,
      "step": 9495
    },
    {
      "entropy": 5.702088832855225,
      "epoch": 0.9428344581183009,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004919875098797886,
      "loss": 5.5955,
      "mean_token_accuracy": 0.13837794438004494,
      "num_tokens": 20816445.0,
      "step": 9500
    },
    {
      "entropy": 5.659405851364136,
      "epoch": 0.9433306867804684,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004919781359782407,
      "loss": 5.4045,
      "mean_token_accuracy": 0.15390849560499192,
      "num_tokens": 20827141.0,
      "step": 9505
    },
    {
      "entropy": 5.607259464263916,
      "epoch": 0.943826915442636,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004919687566961027,
      "loss": 5.4692,
      "mean_token_accuracy": 0.13972658291459084,
      "num_tokens": 20838291.0,
      "step": 9510
    },
    {
      "entropy": 5.674178791046143,
      "epoch": 0.9443231441048034,
      "grad_norm": 1.34375,
      "learning_rate": 0.000491959372033607,
      "loss": 5.4781,
      "mean_token_accuracy": 0.14293897300958633,
      "num_tokens": 20848839.0,
      "step": 9515
    },
    {
      "entropy": 5.645138216018677,
      "epoch": 0.944819372766971,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004919499819909865,
      "loss": 5.4479,
      "mean_token_accuracy": 0.14682312458753585,
      "num_tokens": 20859961.0,
      "step": 9520
    },
    {
      "entropy": 5.548512697219849,
      "epoch": 0.9453156014291385,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004919405865684738,
      "loss": 5.3977,
      "mean_token_accuracy": 0.15042057633399963,
      "num_tokens": 20869720.0,
      "step": 9525
    },
    {
      "entropy": 5.622430849075317,
      "epoch": 0.9458118300913061,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000491931185766302,
      "loss": 5.533,
      "mean_token_accuracy": 0.1509256437420845,
      "num_tokens": 20880513.0,
      "step": 9530
    },
    {
      "entropy": 5.682025289535522,
      "epoch": 0.9463080587534736,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004919217795847044,
      "loss": 5.4617,
      "mean_token_accuracy": 0.15055643171072006,
      "num_tokens": 20892456.0,
      "step": 9535
    },
    {
      "entropy": 5.600494289398194,
      "epoch": 0.9468042874156412,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004919123680239139,
      "loss": 5.4669,
      "mean_token_accuracy": 0.1442364275455475,
      "num_tokens": 20904500.0,
      "step": 9540
    },
    {
      "entropy": 5.618598365783692,
      "epoch": 0.9473005160778086,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004919029510841641,
      "loss": 5.4628,
      "mean_token_accuracy": 0.1481289342045784,
      "num_tokens": 20915133.0,
      "step": 9545
    },
    {
      "entropy": 5.585205030441284,
      "epoch": 0.9477967447399762,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004918935287656887,
      "loss": 5.4265,
      "mean_token_accuracy": 0.1467646636068821,
      "num_tokens": 20924779.0,
      "step": 9550
    },
    {
      "entropy": 5.5477362155914305,
      "epoch": 0.9482929734021437,
      "grad_norm": 1.1875,
      "learning_rate": 0.000491884101068721,
      "loss": 5.3871,
      "mean_token_accuracy": 0.14474384784698485,
      "num_tokens": 20935387.0,
      "step": 9555
    },
    {
      "entropy": 5.64253306388855,
      "epoch": 0.9487892020643113,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004918746679934949,
      "loss": 5.5494,
      "mean_token_accuracy": 0.1444346785545349,
      "num_tokens": 20945432.0,
      "step": 9560
    },
    {
      "entropy": 5.622274875640869,
      "epoch": 0.9492854307264788,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004918652295402444,
      "loss": 5.4549,
      "mean_token_accuracy": 0.150359857827425,
      "num_tokens": 20956777.0,
      "step": 9565
    },
    {
      "entropy": 5.576067399978638,
      "epoch": 0.9497816593886463,
      "grad_norm": 1.625,
      "learning_rate": 0.0004918557857092036,
      "loss": 5.4692,
      "mean_token_accuracy": 0.1470870390534401,
      "num_tokens": 20966057.0,
      "step": 9570
    },
    {
      "entropy": 5.642852306365967,
      "epoch": 0.9502778880508138,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004918463365006065,
      "loss": 5.5685,
      "mean_token_accuracy": 0.14312680512666703,
      "num_tokens": 20976838.0,
      "step": 9575
    },
    {
      "entropy": 5.733105897903442,
      "epoch": 0.9507741167129813,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004918368819146875,
      "loss": 5.4965,
      "mean_token_accuracy": 0.14506655633449556,
      "num_tokens": 20988553.0,
      "step": 9580
    },
    {
      "entropy": 5.630043697357178,
      "epoch": 0.9512703453751489,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004918274219516811,
      "loss": 5.473,
      "mean_token_accuracy": 0.14528999328613282,
      "num_tokens": 20998983.0,
      "step": 9585
    },
    {
      "entropy": 5.677074146270752,
      "epoch": 0.9517665740373163,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004918179566118218,
      "loss": 5.568,
      "mean_token_accuracy": 0.13747491985559462,
      "num_tokens": 21010008.0,
      "step": 9590
    },
    {
      "entropy": 5.632538175582885,
      "epoch": 0.9522628026994839,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004918084858953445,
      "loss": 5.5566,
      "mean_token_accuracy": 0.14554011598229408,
      "num_tokens": 21022826.0,
      "step": 9595
    },
    {
      "entropy": 5.648578357696533,
      "epoch": 0.9527590313616514,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004917990098024839,
      "loss": 5.4706,
      "mean_token_accuracy": 0.14844711273908615,
      "num_tokens": 21033457.0,
      "step": 9600
    },
    {
      "entropy": 5.569852495193482,
      "epoch": 0.953255260023819,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004917895283334749,
      "loss": 5.5081,
      "mean_token_accuracy": 0.14212347865104674,
      "num_tokens": 21044640.0,
      "step": 9605
    },
    {
      "entropy": 5.669201803207398,
      "epoch": 0.9537514886859865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000491780041488553,
      "loss": 5.4886,
      "mean_token_accuracy": 0.14830565080046654,
      "num_tokens": 21055728.0,
      "step": 9610
    },
    {
      "entropy": 5.641808748245239,
      "epoch": 0.9542477173481541,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004917705492679529,
      "loss": 5.4773,
      "mean_token_accuracy": 0.14648034423589706,
      "num_tokens": 21066057.0,
      "step": 9615
    },
    {
      "entropy": 5.643576669692993,
      "epoch": 0.9547439460103215,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004917610516719104,
      "loss": 5.4721,
      "mean_token_accuracy": 0.1409212000668049,
      "num_tokens": 21075717.0,
      "step": 9620
    },
    {
      "entropy": 5.710641622543335,
      "epoch": 0.9552401746724891,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004917515487006609,
      "loss": 5.5806,
      "mean_token_accuracy": 0.13675020188093184,
      "num_tokens": 21087120.0,
      "step": 9625
    },
    {
      "entropy": 5.717125320434571,
      "epoch": 0.9557364033346566,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00049174204035444,
      "loss": 5.5891,
      "mean_token_accuracy": 0.1364285506308079,
      "num_tokens": 21100110.0,
      "step": 9630
    },
    {
      "entropy": 5.735737657546997,
      "epoch": 0.9562326319968242,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004917325266334836,
      "loss": 5.5047,
      "mean_token_accuracy": 0.14622458070516586,
      "num_tokens": 21110411.0,
      "step": 9635
    },
    {
      "entropy": 5.542178201675415,
      "epoch": 0.9567288606589917,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004917230075380276,
      "loss": 5.4156,
      "mean_token_accuracy": 0.14776762574911118,
      "num_tokens": 21120303.0,
      "step": 9640
    },
    {
      "entropy": 5.619141960144043,
      "epoch": 0.9572250893211592,
      "grad_norm": 1.25,
      "learning_rate": 0.0004917134830683079,
      "loss": 5.5285,
      "mean_token_accuracy": 0.14223454743623734,
      "num_tokens": 21132312.0,
      "step": 9645
    },
    {
      "entropy": 5.637432813644409,
      "epoch": 0.9577213179833267,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004917039532245608,
      "loss": 5.4779,
      "mean_token_accuracy": 0.14544432908296584,
      "num_tokens": 21142472.0,
      "step": 9650
    },
    {
      "entropy": 5.602258157730103,
      "epoch": 0.9582175466454942,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004916944180070226,
      "loss": 5.4867,
      "mean_token_accuracy": 0.13940309807658197,
      "num_tokens": 21153306.0,
      "step": 9655
    },
    {
      "entropy": 5.731304502487182,
      "epoch": 0.9587137753076618,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004916848774159299,
      "loss": 5.5425,
      "mean_token_accuracy": 0.13948502615094185,
      "num_tokens": 21165606.0,
      "step": 9660
    },
    {
      "entropy": 5.662382411956787,
      "epoch": 0.9592100039698293,
      "grad_norm": 1.359375,
      "learning_rate": 0.000491675331451519,
      "loss": 5.4784,
      "mean_token_accuracy": 0.1481885939836502,
      "num_tokens": 21176734.0,
      "step": 9665
    },
    {
      "entropy": 5.611044979095459,
      "epoch": 0.9597062326319968,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004916657801140269,
      "loss": 5.4188,
      "mean_token_accuracy": 0.1522512599825859,
      "num_tokens": 21187109.0,
      "step": 9670
    },
    {
      "entropy": 5.610244989395142,
      "epoch": 0.9602024612941643,
      "grad_norm": 1.375,
      "learning_rate": 0.0004916562234036903,
      "loss": 5.4611,
      "mean_token_accuracy": 0.14884927421808242,
      "num_tokens": 21197281.0,
      "step": 9675
    },
    {
      "entropy": 5.592353200912475,
      "epoch": 0.9606986899563319,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004916466613207462,
      "loss": 5.5054,
      "mean_token_accuracy": 0.14292329400777817,
      "num_tokens": 21208767.0,
      "step": 9680
    },
    {
      "entropy": 5.593602037429809,
      "epoch": 0.9611949186184994,
      "grad_norm": 1.5,
      "learning_rate": 0.0004916370938654317,
      "loss": 5.4869,
      "mean_token_accuracy": 0.14479880556464195,
      "num_tokens": 21219328.0,
      "step": 9685
    },
    {
      "entropy": 5.733356189727783,
      "epoch": 0.961691147280667,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004916275210379841,
      "loss": 5.6132,
      "mean_token_accuracy": 0.14567993432283402,
      "num_tokens": 21231476.0,
      "step": 9690
    },
    {
      "entropy": 5.687578916549683,
      "epoch": 0.9621873759428344,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004916179428386408,
      "loss": 5.5485,
      "mean_token_accuracy": 0.14798671901226043,
      "num_tokens": 21242445.0,
      "step": 9695
    },
    {
      "entropy": 5.597250032424927,
      "epoch": 0.962683604605002,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004916083592676392,
      "loss": 5.3663,
      "mean_token_accuracy": 0.1499406322836876,
      "num_tokens": 21253179.0,
      "step": 9700
    },
    {
      "entropy": 5.663527631759644,
      "epoch": 0.9631798332671695,
      "grad_norm": 1.75,
      "learning_rate": 0.0004915987703252171,
      "loss": 5.4011,
      "mean_token_accuracy": 0.15301899015903472,
      "num_tokens": 21263460.0,
      "step": 9705
    },
    {
      "entropy": 5.539344453811646,
      "epoch": 0.9636760619293371,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004915891760116123,
      "loss": 5.451,
      "mean_token_accuracy": 0.14565740525722504,
      "num_tokens": 21274819.0,
      "step": 9710
    },
    {
      "entropy": 5.586298274993896,
      "epoch": 0.9641722905915046,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004915795763270626,
      "loss": 5.4399,
      "mean_token_accuracy": 0.14477425664663315,
      "num_tokens": 21286215.0,
      "step": 9715
    },
    {
      "entropy": 5.613454580307007,
      "epoch": 0.9646685192536721,
      "grad_norm": 1.40625,
      "learning_rate": 0.000491569971271806,
      "loss": 5.4381,
      "mean_token_accuracy": 0.14988733530044557,
      "num_tokens": 21297033.0,
      "step": 9720
    },
    {
      "entropy": 5.536926984786987,
      "epoch": 0.9651647479158396,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004915603608460809,
      "loss": 5.4166,
      "mean_token_accuracy": 0.14530788511037826,
      "num_tokens": 21308019.0,
      "step": 9725
    },
    {
      "entropy": 5.616249704360962,
      "epoch": 0.9656609765780071,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004915507450501254,
      "loss": 5.5407,
      "mean_token_accuracy": 0.14036298021674157,
      "num_tokens": 21319313.0,
      "step": 9730
    },
    {
      "entropy": 5.64888825416565,
      "epoch": 0.9661572052401747,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000491541123884178,
      "loss": 5.4607,
      "mean_token_accuracy": 0.1480522185564041,
      "num_tokens": 21329950.0,
      "step": 9735
    },
    {
      "entropy": 5.626601791381836,
      "epoch": 0.9666534339023422,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004915314973484775,
      "loss": 5.4321,
      "mean_token_accuracy": 0.14527367055416107,
      "num_tokens": 21339061.0,
      "step": 9740
    },
    {
      "entropy": 5.582049322128296,
      "epoch": 0.9671496625645097,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004915218654432624,
      "loss": 5.5192,
      "mean_token_accuracy": 0.14555648863315582,
      "num_tokens": 21350114.0,
      "step": 9745
    },
    {
      "entropy": 5.604274845123291,
      "epoch": 0.9676458912266772,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004915122281687716,
      "loss": 5.4379,
      "mean_token_accuracy": 0.14853436946868898,
      "num_tokens": 21361219.0,
      "step": 9750
    },
    {
      "entropy": 5.6261931419372555,
      "epoch": 0.9681421198888448,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004915025855252441,
      "loss": 5.4433,
      "mean_token_accuracy": 0.14572725892066957,
      "num_tokens": 21372722.0,
      "step": 9755
    },
    {
      "entropy": 5.6248949527740475,
      "epoch": 0.9686383485510123,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004914929375129188,
      "loss": 5.4698,
      "mean_token_accuracy": 0.1452346235513687,
      "num_tokens": 21383350.0,
      "step": 9760
    },
    {
      "entropy": 5.640792560577393,
      "epoch": 0.9691345772131799,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004914832841320353,
      "loss": 5.5147,
      "mean_token_accuracy": 0.1461635112762451,
      "num_tokens": 21394656.0,
      "step": 9765
    },
    {
      "entropy": 5.636889934539795,
      "epoch": 0.9696308058753473,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000491473625382833,
      "loss": 5.4989,
      "mean_token_accuracy": 0.14275254160165787,
      "num_tokens": 21406163.0,
      "step": 9770
    },
    {
      "entropy": 5.526058626174927,
      "epoch": 0.9701270345375149,
      "grad_norm": 1.53125,
      "learning_rate": 0.000491463961265551,
      "loss": 5.4081,
      "mean_token_accuracy": 0.14551704972982407,
      "num_tokens": 21418703.0,
      "step": 9775
    },
    {
      "entropy": 5.577323913574219,
      "epoch": 0.9706232631996824,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004914542917804292,
      "loss": 5.4049,
      "mean_token_accuracy": 0.15396643429994583,
      "num_tokens": 21428511.0,
      "step": 9780
    },
    {
      "entropy": 5.612396335601806,
      "epoch": 0.97111949186185,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004914446169277075,
      "loss": 5.4489,
      "mean_token_accuracy": 0.1477150410413742,
      "num_tokens": 21440350.0,
      "step": 9785
    },
    {
      "entropy": 5.689861154556274,
      "epoch": 0.9716157205240175,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004914349367076258,
      "loss": 5.5426,
      "mean_token_accuracy": 0.14394681751728058,
      "num_tokens": 21451262.0,
      "step": 9790
    },
    {
      "entropy": 5.625470352172852,
      "epoch": 0.972111949186185,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004914252511204238,
      "loss": 5.5079,
      "mean_token_accuracy": 0.14273252710700035,
      "num_tokens": 21462587.0,
      "step": 9795
    },
    {
      "entropy": 5.614661645889282,
      "epoch": 0.9726081778483525,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004914155601663422,
      "loss": 5.4639,
      "mean_token_accuracy": 0.1487550415098667,
      "num_tokens": 21473452.0,
      "step": 9800
    },
    {
      "entropy": 5.645523500442505,
      "epoch": 0.97310440651052,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004914058638456208,
      "loss": 5.4392,
      "mean_token_accuracy": 0.14208987951278687,
      "num_tokens": 21484011.0,
      "step": 9805
    },
    {
      "entropy": 5.5301783084869385,
      "epoch": 0.9736006351726876,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004913961621585003,
      "loss": 5.414,
      "mean_token_accuracy": 0.14731867760419845,
      "num_tokens": 21495959.0,
      "step": 9810
    },
    {
      "entropy": 5.596349143981934,
      "epoch": 0.974096863834855,
      "grad_norm": 1.375,
      "learning_rate": 0.0004913864551052214,
      "loss": 5.4552,
      "mean_token_accuracy": 0.1569521501660347,
      "num_tokens": 21507496.0,
      "step": 9815
    },
    {
      "entropy": 5.67779769897461,
      "epoch": 0.9745930924970226,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004913767426860246,
      "loss": 5.5331,
      "mean_token_accuracy": 0.14126116633415223,
      "num_tokens": 21517532.0,
      "step": 9820
    },
    {
      "entropy": 5.564006996154785,
      "epoch": 0.9750893211591901,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004913670249011508,
      "loss": 5.3529,
      "mean_token_accuracy": 0.15146676301956177,
      "num_tokens": 21527969.0,
      "step": 9825
    },
    {
      "entropy": 5.559222412109375,
      "epoch": 0.9755855498213577,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004913573017508411,
      "loss": 5.4594,
      "mean_token_accuracy": 0.14351911693811417,
      "num_tokens": 21538243.0,
      "step": 9830
    },
    {
      "entropy": 5.608201599121093,
      "epoch": 0.9760817784835252,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004913475732353365,
      "loss": 5.4769,
      "mean_token_accuracy": 0.14904194995760917,
      "num_tokens": 21548702.0,
      "step": 9835
    },
    {
      "entropy": 5.5619007587432865,
      "epoch": 0.9765780071456928,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004913378393548784,
      "loss": 5.4338,
      "mean_token_accuracy": 0.1480866104364395,
      "num_tokens": 21558583.0,
      "step": 9840
    },
    {
      "entropy": 5.608968305587768,
      "epoch": 0.9770742358078602,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004913281001097078,
      "loss": 5.4158,
      "mean_token_accuracy": 0.1543859586119652,
      "num_tokens": 21569845.0,
      "step": 9845
    },
    {
      "entropy": 5.680455255508423,
      "epoch": 0.9775704644700278,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004913183555000667,
      "loss": 5.5675,
      "mean_token_accuracy": 0.14055220857262612,
      "num_tokens": 21580540.0,
      "step": 9850
    },
    {
      "entropy": 5.67117486000061,
      "epoch": 0.9780666931321953,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004913086055261963,
      "loss": 5.4126,
      "mean_token_accuracy": 0.14745607674121858,
      "num_tokens": 21590982.0,
      "step": 9855
    },
    {
      "entropy": 5.627556467056275,
      "epoch": 0.9785629217943629,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004912988501883387,
      "loss": 5.4467,
      "mean_token_accuracy": 0.14832669347524643,
      "num_tokens": 21602817.0,
      "step": 9860
    },
    {
      "entropy": 5.627848768234253,
      "epoch": 0.9790591504565304,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004912890894867357,
      "loss": 5.4892,
      "mean_token_accuracy": 0.1451600819826126,
      "num_tokens": 21613422.0,
      "step": 9865
    },
    {
      "entropy": 5.60218997001648,
      "epoch": 0.9795553791186979,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004912793234216293,
      "loss": 5.4723,
      "mean_token_accuracy": 0.1434368908405304,
      "num_tokens": 21624241.0,
      "step": 9870
    },
    {
      "entropy": 5.576246166229248,
      "epoch": 0.9800516077808654,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004912695519932618,
      "loss": 5.4731,
      "mean_token_accuracy": 0.14662786722183227,
      "num_tokens": 21636243.0,
      "step": 9875
    },
    {
      "entropy": 5.635243892669678,
      "epoch": 0.9805478364430329,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004912597752018753,
      "loss": 5.5095,
      "mean_token_accuracy": 0.14378939121961593,
      "num_tokens": 21647918.0,
      "step": 9880
    },
    {
      "entropy": 5.663714075088501,
      "epoch": 0.9810440651052005,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004912499930477126,
      "loss": 5.4868,
      "mean_token_accuracy": 0.14592071175575255,
      "num_tokens": 21659236.0,
      "step": 9885
    },
    {
      "entropy": 5.603769731521607,
      "epoch": 0.981540293767368,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004912402055310158,
      "loss": 5.4583,
      "mean_token_accuracy": 0.150116203725338,
      "num_tokens": 21669647.0,
      "step": 9890
    },
    {
      "entropy": 5.563555717468262,
      "epoch": 0.9820365224295355,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004912304126520279,
      "loss": 5.5222,
      "mean_token_accuracy": 0.14203586801886559,
      "num_tokens": 21680369.0,
      "step": 9895
    },
    {
      "entropy": 5.683042287826538,
      "epoch": 0.982532751091703,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004912206144109917,
      "loss": 5.5281,
      "mean_token_accuracy": 0.14423670917749404,
      "num_tokens": 21690942.0,
      "step": 9900
    },
    {
      "entropy": 5.681362342834473,
      "epoch": 0.9830289797538706,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004912108108081503,
      "loss": 5.4512,
      "mean_token_accuracy": 0.14615751057863235,
      "num_tokens": 21701945.0,
      "step": 9905
    },
    {
      "entropy": 5.620236158370972,
      "epoch": 0.9835252084160381,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004912010018437465,
      "loss": 5.486,
      "mean_token_accuracy": 0.1471480518579483,
      "num_tokens": 21712246.0,
      "step": 9910
    },
    {
      "entropy": 5.659247636795044,
      "epoch": 0.9840214370782057,
      "grad_norm": 1.3125,
      "learning_rate": 0.000491191187518024,
      "loss": 5.496,
      "mean_token_accuracy": 0.14491796046495437,
      "num_tokens": 21721660.0,
      "step": 9915
    },
    {
      "entropy": 5.561011075973511,
      "epoch": 0.9845176657403731,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004911813678312257,
      "loss": 5.3657,
      "mean_token_accuracy": 0.14719849079847336,
      "num_tokens": 21732157.0,
      "step": 9920
    },
    {
      "entropy": 5.612678718566895,
      "epoch": 0.9850138944025407,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004911715427835953,
      "loss": 5.5023,
      "mean_token_accuracy": 0.14685902297496795,
      "num_tokens": 21741727.0,
      "step": 9925
    },
    {
      "entropy": 5.642323541641235,
      "epoch": 0.9855101230647082,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004911617123753765,
      "loss": 5.4597,
      "mean_token_accuracy": 0.14912796393036842,
      "num_tokens": 21752596.0,
      "step": 9930
    },
    {
      "entropy": 5.598133373260498,
      "epoch": 0.9860063517268758,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004911518766068131,
      "loss": 5.3547,
      "mean_token_accuracy": 0.15120711028575898,
      "num_tokens": 21763978.0,
      "step": 9935
    },
    {
      "entropy": 5.6020303726196286,
      "epoch": 0.9865025803890433,
      "grad_norm": 4.84375,
      "learning_rate": 0.0004911420354781487,
      "loss": 5.4994,
      "mean_token_accuracy": 0.13747098445892333,
      "num_tokens": 21774961.0,
      "step": 9940
    },
    {
      "entropy": 5.613154697418213,
      "epoch": 0.9869988090512108,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004911321889896278,
      "loss": 5.4121,
      "mean_token_accuracy": 0.1527455672621727,
      "num_tokens": 21785387.0,
      "step": 9945
    },
    {
      "entropy": 5.583154678344727,
      "epoch": 0.9874950377133783,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004911223371414943,
      "loss": 5.4714,
      "mean_token_accuracy": 0.1485290363430977,
      "num_tokens": 21795982.0,
      "step": 9950
    },
    {
      "entropy": 5.628645992279052,
      "epoch": 0.9879912663755459,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004911124799339924,
      "loss": 5.4769,
      "mean_token_accuracy": 0.14477028250694274,
      "num_tokens": 21806768.0,
      "step": 9955
    },
    {
      "entropy": 5.610444784164429,
      "epoch": 0.9884874950377134,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004911026173673667,
      "loss": 5.4681,
      "mean_token_accuracy": 0.1461367890238762,
      "num_tokens": 21818563.0,
      "step": 9960
    },
    {
      "entropy": 5.659408664703369,
      "epoch": 0.9889837236998809,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004910927494418619,
      "loss": 5.5229,
      "mean_token_accuracy": 0.14659809470176696,
      "num_tokens": 21830253.0,
      "step": 9965
    },
    {
      "entropy": 5.680997705459594,
      "epoch": 0.9894799523620484,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004910828761577224,
      "loss": 5.5378,
      "mean_token_accuracy": 0.14875530898571016,
      "num_tokens": 21842353.0,
      "step": 9970
    },
    {
      "entropy": 5.6365868091583256,
      "epoch": 0.9899761810242159,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004910729975151933,
      "loss": 5.4273,
      "mean_token_accuracy": 0.1521955206990242,
      "num_tokens": 21852066.0,
      "step": 9975
    },
    {
      "entropy": 5.544122934341431,
      "epoch": 0.9904724096863835,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004910631135145194,
      "loss": 5.4941,
      "mean_token_accuracy": 0.14993072897195817,
      "num_tokens": 21862754.0,
      "step": 9980
    },
    {
      "entropy": 5.514778423309326,
      "epoch": 0.990968638348551,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004910532241559458,
      "loss": 5.3948,
      "mean_token_accuracy": 0.15903548896312714,
      "num_tokens": 21873531.0,
      "step": 9985
    },
    {
      "entropy": 5.4950783252716064,
      "epoch": 0.9914648670107186,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004910433294397179,
      "loss": 5.3236,
      "mean_token_accuracy": 0.15211291164159774,
      "num_tokens": 21884339.0,
      "step": 9990
    },
    {
      "entropy": 5.568712329864502,
      "epoch": 0.991961095672886,
      "grad_norm": 1.8359375,
      "learning_rate": 0.000491033429366081,
      "loss": 5.4477,
      "mean_token_accuracy": 0.14405056238174438,
      "num_tokens": 21895489.0,
      "step": 9995
    },
    {
      "entropy": 5.612151765823365,
      "epoch": 0.9924573243350536,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004910235239352804,
      "loss": 5.4091,
      "mean_token_accuracy": 0.15082173496484758,
      "num_tokens": 21906832.0,
      "step": 10000
    },
    {
      "entropy": 5.5907598495483395,
      "epoch": 0.9929535529972211,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000491013613147562,
      "loss": 5.4625,
      "mean_token_accuracy": 0.14950812011957168,
      "num_tokens": 21916979.0,
      "step": 10005
    },
    {
      "entropy": 5.61099648475647,
      "epoch": 0.9934497816593887,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004910036970031715,
      "loss": 5.3738,
      "mean_token_accuracy": 0.15420809984207154,
      "num_tokens": 21927352.0,
      "step": 10010
    },
    {
      "entropy": 5.6398192882537845,
      "epoch": 0.9939460103215562,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004909937755023547,
      "loss": 5.498,
      "mean_token_accuracy": 0.14480521380901337,
      "num_tokens": 21937765.0,
      "step": 10015
    },
    {
      "entropy": 5.548859930038452,
      "epoch": 0.9944422389837237,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004909838486453577,
      "loss": 5.3895,
      "mean_token_accuracy": 0.15133388638496398,
      "num_tokens": 21948659.0,
      "step": 10020
    },
    {
      "entropy": 5.5693504333496096,
      "epoch": 0.9949384676458912,
      "grad_norm": 1.75,
      "learning_rate": 0.0004909739164324267,
      "loss": 5.4944,
      "mean_token_accuracy": 0.14375930577516555,
      "num_tokens": 21959356.0,
      "step": 10025
    },
    {
      "entropy": 5.779584264755249,
      "epoch": 0.9954346963080588,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004909639788638079,
      "loss": 5.6128,
      "mean_token_accuracy": 0.13497235625982285,
      "num_tokens": 21971321.0,
      "step": 10030
    },
    {
      "entropy": 5.638870000839233,
      "epoch": 0.9959309249702263,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004909540359397477,
      "loss": 5.4735,
      "mean_token_accuracy": 0.13892346769571304,
      "num_tokens": 21981252.0,
      "step": 10035
    },
    {
      "entropy": 5.623515701293945,
      "epoch": 0.9964271536323938,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004909440876604929,
      "loss": 5.4795,
      "mean_token_accuracy": 0.14212223440408706,
      "num_tokens": 21991820.0,
      "step": 10040
    },
    {
      "entropy": 5.596395778656006,
      "epoch": 0.9969233822945613,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00049093413402629,
      "loss": 5.4306,
      "mean_token_accuracy": 0.15245020687580108,
      "num_tokens": 22003619.0,
      "step": 10045
    },
    {
      "entropy": 5.62937297821045,
      "epoch": 0.9974196109567288,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004909241750373858,
      "loss": 5.4986,
      "mean_token_accuracy": 0.13931187242269516,
      "num_tokens": 22016116.0,
      "step": 10050
    },
    {
      "entropy": 5.6824909210205075,
      "epoch": 0.9979158396188964,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004909142106940275,
      "loss": 5.4176,
      "mean_token_accuracy": 0.14956295788288115,
      "num_tokens": 22026115.0,
      "step": 10055
    },
    {
      "entropy": 5.652971982955933,
      "epoch": 0.9984120682810639,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000490904240996462,
      "loss": 5.4954,
      "mean_token_accuracy": 0.15159289687871932,
      "num_tokens": 22038889.0,
      "step": 10060
    },
    {
      "entropy": 5.537805795669556,
      "epoch": 0.9989082969432315,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004908942659449364,
      "loss": 5.4501,
      "mean_token_accuracy": 0.14977432787418365,
      "num_tokens": 22050773.0,
      "step": 10065
    },
    {
      "entropy": 5.61877670288086,
      "epoch": 0.9994045256053989,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004908842855396982,
      "loss": 5.4489,
      "mean_token_accuracy": 0.14058464169502258,
      "num_tokens": 22062869.0,
      "step": 10070
    },
    {
      "entropy": 5.647386646270752,
      "epoch": 0.9999007542675665,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004908742997809951,
      "loss": 5.4774,
      "mean_token_accuracy": 0.1462644025683403,
      "num_tokens": 22073898.0,
      "step": 10075
    },
    {
      "entropy": 5.650941514968872,
      "epoch": 1.0003969829297341,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004908643086690744,
      "loss": 5.3676,
      "mean_token_accuracy": 0.147015330940485,
      "num_tokens": 22084383.0,
      "step": 10080
    },
    {
      "entropy": 5.622048616409302,
      "epoch": 1.0008932115919016,
      "grad_norm": 1.78125,
      "learning_rate": 0.000490854312204184,
      "loss": 5.4314,
      "mean_token_accuracy": 0.14675159454345704,
      "num_tokens": 22094990.0,
      "step": 10085
    },
    {
      "entropy": 5.59576826095581,
      "epoch": 1.001389440254069,
      "grad_norm": 1.625,
      "learning_rate": 0.0004908443103865719,
      "loss": 5.4411,
      "mean_token_accuracy": 0.14362398758530617,
      "num_tokens": 22105259.0,
      "step": 10090
    },
    {
      "entropy": 5.594568300247192,
      "epoch": 1.0018856689162365,
      "grad_norm": 1.375,
      "learning_rate": 0.0004908343032164859,
      "loss": 5.4128,
      "mean_token_accuracy": 0.14742902517318726,
      "num_tokens": 22116348.0,
      "step": 10095
    },
    {
      "entropy": 5.626964902877807,
      "epoch": 1.0023818975784042,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004908242906941743,
      "loss": 5.4403,
      "mean_token_accuracy": 0.1497355319559574,
      "num_tokens": 22127311.0,
      "step": 10100
    },
    {
      "entropy": 5.662290906906128,
      "epoch": 1.0028781262405717,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004908142728198853,
      "loss": 5.4368,
      "mean_token_accuracy": 0.1441744588315487,
      "num_tokens": 22137929.0,
      "step": 10105
    },
    {
      "entropy": 5.697361755371094,
      "epoch": 1.0033743549027392,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004908042495938673,
      "loss": 5.5344,
      "mean_token_accuracy": 0.14043795987963675,
      "num_tokens": 22148202.0,
      "step": 10110
    },
    {
      "entropy": 5.557501029968262,
      "epoch": 1.0038705835649067,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004907942210163692,
      "loss": 5.3285,
      "mean_token_accuracy": 0.14917605817317964,
      "num_tokens": 22158955.0,
      "step": 10115
    },
    {
      "entropy": 5.568927049636841,
      "epoch": 1.0043668122270741,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004907841870876392,
      "loss": 5.3533,
      "mean_token_accuracy": 0.1470411777496338,
      "num_tokens": 22169569.0,
      "step": 10120
    },
    {
      "entropy": 5.5751824378967285,
      "epoch": 1.0048630408892418,
      "grad_norm": 1.5,
      "learning_rate": 0.0004907741478079263,
      "loss": 5.337,
      "mean_token_accuracy": 0.14379023760557175,
      "num_tokens": 22180171.0,
      "step": 10125
    },
    {
      "entropy": 5.623293352127075,
      "epoch": 1.0053592695514093,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004907641031774796,
      "loss": 5.4617,
      "mean_token_accuracy": 0.14716817140579225,
      "num_tokens": 22190531.0,
      "step": 10130
    },
    {
      "entropy": 5.602275609970093,
      "epoch": 1.0058554982135768,
      "grad_norm": 1.390625,
      "learning_rate": 0.000490754053196548,
      "loss": 5.3432,
      "mean_token_accuracy": 0.1535037025809288,
      "num_tokens": 22201495.0,
      "step": 10135
    },
    {
      "entropy": 5.666479682922363,
      "epoch": 1.0063517268757443,
      "grad_norm": 1.5,
      "learning_rate": 0.0004907439978653809,
      "loss": 5.4217,
      "mean_token_accuracy": 0.14277777224779128,
      "num_tokens": 22212941.0,
      "step": 10140
    },
    {
      "entropy": 5.662691402435303,
      "epoch": 1.006847955537912,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004907339371842275,
      "loss": 5.4643,
      "mean_token_accuracy": 0.14662302583456038,
      "num_tokens": 22225212.0,
      "step": 10145
    },
    {
      "entropy": 5.585462760925293,
      "epoch": 1.0073441842000794,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004907238711533373,
      "loss": 5.4142,
      "mean_token_accuracy": 0.1464483305811882,
      "num_tokens": 22236370.0,
      "step": 10150
    },
    {
      "entropy": 5.590307950973511,
      "epoch": 1.007840412862247,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004907137997729599,
      "loss": 5.397,
      "mean_token_accuracy": 0.14983568489551544,
      "num_tokens": 22247641.0,
      "step": 10155
    },
    {
      "entropy": 5.584212398529052,
      "epoch": 1.0083366415244144,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000490703723043345,
      "loss": 5.4477,
      "mean_token_accuracy": 0.14765015691518785,
      "num_tokens": 22258599.0,
      "step": 10160
    },
    {
      "entropy": 5.593160247802734,
      "epoch": 1.008832870186582,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004906936409647428,
      "loss": 5.3714,
      "mean_token_accuracy": 0.1448495529592037,
      "num_tokens": 22270913.0,
      "step": 10165
    },
    {
      "entropy": 5.546239995956421,
      "epoch": 1.0093290988487496,
      "grad_norm": 1.421875,
      "learning_rate": 0.000490683553537403,
      "loss": 5.2793,
      "mean_token_accuracy": 0.15433021634817123,
      "num_tokens": 22281053.0,
      "step": 10170
    },
    {
      "entropy": 5.5640322208404545,
      "epoch": 1.009825327510917,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004906734607615758,
      "loss": 5.3025,
      "mean_token_accuracy": 0.15013012886047364,
      "num_tokens": 22292570.0,
      "step": 10175
    },
    {
      "entropy": 5.571701288223267,
      "epoch": 1.0103215561730845,
      "grad_norm": 2.5,
      "learning_rate": 0.0004906633626375114,
      "loss": 5.376,
      "mean_token_accuracy": 0.1442698732018471,
      "num_tokens": 22303548.0,
      "step": 10180
    },
    {
      "entropy": 5.585268878936768,
      "epoch": 1.010817784835252,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004906532591654604,
      "loss": 5.4148,
      "mean_token_accuracy": 0.15121050477027892,
      "num_tokens": 22314133.0,
      "step": 10185
    },
    {
      "entropy": 5.619942092895508,
      "epoch": 1.0113140134974197,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004906431503456733,
      "loss": 5.3925,
      "mean_token_accuracy": 0.14394719451665877,
      "num_tokens": 22326431.0,
      "step": 10190
    },
    {
      "entropy": 5.575724315643311,
      "epoch": 1.0118102421595871,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004906330361784007,
      "loss": 5.411,
      "mean_token_accuracy": 0.14873455464839935,
      "num_tokens": 22338507.0,
      "step": 10195
    },
    {
      "entropy": 5.613606929779053,
      "epoch": 1.0123064708217546,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004906229166638936,
      "loss": 5.3958,
      "mean_token_accuracy": 0.14828364849090575,
      "num_tokens": 22348863.0,
      "step": 10200
    },
    {
      "entropy": 5.548977184295654,
      "epoch": 1.012802699483922,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004906127918024026,
      "loss": 5.3309,
      "mean_token_accuracy": 0.15180794298648834,
      "num_tokens": 22359000.0,
      "step": 10205
    },
    {
      "entropy": 5.65914134979248,
      "epoch": 1.0132989281460898,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004906026615941791,
      "loss": 5.4779,
      "mean_token_accuracy": 0.14160289168357848,
      "num_tokens": 22369539.0,
      "step": 10210
    },
    {
      "entropy": 5.630538845062256,
      "epoch": 1.0137951568082573,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004905925260394741,
      "loss": 5.429,
      "mean_token_accuracy": 0.1468749985098839,
      "num_tokens": 22380700.0,
      "step": 10215
    },
    {
      "entropy": 5.593900108337403,
      "epoch": 1.0142913854704247,
      "grad_norm": 1.3125,
      "learning_rate": 0.000490582385138539,
      "loss": 5.4102,
      "mean_token_accuracy": 0.14209697097539903,
      "num_tokens": 22392151.0,
      "step": 10220
    },
    {
      "entropy": 5.594948101043701,
      "epoch": 1.0147876141325922,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004905722388916254,
      "loss": 5.3907,
      "mean_token_accuracy": 0.151438869535923,
      "num_tokens": 22401735.0,
      "step": 10225
    },
    {
      "entropy": 5.615007448196411,
      "epoch": 1.01528384279476,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004905620872989846,
      "loss": 5.4655,
      "mean_token_accuracy": 0.14919091165065765,
      "num_tokens": 22412549.0,
      "step": 10230
    },
    {
      "entropy": 5.66500678062439,
      "epoch": 1.0157800714569274,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004905519303608687,
      "loss": 5.4432,
      "mean_token_accuracy": 0.1483164981007576,
      "num_tokens": 22424728.0,
      "step": 10235
    },
    {
      "entropy": 5.563656616210937,
      "epoch": 1.0162763001190949,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004905417680775294,
      "loss": 5.2982,
      "mean_token_accuracy": 0.152158984541893,
      "num_tokens": 22434197.0,
      "step": 10240
    },
    {
      "entropy": 5.644486570358277,
      "epoch": 1.0167725287812623,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004905316004492186,
      "loss": 5.5004,
      "mean_token_accuracy": 0.14115079641342163,
      "num_tokens": 22444671.0,
      "step": 10245
    },
    {
      "entropy": 5.5864707946777346,
      "epoch": 1.01726875744343,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004905214274761885,
      "loss": 5.3779,
      "mean_token_accuracy": 0.14497001543641092,
      "num_tokens": 22456636.0,
      "step": 10250
    },
    {
      "entropy": 5.553546714782715,
      "epoch": 1.0177649861055975,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004905112491586915,
      "loss": 5.308,
      "mean_token_accuracy": 0.15372665971517563,
      "num_tokens": 22467011.0,
      "step": 10255
    },
    {
      "entropy": 5.531224632263184,
      "epoch": 1.018261214767765,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00049050106549698,
      "loss": 5.4032,
      "mean_token_accuracy": 0.14761361479759216,
      "num_tokens": 22477630.0,
      "step": 10260
    },
    {
      "entropy": 5.529613590240478,
      "epoch": 1.0187574434299325,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004904908764913063,
      "loss": 5.349,
      "mean_token_accuracy": 0.15434909909963607,
      "num_tokens": 22488132.0,
      "step": 10265
    },
    {
      "entropy": 5.555251884460449,
      "epoch": 1.0192536720921,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004904806821419233,
      "loss": 5.4272,
      "mean_token_accuracy": 0.14391227215528488,
      "num_tokens": 22499430.0,
      "step": 10270
    },
    {
      "entropy": 5.557052803039551,
      "epoch": 1.0197499007542676,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004904704824490835,
      "loss": 5.3337,
      "mean_token_accuracy": 0.15297246277332305,
      "num_tokens": 22509728.0,
      "step": 10275
    },
    {
      "entropy": 5.611488914489746,
      "epoch": 1.020246129416435,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004904602774130402,
      "loss": 5.4415,
      "mean_token_accuracy": 0.1439923509955406,
      "num_tokens": 22520955.0,
      "step": 10280
    },
    {
      "entropy": 5.679212188720703,
      "epoch": 1.0207423580786026,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004904500670340463,
      "loss": 5.4681,
      "mean_token_accuracy": 0.14176828786730766,
      "num_tokens": 22530440.0,
      "step": 10285
    },
    {
      "entropy": 5.595128917694092,
      "epoch": 1.02123858674077,
      "grad_norm": 1.203125,
      "learning_rate": 0.000490439851312355,
      "loss": 5.4346,
      "mean_token_accuracy": 0.15255299359560012,
      "num_tokens": 22541539.0,
      "step": 10290
    },
    {
      "entropy": 5.556751346588134,
      "epoch": 1.0217348154029378,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004904296302482196,
      "loss": 5.4218,
      "mean_token_accuracy": 0.1473092779517174,
      "num_tokens": 22555717.0,
      "step": 10295
    },
    {
      "entropy": 5.668336582183838,
      "epoch": 1.0222310440651052,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004904194038418935,
      "loss": 5.5024,
      "mean_token_accuracy": 0.14512292593717574,
      "num_tokens": 22566938.0,
      "step": 10300
    },
    {
      "entropy": 5.67635817527771,
      "epoch": 1.0227272727272727,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004904091720936305,
      "loss": 5.3815,
      "mean_token_accuracy": 0.1459490731358528,
      "num_tokens": 22578725.0,
      "step": 10305
    },
    {
      "entropy": 5.5460607528686525,
      "epoch": 1.0232235013894402,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004903989350036841,
      "loss": 5.3195,
      "mean_token_accuracy": 0.15639133453369142,
      "num_tokens": 22588476.0,
      "step": 10310
    },
    {
      "entropy": 5.623313951492309,
      "epoch": 1.0237197300516079,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004903886925723083,
      "loss": 5.5133,
      "mean_token_accuracy": 0.1374771013855934,
      "num_tokens": 22600294.0,
      "step": 10315
    },
    {
      "entropy": 5.600350475311279,
      "epoch": 1.0242159587137754,
      "grad_norm": 1.15625,
      "learning_rate": 0.000490378444799757,
      "loss": 5.3916,
      "mean_token_accuracy": 0.14845612347126008,
      "num_tokens": 22610578.0,
      "step": 10320
    },
    {
      "entropy": 5.611328649520874,
      "epoch": 1.0247121873759428,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004903681916862844,
      "loss": 5.4111,
      "mean_token_accuracy": 0.1456850215792656,
      "num_tokens": 22620988.0,
      "step": 10325
    },
    {
      "entropy": 5.5753821849823,
      "epoch": 1.0252084160381103,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004903579332321447,
      "loss": 5.3993,
      "mean_token_accuracy": 0.148382967710495,
      "num_tokens": 22632198.0,
      "step": 10330
    },
    {
      "entropy": 5.598408126831055,
      "epoch": 1.0257046447002778,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004903476694375924,
      "loss": 5.3905,
      "mean_token_accuracy": 0.1537804886698723,
      "num_tokens": 22642543.0,
      "step": 10335
    },
    {
      "entropy": 5.635063791275025,
      "epoch": 1.0262008733624455,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004903374003028819,
      "loss": 5.4037,
      "mean_token_accuracy": 0.14648714661598206,
      "num_tokens": 22655453.0,
      "step": 10340
    },
    {
      "entropy": 5.567618179321289,
      "epoch": 1.026697102024613,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004903271258282678,
      "loss": 5.3534,
      "mean_token_accuracy": 0.15555230379104615,
      "num_tokens": 22667342.0,
      "step": 10345
    },
    {
      "entropy": 5.6181230545043945,
      "epoch": 1.0271933306867804,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004903168460140049,
      "loss": 5.4349,
      "mean_token_accuracy": 0.15233862847089769,
      "num_tokens": 22676890.0,
      "step": 10350
    },
    {
      "entropy": 5.5478263854980465,
      "epoch": 1.027689559348948,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004903065608603484,
      "loss": 5.417,
      "mean_token_accuracy": 0.15019289404153824,
      "num_tokens": 22687809.0,
      "step": 10355
    },
    {
      "entropy": 5.584484195709228,
      "epoch": 1.0281857880111156,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000490296270367553,
      "loss": 5.422,
      "mean_token_accuracy": 0.14731218665838242,
      "num_tokens": 22697938.0,
      "step": 10360
    },
    {
      "entropy": 5.658752298355102,
      "epoch": 1.028682016673283,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004902859745358741,
      "loss": 5.4216,
      "mean_token_accuracy": 0.15581315010786057,
      "num_tokens": 22708303.0,
      "step": 10365
    },
    {
      "entropy": 5.589664173126221,
      "epoch": 1.0291782453354505,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004902756733655668,
      "loss": 5.335,
      "mean_token_accuracy": 0.1559376209974289,
      "num_tokens": 22717987.0,
      "step": 10370
    },
    {
      "entropy": 5.581808137893677,
      "epoch": 1.029674473997618,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004902653668568868,
      "loss": 5.4146,
      "mean_token_accuracy": 0.15006666481494904,
      "num_tokens": 22729672.0,
      "step": 10375
    },
    {
      "entropy": 5.6060303211212155,
      "epoch": 1.0301707026597857,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004902550550100896,
      "loss": 5.4018,
      "mean_token_accuracy": 0.15362791419029237,
      "num_tokens": 22741296.0,
      "step": 10380
    },
    {
      "entropy": 5.577283239364624,
      "epoch": 1.0306669313219532,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004902447378254307,
      "loss": 5.4069,
      "mean_token_accuracy": 0.1490415155887604,
      "num_tokens": 22752337.0,
      "step": 10385
    },
    {
      "entropy": 5.5881676197052,
      "epoch": 1.0311631599841207,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004902344153031663,
      "loss": 5.4105,
      "mean_token_accuracy": 0.1446045942604542,
      "num_tokens": 22762681.0,
      "step": 10390
    },
    {
      "entropy": 5.648207092285157,
      "epoch": 1.0316593886462881,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000490224087443552,
      "loss": 5.4619,
      "mean_token_accuracy": 0.14346129670739174,
      "num_tokens": 22774275.0,
      "step": 10395
    },
    {
      "entropy": 5.587457323074341,
      "epoch": 1.0321556173084558,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004902137542468442,
      "loss": 5.3346,
      "mean_token_accuracy": 0.15190129727125168,
      "num_tokens": 22785944.0,
      "step": 10400
    },
    {
      "entropy": 5.484256792068481,
      "epoch": 1.0326518459706233,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004902034157132991,
      "loss": 5.2612,
      "mean_token_accuracy": 0.15184066742658614,
      "num_tokens": 22796543.0,
      "step": 10405
    },
    {
      "entropy": 5.521667575836181,
      "epoch": 1.0331480746327908,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004901930718431729,
      "loss": 5.3835,
      "mean_token_accuracy": 0.15052640438079834,
      "num_tokens": 22807957.0,
      "step": 10410
    },
    {
      "entropy": 5.674728584289551,
      "epoch": 1.0336443032949583,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004901827226367223,
      "loss": 5.4726,
      "mean_token_accuracy": 0.14678584933280944,
      "num_tokens": 22818903.0,
      "step": 10415
    },
    {
      "entropy": 5.650951671600342,
      "epoch": 1.0341405319571257,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004901723680942038,
      "loss": 5.4862,
      "mean_token_accuracy": 0.14566054940223694,
      "num_tokens": 22829286.0,
      "step": 10420
    },
    {
      "entropy": 5.650074005126953,
      "epoch": 1.0346367606192934,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004901620082158742,
      "loss": 5.436,
      "mean_token_accuracy": 0.1431647278368473,
      "num_tokens": 22839860.0,
      "step": 10425
    },
    {
      "entropy": 5.655006313323975,
      "epoch": 1.035132989281461,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004901516430019905,
      "loss": 5.4822,
      "mean_token_accuracy": 0.13847592622041702,
      "num_tokens": 22852238.0,
      "step": 10430
    },
    {
      "entropy": 5.60273232460022,
      "epoch": 1.0356292179436284,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004901412724528097,
      "loss": 5.4168,
      "mean_token_accuracy": 0.14669334888458252,
      "num_tokens": 22863013.0,
      "step": 10435
    },
    {
      "entropy": 5.554064416885376,
      "epoch": 1.0361254466057959,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004901308965685889,
      "loss": 5.3119,
      "mean_token_accuracy": 0.15089523494243623,
      "num_tokens": 22874233.0,
      "step": 10440
    },
    {
      "entropy": 5.6020301342010494,
      "epoch": 1.0366216752679636,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004901205153495853,
      "loss": 5.4694,
      "mean_token_accuracy": 0.147039183229208,
      "num_tokens": 22885313.0,
      "step": 10445
    },
    {
      "entropy": 5.5912376880645756,
      "epoch": 1.037117903930131,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004901101287960566,
      "loss": 5.3551,
      "mean_token_accuracy": 0.1534878969192505,
      "num_tokens": 22895365.0,
      "step": 10450
    },
    {
      "entropy": 5.539560604095459,
      "epoch": 1.0376141325922985,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004900997369082602,
      "loss": 5.3405,
      "mean_token_accuracy": 0.14811833202838898,
      "num_tokens": 22906392.0,
      "step": 10455
    },
    {
      "entropy": 5.527565908432007,
      "epoch": 1.038110361254466,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004900893396864539,
      "loss": 5.3297,
      "mean_token_accuracy": 0.1574187308549881,
      "num_tokens": 22916760.0,
      "step": 10460
    },
    {
      "entropy": 5.599536275863647,
      "epoch": 1.0386065899166337,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004900789371308953,
      "loss": 5.3851,
      "mean_token_accuracy": 0.14612501859664917,
      "num_tokens": 22928070.0,
      "step": 10465
    },
    {
      "entropy": 5.5075523853302,
      "epoch": 1.0391028185788012,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004900685292418425,
      "loss": 5.3491,
      "mean_token_accuracy": 0.15478256195783616,
      "num_tokens": 22938815.0,
      "step": 10470
    },
    {
      "entropy": 5.565003824234009,
      "epoch": 1.0395990472409686,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004900581160195537,
      "loss": 5.4244,
      "mean_token_accuracy": 0.14336036145687103,
      "num_tokens": 22949806.0,
      "step": 10475
    },
    {
      "entropy": 5.530713653564453,
      "epoch": 1.040095275903136,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000490047697464287,
      "loss": 5.2752,
      "mean_token_accuracy": 0.1558382824063301,
      "num_tokens": 22960880.0,
      "step": 10480
    },
    {
      "entropy": 5.591424131393433,
      "epoch": 1.0405915045653038,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004900372735763009,
      "loss": 5.4269,
      "mean_token_accuracy": 0.14368789196014403,
      "num_tokens": 22972828.0,
      "step": 10485
    },
    {
      "entropy": 5.480336284637451,
      "epoch": 1.0410877332274713,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004900268443558536,
      "loss": 5.3062,
      "mean_token_accuracy": 0.1525890201330185,
      "num_tokens": 22981869.0,
      "step": 10490
    },
    {
      "entropy": 5.633327960968018,
      "epoch": 1.0415839618896388,
      "grad_norm": 1.296875,
      "learning_rate": 0.000490016409803204,
      "loss": 5.4299,
      "mean_token_accuracy": 0.14455235078930856,
      "num_tokens": 22993111.0,
      "step": 10495
    },
    {
      "entropy": 5.613408374786377,
      "epoch": 1.0420801905518062,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004900059699186108,
      "loss": 5.4131,
      "mean_token_accuracy": 0.1455287143588066,
      "num_tokens": 23003115.0,
      "step": 10500
    },
    {
      "entropy": 5.616667127609253,
      "epoch": 1.0425764192139737,
      "grad_norm": 1.375,
      "learning_rate": 0.0004899955247023328,
      "loss": 5.4327,
      "mean_token_accuracy": 0.14665038958191873,
      "num_tokens": 23012587.0,
      "step": 10505
    },
    {
      "entropy": 5.5872313499450685,
      "epoch": 1.0430726478761414,
      "grad_norm": 1.328125,
      "learning_rate": 0.000489985074154629,
      "loss": 5.4736,
      "mean_token_accuracy": 0.14407994598150253,
      "num_tokens": 23024455.0,
      "step": 10510
    },
    {
      "entropy": 5.589589786529541,
      "epoch": 1.0435688765383089,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004899746182757587,
      "loss": 5.3818,
      "mean_token_accuracy": 0.14735177233815194,
      "num_tokens": 23034430.0,
      "step": 10515
    },
    {
      "entropy": 5.653596591949463,
      "epoch": 1.0440651052004764,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004899641570659811,
      "loss": 5.4502,
      "mean_token_accuracy": 0.14837532341480256,
      "num_tokens": 23044811.0,
      "step": 10520
    },
    {
      "entropy": 5.624305963516235,
      "epoch": 1.0445613338626438,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004899536905255555,
      "loss": 5.4839,
      "mean_token_accuracy": 0.14939549714326858,
      "num_tokens": 23056179.0,
      "step": 10525
    },
    {
      "entropy": 5.5613785743713375,
      "epoch": 1.0450575625248115,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004899432186547416,
      "loss": 5.4101,
      "mean_token_accuracy": 0.1484897680580616,
      "num_tokens": 23068331.0,
      "step": 10530
    },
    {
      "entropy": 5.58583550453186,
      "epoch": 1.045553791186979,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004899327414537991,
      "loss": 5.3901,
      "mean_token_accuracy": 0.14447612017393113,
      "num_tokens": 23079213.0,
      "step": 10535
    },
    {
      "entropy": 5.695864486694336,
      "epoch": 1.0460500198491465,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004899222589229877,
      "loss": 5.503,
      "mean_token_accuracy": 0.1394645892083645,
      "num_tokens": 23089860.0,
      "step": 10540
    },
    {
      "entropy": 5.55170316696167,
      "epoch": 1.046546248511314,
      "grad_norm": 1.875,
      "learning_rate": 0.0004899117710625674,
      "loss": 5.3717,
      "mean_token_accuracy": 0.1557852789759636,
      "num_tokens": 23099431.0,
      "step": 10545
    },
    {
      "entropy": 5.596436500549316,
      "epoch": 1.0470424771734816,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004899012778727983,
      "loss": 5.4584,
      "mean_token_accuracy": 0.14351726695895195,
      "num_tokens": 23110947.0,
      "step": 10550
    },
    {
      "entropy": 5.620785188674927,
      "epoch": 1.0475387058356491,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004898907793539406,
      "loss": 5.5167,
      "mean_token_accuracy": 0.14646417573094367,
      "num_tokens": 23122782.0,
      "step": 10555
    },
    {
      "entropy": 5.672925615310669,
      "epoch": 1.0480349344978166,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004898802755062546,
      "loss": 5.4324,
      "mean_token_accuracy": 0.14839882999658585,
      "num_tokens": 23134254.0,
      "step": 10560
    },
    {
      "entropy": 5.557016181945801,
      "epoch": 1.048531163159984,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004898697663300007,
      "loss": 5.2548,
      "mean_token_accuracy": 0.1571120262145996,
      "num_tokens": 23143851.0,
      "step": 10565
    },
    {
      "entropy": 5.579274654388428,
      "epoch": 1.0490273918221515,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004898592518254398,
      "loss": 5.3969,
      "mean_token_accuracy": 0.14896284341812133,
      "num_tokens": 23155966.0,
      "step": 10570
    },
    {
      "entropy": 5.558279657363892,
      "epoch": 1.0495236204843192,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004898487319928322,
      "loss": 5.396,
      "mean_token_accuracy": 0.14339442625641824,
      "num_tokens": 23165896.0,
      "step": 10575
    },
    {
      "entropy": 5.574773216247559,
      "epoch": 1.0500198491464867,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004898382068324392,
      "loss": 5.3838,
      "mean_token_accuracy": 0.14713851362466812,
      "num_tokens": 23177821.0,
      "step": 10580
    },
    {
      "entropy": 5.632523155212402,
      "epoch": 1.0505160778086542,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004898276763445215,
      "loss": 5.4441,
      "mean_token_accuracy": 0.14489770233631133,
      "num_tokens": 23188916.0,
      "step": 10585
    },
    {
      "entropy": 5.639192056655884,
      "epoch": 1.0510123064708217,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004898171405293404,
      "loss": 5.425,
      "mean_token_accuracy": 0.1547326222062111,
      "num_tokens": 23200490.0,
      "step": 10590
    },
    {
      "entropy": 5.513908720016479,
      "epoch": 1.0515085351329894,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004898065993871571,
      "loss": 5.3747,
      "mean_token_accuracy": 0.1476880818605423,
      "num_tokens": 23212991.0,
      "step": 10595
    },
    {
      "entropy": 5.618760442733764,
      "epoch": 1.0520047637951568,
      "grad_norm": 1.25,
      "learning_rate": 0.000489796052918233,
      "loss": 5.4825,
      "mean_token_accuracy": 0.1472497045993805,
      "num_tokens": 23223361.0,
      "step": 10600
    },
    {
      "entropy": 5.6412327766418455,
      "epoch": 1.0525009924573243,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004897855011228295,
      "loss": 5.3962,
      "mean_token_accuracy": 0.15113023519515992,
      "num_tokens": 23233765.0,
      "step": 10605
    },
    {
      "entropy": 5.530994987487793,
      "epoch": 1.0529972211194918,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004897749440012085,
      "loss": 5.339,
      "mean_token_accuracy": 0.15162474811077117,
      "num_tokens": 23246562.0,
      "step": 10610
    },
    {
      "entropy": 5.537393856048584,
      "epoch": 1.0534934497816595,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004897643815536316,
      "loss": 5.3527,
      "mean_token_accuracy": 0.15241698026657105,
      "num_tokens": 23256648.0,
      "step": 10615
    },
    {
      "entropy": 5.575681400299072,
      "epoch": 1.053989678443827,
      "grad_norm": 1.171875,
      "learning_rate": 0.000489753813780361,
      "loss": 5.4059,
      "mean_token_accuracy": 0.14553543776273728,
      "num_tokens": 23267428.0,
      "step": 10620
    },
    {
      "entropy": 5.5764764785766605,
      "epoch": 1.0544859071059944,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004897432406816585,
      "loss": 5.3351,
      "mean_token_accuracy": 0.14827754944562913,
      "num_tokens": 23278440.0,
      "step": 10625
    },
    {
      "entropy": 5.586294031143188,
      "epoch": 1.054982135768162,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004897326622577863,
      "loss": 5.401,
      "mean_token_accuracy": 0.1444564491510391,
      "num_tokens": 23290571.0,
      "step": 10630
    },
    {
      "entropy": 5.664337825775147,
      "epoch": 1.0554783644303294,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004897220785090067,
      "loss": 5.4719,
      "mean_token_accuracy": 0.1463848441839218,
      "num_tokens": 23301634.0,
      "step": 10635
    },
    {
      "entropy": 5.5513616561889645,
      "epoch": 1.055974593092497,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004897114894355824,
      "loss": 5.333,
      "mean_token_accuracy": 0.15088742822408677,
      "num_tokens": 23313125.0,
      "step": 10640
    },
    {
      "entropy": 5.520159006118774,
      "epoch": 1.0564708217546646,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004897008950377758,
      "loss": 5.3443,
      "mean_token_accuracy": 0.15069437474012376,
      "num_tokens": 23323610.0,
      "step": 10645
    },
    {
      "entropy": 5.509404993057251,
      "epoch": 1.056967050416832,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004896902953158497,
      "loss": 5.3398,
      "mean_token_accuracy": 0.1503956899046898,
      "num_tokens": 23333652.0,
      "step": 10650
    },
    {
      "entropy": 5.533861112594605,
      "epoch": 1.0574632790789995,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004896796902700667,
      "loss": 5.3791,
      "mean_token_accuracy": 0.15427400469779967,
      "num_tokens": 23345954.0,
      "step": 10655
    },
    {
      "entropy": 5.559717464447021,
      "epoch": 1.0579595077411672,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004896690799006901,
      "loss": 5.3635,
      "mean_token_accuracy": 0.14997858107089995,
      "num_tokens": 23357052.0,
      "step": 10660
    },
    {
      "entropy": 5.589969444274902,
      "epoch": 1.0584557364033347,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004896584642079828,
      "loss": 5.3727,
      "mean_token_accuracy": 0.15361550748348235,
      "num_tokens": 23368837.0,
      "step": 10665
    },
    {
      "entropy": 5.513894414901733,
      "epoch": 1.0589519650655022,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004896478431922083,
      "loss": 5.3471,
      "mean_token_accuracy": 0.15341353863477708,
      "num_tokens": 23378608.0,
      "step": 10670
    },
    {
      "entropy": 5.605801200866699,
      "epoch": 1.0594481937276696,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004896372168536297,
      "loss": 5.4333,
      "mean_token_accuracy": 0.15041675567626953,
      "num_tokens": 23390094.0,
      "step": 10675
    },
    {
      "entropy": 5.562920999526978,
      "epoch": 1.0599444223898373,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004896265851925107,
      "loss": 5.2906,
      "mean_token_accuracy": 0.1569736734032631,
      "num_tokens": 23399991.0,
      "step": 10680
    },
    {
      "entropy": 5.577964639663696,
      "epoch": 1.0604406510520048,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004896159482091148,
      "loss": 5.4859,
      "mean_token_accuracy": 0.14056169614195824,
      "num_tokens": 23410705.0,
      "step": 10685
    },
    {
      "entropy": 5.555141925811768,
      "epoch": 1.0609368797141723,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004896053059037058,
      "loss": 5.2767,
      "mean_token_accuracy": 0.16308921873569487,
      "num_tokens": 23421939.0,
      "step": 10690
    },
    {
      "entropy": 5.566424894332886,
      "epoch": 1.0614331083763398,
      "grad_norm": 1.375,
      "learning_rate": 0.0004895946582765477,
      "loss": 5.4118,
      "mean_token_accuracy": 0.1455625757575035,
      "num_tokens": 23433252.0,
      "step": 10695
    },
    {
      "entropy": 5.651719570159912,
      "epoch": 1.0619293370385074,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004895840053279045,
      "loss": 5.475,
      "mean_token_accuracy": 0.14326800853013993,
      "num_tokens": 23445582.0,
      "step": 10700
    },
    {
      "entropy": 5.637252616882324,
      "epoch": 1.062425565700675,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004895733470580404,
      "loss": 5.4431,
      "mean_token_accuracy": 0.14277275502681733,
      "num_tokens": 23456360.0,
      "step": 10705
    },
    {
      "entropy": 5.577129220962524,
      "epoch": 1.0629217943628424,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004895626834672196,
      "loss": 5.3253,
      "mean_token_accuracy": 0.15172981470823288,
      "num_tokens": 23466576.0,
      "step": 10710
    },
    {
      "entropy": 5.49441876411438,
      "epoch": 1.0634180230250099,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004895520145557066,
      "loss": 5.2776,
      "mean_token_accuracy": 0.15787552148103715,
      "num_tokens": 23476200.0,
      "step": 10715
    },
    {
      "entropy": 5.623633909225464,
      "epoch": 1.0639142516871773,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004895413403237659,
      "loss": 5.4755,
      "mean_token_accuracy": 0.1498984821140766,
      "num_tokens": 23487776.0,
      "step": 10720
    },
    {
      "entropy": 5.543395709991455,
      "epoch": 1.064410480349345,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004895306607716624,
      "loss": 5.3454,
      "mean_token_accuracy": 0.1497153490781784,
      "num_tokens": 23497517.0,
      "step": 10725
    },
    {
      "entropy": 5.553286695480347,
      "epoch": 1.0649067090115125,
      "grad_norm": 2.828125,
      "learning_rate": 0.0004895199758996607,
      "loss": 5.3554,
      "mean_token_accuracy": 0.15442114919424058,
      "num_tokens": 23507436.0,
      "step": 10730
    },
    {
      "entropy": 5.552494049072266,
      "epoch": 1.06540293767368,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004895092857080258,
      "loss": 5.3097,
      "mean_token_accuracy": 0.14342986270785332,
      "num_tokens": 23517170.0,
      "step": 10735
    },
    {
      "entropy": 5.532317781448365,
      "epoch": 1.0658991663358475,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004894985901970229,
      "loss": 5.3143,
      "mean_token_accuracy": 0.153928504884243,
      "num_tokens": 23527329.0,
      "step": 10740
    },
    {
      "entropy": 5.552791786193848,
      "epoch": 1.0663953949980152,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004894878893669171,
      "loss": 5.4207,
      "mean_token_accuracy": 0.14376751333475113,
      "num_tokens": 23538341.0,
      "step": 10745
    },
    {
      "entropy": 5.579165887832642,
      "epoch": 1.0668916236601826,
      "grad_norm": 1.125,
      "learning_rate": 0.0004894771832179739,
      "loss": 5.3448,
      "mean_token_accuracy": 0.1490940883755684,
      "num_tokens": 23549522.0,
      "step": 10750
    },
    {
      "entropy": 5.597357177734375,
      "epoch": 1.0673878523223501,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004894664717504587,
      "loss": 5.384,
      "mean_token_accuracy": 0.147237104177475,
      "num_tokens": 23559274.0,
      "step": 10755
    },
    {
      "entropy": 5.564702463150025,
      "epoch": 1.0678840809845176,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004894557549646371,
      "loss": 5.35,
      "mean_token_accuracy": 0.15608398765325546,
      "num_tokens": 23570852.0,
      "step": 10760
    },
    {
      "entropy": 5.549319696426392,
      "epoch": 1.0683803096466853,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004894450328607749,
      "loss": 5.3614,
      "mean_token_accuracy": 0.14895587265491486,
      "num_tokens": 23582918.0,
      "step": 10765
    },
    {
      "entropy": 5.54530553817749,
      "epoch": 1.0688765383088528,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004894343054391379,
      "loss": 5.3834,
      "mean_token_accuracy": 0.15435569137334823,
      "num_tokens": 23594177.0,
      "step": 10770
    },
    {
      "entropy": 5.549611139297485,
      "epoch": 1.0693727669710202,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004894235726999924,
      "loss": 5.3566,
      "mean_token_accuracy": 0.15240906327962875,
      "num_tokens": 23605901.0,
      "step": 10775
    },
    {
      "entropy": 5.548908758163452,
      "epoch": 1.0698689956331877,
      "grad_norm": 1.25,
      "learning_rate": 0.0004894128346436042,
      "loss": 5.3956,
      "mean_token_accuracy": 0.15026365518569945,
      "num_tokens": 23617354.0,
      "step": 10780
    },
    {
      "entropy": 5.597808885574341,
      "epoch": 1.0703652242953554,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004894020912702397,
      "loss": 5.4317,
      "mean_token_accuracy": 0.1465703323483467,
      "num_tokens": 23627603.0,
      "step": 10785
    },
    {
      "entropy": 5.497972774505615,
      "epoch": 1.0708614529575229,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004893913425801653,
      "loss": 5.2418,
      "mean_token_accuracy": 0.161360202729702,
      "num_tokens": 23637280.0,
      "step": 10790
    },
    {
      "entropy": 5.553600740432739,
      "epoch": 1.0713576816196904,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004893805885736477,
      "loss": 5.4445,
      "mean_token_accuracy": 0.14513869136571883,
      "num_tokens": 23649326.0,
      "step": 10795
    },
    {
      "entropy": 5.623505258560181,
      "epoch": 1.0718539102818578,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004893698292509534,
      "loss": 5.4588,
      "mean_token_accuracy": 0.1441636860370636,
      "num_tokens": 23659561.0,
      "step": 10800
    },
    {
      "entropy": 5.624791002273559,
      "epoch": 1.0723501389440253,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004893590646123492,
      "loss": 5.3693,
      "mean_token_accuracy": 0.1506016656756401,
      "num_tokens": 23673206.0,
      "step": 10805
    },
    {
      "entropy": 5.5633055686950685,
      "epoch": 1.072846367606193,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004893482946581022,
      "loss": 5.4047,
      "mean_token_accuracy": 0.1424211546778679,
      "num_tokens": 23683920.0,
      "step": 10810
    },
    {
      "entropy": 5.531470155715942,
      "epoch": 1.0733425962683605,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004893375193884793,
      "loss": 5.32,
      "mean_token_accuracy": 0.15341602712869645,
      "num_tokens": 23693990.0,
      "step": 10815
    },
    {
      "entropy": 5.602037811279297,
      "epoch": 1.073838824930528,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004893267388037478,
      "loss": 5.4292,
      "mean_token_accuracy": 0.14800168201327324,
      "num_tokens": 23704870.0,
      "step": 10820
    },
    {
      "entropy": 5.536337614059448,
      "epoch": 1.0743350535926954,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000489315952904175,
      "loss": 5.4043,
      "mean_token_accuracy": 0.15073905140161514,
      "num_tokens": 23715159.0,
      "step": 10825
    },
    {
      "entropy": 5.5771567821502686,
      "epoch": 1.0748312822548631,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004893051616900283,
      "loss": 5.3341,
      "mean_token_accuracy": 0.15016917288303375,
      "num_tokens": 23725252.0,
      "step": 10830
    },
    {
      "entropy": 5.553519201278687,
      "epoch": 1.0753275109170306,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004892943651615756,
      "loss": 5.3646,
      "mean_token_accuracy": 0.1509012058377266,
      "num_tokens": 23736472.0,
      "step": 10835
    },
    {
      "entropy": 5.568573045730591,
      "epoch": 1.075823739579198,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004892835633190843,
      "loss": 5.2901,
      "mean_token_accuracy": 0.14610609263181687,
      "num_tokens": 23747565.0,
      "step": 10840
    },
    {
      "entropy": 5.573373651504516,
      "epoch": 1.0763199682413656,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004892727561628223,
      "loss": 5.3693,
      "mean_token_accuracy": 0.1465209037065506,
      "num_tokens": 23758201.0,
      "step": 10845
    },
    {
      "entropy": 5.612726068496704,
      "epoch": 1.0768161969035333,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004892619436930577,
      "loss": 5.4227,
      "mean_token_accuracy": 0.1500130921602249,
      "num_tokens": 23769087.0,
      "step": 10850
    },
    {
      "entropy": 5.575374412536621,
      "epoch": 1.0773124255657007,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004892511259100585,
      "loss": 5.4187,
      "mean_token_accuracy": 0.14716205820441247,
      "num_tokens": 23780598.0,
      "step": 10855
    },
    {
      "entropy": 5.57796630859375,
      "epoch": 1.0778086542278682,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004892403028140932,
      "loss": 5.3596,
      "mean_token_accuracy": 0.14485704600811006,
      "num_tokens": 23791449.0,
      "step": 10860
    },
    {
      "entropy": 5.541803884506225,
      "epoch": 1.0783048828900357,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048922947440543,
      "loss": 5.3271,
      "mean_token_accuracy": 0.15452092587947847,
      "num_tokens": 23803071.0,
      "step": 10865
    },
    {
      "entropy": 5.596189880371094,
      "epoch": 1.0788011115522032,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004892186406843374,
      "loss": 5.4208,
      "mean_token_accuracy": 0.1426899902522564,
      "num_tokens": 23813372.0,
      "step": 10870
    },
    {
      "entropy": 5.59133882522583,
      "epoch": 1.0792973402143708,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004892078016510841,
      "loss": 5.408,
      "mean_token_accuracy": 0.15046855360269545,
      "num_tokens": 23824979.0,
      "step": 10875
    },
    {
      "entropy": 5.575678062438965,
      "epoch": 1.0797935688765383,
      "grad_norm": 1.40625,
      "learning_rate": 0.000489196957305939,
      "loss": 5.3771,
      "mean_token_accuracy": 0.14679014757275582,
      "num_tokens": 23836335.0,
      "step": 10880
    },
    {
      "entropy": 5.543770027160645,
      "epoch": 1.0802897975387058,
      "grad_norm": 2.890625,
      "learning_rate": 0.0004891861076491709,
      "loss": 5.3642,
      "mean_token_accuracy": 0.14835133999586106,
      "num_tokens": 23848147.0,
      "step": 10885
    },
    {
      "entropy": 5.480254316329956,
      "epoch": 1.0807860262008733,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004891752526810488,
      "loss": 5.3133,
      "mean_token_accuracy": 0.15273910760879517,
      "num_tokens": 23859502.0,
      "step": 10890
    },
    {
      "entropy": 5.597673606872559,
      "epoch": 1.081282254863041,
      "grad_norm": 1.28125,
      "learning_rate": 0.000489164392401842,
      "loss": 5.4575,
      "mean_token_accuracy": 0.14838099330663682,
      "num_tokens": 23870433.0,
      "step": 10895
    },
    {
      "entropy": 5.6021904945373535,
      "epoch": 1.0817784835252084,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004891535268118196,
      "loss": 5.2924,
      "mean_token_accuracy": 0.15119415819644927,
      "num_tokens": 23879581.0,
      "step": 10900
    },
    {
      "entropy": 5.48640456199646,
      "epoch": 1.082274712187376,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004891426559112514,
      "loss": 5.3469,
      "mean_token_accuracy": 0.15342123955488204,
      "num_tokens": 23890889.0,
      "step": 10905
    },
    {
      "entropy": 5.528471899032593,
      "epoch": 1.0827709408495434,
      "grad_norm": 1.375,
      "learning_rate": 0.0004891317797004066,
      "loss": 5.3374,
      "mean_token_accuracy": 0.15530709624290467,
      "num_tokens": 23901577.0,
      "step": 10910
    },
    {
      "entropy": 5.564786529541015,
      "epoch": 1.083267169511711,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004891208981795552,
      "loss": 5.415,
      "mean_token_accuracy": 0.1486767366528511,
      "num_tokens": 23912698.0,
      "step": 10915
    },
    {
      "entropy": 5.510092401504517,
      "epoch": 1.0837633981738786,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004891100113489669,
      "loss": 5.2854,
      "mean_token_accuracy": 0.15407822281122208,
      "num_tokens": 23923056.0,
      "step": 10920
    },
    {
      "entropy": 5.531494951248169,
      "epoch": 1.084259626836046,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004890991192089116,
      "loss": 5.4332,
      "mean_token_accuracy": 0.15010141283273698,
      "num_tokens": 23933998.0,
      "step": 10925
    },
    {
      "entropy": 5.565519142150879,
      "epoch": 1.0847558554982135,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004890882217596595,
      "loss": 5.312,
      "mean_token_accuracy": 0.15542884320020675,
      "num_tokens": 23943980.0,
      "step": 10930
    },
    {
      "entropy": 5.628781270980835,
      "epoch": 1.085252084160381,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004890773190014809,
      "loss": 5.4015,
      "mean_token_accuracy": 0.14728218764066697,
      "num_tokens": 23955646.0,
      "step": 10935
    },
    {
      "entropy": 5.545685338973999,
      "epoch": 1.0857483128225487,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000489066410934646,
      "loss": 5.3499,
      "mean_token_accuracy": 0.1557690232992172,
      "num_tokens": 23966381.0,
      "step": 10940
    },
    {
      "entropy": 5.603308534622192,
      "epoch": 1.0862445414847162,
      "grad_norm": 1.125,
      "learning_rate": 0.0004890554975594255,
      "loss": 5.4333,
      "mean_token_accuracy": 0.14232421517372132,
      "num_tokens": 23978669.0,
      "step": 10945
    },
    {
      "entropy": 5.683228397369385,
      "epoch": 1.0867407701468836,
      "grad_norm": 1.5,
      "learning_rate": 0.0004890445788760898,
      "loss": 5.5146,
      "mean_token_accuracy": 0.1384425513446331,
      "num_tokens": 23989467.0,
      "step": 10950
    },
    {
      "entropy": 5.570678043365478,
      "epoch": 1.0872369988090511,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004890336548849098,
      "loss": 5.3227,
      "mean_token_accuracy": 0.1506729707121849,
      "num_tokens": 24000797.0,
      "step": 10955
    },
    {
      "entropy": 5.5994288444519045,
      "epoch": 1.0877332274712188,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004890227255861564,
      "loss": 5.462,
      "mean_token_accuracy": 0.14822121411561967,
      "num_tokens": 24013182.0,
      "step": 10960
    },
    {
      "entropy": 5.5483075141906735,
      "epoch": 1.0882294561333863,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004890117909801005,
      "loss": 5.3575,
      "mean_token_accuracy": 0.1505046159029007,
      "num_tokens": 24024005.0,
      "step": 10965
    },
    {
      "entropy": 5.575407314300537,
      "epoch": 1.0887256847955538,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004890008510670134,
      "loss": 5.336,
      "mean_token_accuracy": 0.14957382827997207,
      "num_tokens": 24034524.0,
      "step": 10970
    },
    {
      "entropy": 5.561434364318847,
      "epoch": 1.0892219134577212,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004889899058471663,
      "loss": 5.3957,
      "mean_token_accuracy": 0.1522506982088089,
      "num_tokens": 24045944.0,
      "step": 10975
    },
    {
      "entropy": 5.653478813171387,
      "epoch": 1.089718142119889,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004889789553208306,
      "loss": 5.479,
      "mean_token_accuracy": 0.15033394396305083,
      "num_tokens": 24058612.0,
      "step": 10980
    },
    {
      "entropy": 5.655957746505737,
      "epoch": 1.0902143707820564,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000488967999488278,
      "loss": 5.3357,
      "mean_token_accuracy": 0.1527863159775734,
      "num_tokens": 24068912.0,
      "step": 10985
    },
    {
      "entropy": 5.47331428527832,
      "epoch": 1.0907105994442239,
      "grad_norm": 2.484375,
      "learning_rate": 0.00048895703834978,
      "loss": 5.3355,
      "mean_token_accuracy": 0.15577730685472488,
      "num_tokens": 24079879.0,
      "step": 10990
    },
    {
      "entropy": 5.5048990726470945,
      "epoch": 1.0912068281063914,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004889460719056085,
      "loss": 5.3727,
      "mean_token_accuracy": 0.14852165132761003,
      "num_tokens": 24091268.0,
      "step": 10995
    },
    {
      "entropy": 5.572988557815552,
      "epoch": 1.091703056768559,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004889351001560354,
      "loss": 5.3814,
      "mean_token_accuracy": 0.15017618834972382,
      "num_tokens": 24101428.0,
      "step": 11000
    },
    {
      "entropy": 5.532800340652466,
      "epoch": 1.0921992854307265,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004889241231013329,
      "loss": 5.3534,
      "mean_token_accuracy": 0.15488902926445008,
      "num_tokens": 24112920.0,
      "step": 11005
    },
    {
      "entropy": 5.485278987884522,
      "epoch": 1.092695514092894,
      "grad_norm": 1.5625,
      "learning_rate": 0.000488913140741773,
      "loss": 5.3169,
      "mean_token_accuracy": 0.15163280218839645,
      "num_tokens": 24125731.0,
      "step": 11010
    },
    {
      "entropy": 5.529014968872071,
      "epoch": 1.0931917427550615,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004889021530776282,
      "loss": 5.3053,
      "mean_token_accuracy": 0.1526373505592346,
      "num_tokens": 24136606.0,
      "step": 11015
    },
    {
      "entropy": 5.611603784561157,
      "epoch": 1.093687971417229,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004888911601091709,
      "loss": 5.3216,
      "mean_token_accuracy": 0.15624798089265823,
      "num_tokens": 24146783.0,
      "step": 11020
    },
    {
      "entropy": 5.594170141220093,
      "epoch": 1.0941842000793967,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004888801618366737,
      "loss": 5.4398,
      "mean_token_accuracy": 0.14871452748775482,
      "num_tokens": 24157412.0,
      "step": 11025
    },
    {
      "entropy": 5.556654596328736,
      "epoch": 1.0946804287415641,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004888691582604095,
      "loss": 5.4206,
      "mean_token_accuracy": 0.15288988649845123,
      "num_tokens": 24168077.0,
      "step": 11030
    },
    {
      "entropy": 5.571929979324341,
      "epoch": 1.0951766574037316,
      "grad_norm": 1.9765625,
      "learning_rate": 0.000488858149380651,
      "loss": 5.4528,
      "mean_token_accuracy": 0.14549228325486183,
      "num_tokens": 24178913.0,
      "step": 11035
    },
    {
      "entropy": 5.662427854537964,
      "epoch": 1.095672886065899,
      "grad_norm": 1.9375,
      "learning_rate": 0.000488847135197671,
      "loss": 5.4977,
      "mean_token_accuracy": 0.14151506647467613,
      "num_tokens": 24190356.0,
      "step": 11040
    },
    {
      "entropy": 5.64313817024231,
      "epoch": 1.0961691147280668,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000488836115711743,
      "loss": 5.4408,
      "mean_token_accuracy": 0.1450464531779289,
      "num_tokens": 24199772.0,
      "step": 11045
    },
    {
      "entropy": 5.613921880722046,
      "epoch": 1.0966653433902342,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004888250909231401,
      "loss": 5.3913,
      "mean_token_accuracy": 0.15150224044919014,
      "num_tokens": 24211807.0,
      "step": 11050
    },
    {
      "entropy": 5.592180871963501,
      "epoch": 1.0971615720524017,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004888140608321357,
      "loss": 5.3574,
      "mean_token_accuracy": 0.15196514874696732,
      "num_tokens": 24223156.0,
      "step": 11055
    },
    {
      "entropy": 5.6193183898925785,
      "epoch": 1.0976578007145692,
      "grad_norm": 1.9609375,
      "learning_rate": 0.0004888030254390033,
      "loss": 5.421,
      "mean_token_accuracy": 0.1455415517091751,
      "num_tokens": 24234352.0,
      "step": 11060
    },
    {
      "entropy": 5.5739212989807125,
      "epoch": 1.098154029376737,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004887919847440166,
      "loss": 5.3373,
      "mean_token_accuracy": 0.1551565557718277,
      "num_tokens": 24244137.0,
      "step": 11065
    },
    {
      "entropy": 5.56308388710022,
      "epoch": 1.0986502580389044,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004887809387474494,
      "loss": 5.3383,
      "mean_token_accuracy": 0.14976194500923157,
      "num_tokens": 24253250.0,
      "step": 11070
    },
    {
      "entropy": 5.623033666610718,
      "epoch": 1.0991464867010718,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004887698874495755,
      "loss": 5.4584,
      "mean_token_accuracy": 0.14906295388936996,
      "num_tokens": 24264551.0,
      "step": 11075
    },
    {
      "entropy": 5.572436189651489,
      "epoch": 1.0996427153632393,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000488758830850669,
      "loss": 5.266,
      "mean_token_accuracy": 0.1530448704957962,
      "num_tokens": 24275522.0,
      "step": 11080
    },
    {
      "entropy": 5.572006607055664,
      "epoch": 1.100138944025407,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004887477689510042,
      "loss": 5.4219,
      "mean_token_accuracy": 0.1521586999297142,
      "num_tokens": 24286333.0,
      "step": 11085
    },
    {
      "entropy": 5.511431694030762,
      "epoch": 1.1006351726875745,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004887367017508553,
      "loss": 5.4188,
      "mean_token_accuracy": 0.1553326889872551,
      "num_tokens": 24297627.0,
      "step": 11090
    },
    {
      "entropy": 5.521984672546386,
      "epoch": 1.101131401349742,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004887256292504967,
      "loss": 5.3782,
      "mean_token_accuracy": 0.14773138761520385,
      "num_tokens": 24308246.0,
      "step": 11095
    },
    {
      "entropy": 5.6061803817749025,
      "epoch": 1.1016276300119094,
      "grad_norm": 1.6875,
      "learning_rate": 0.000488714551450203,
      "loss": 5.4279,
      "mean_token_accuracy": 0.14293871819972992,
      "num_tokens": 24317718.0,
      "step": 11100
    },
    {
      "entropy": 5.580127239227295,
      "epoch": 1.102123858674077,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000488703468350249,
      "loss": 5.3651,
      "mean_token_accuracy": 0.15238318145275115,
      "num_tokens": 24329242.0,
      "step": 11105
    },
    {
      "entropy": 5.687979602813721,
      "epoch": 1.1026200873362446,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004886923799509095,
      "loss": 5.5305,
      "mean_token_accuracy": 0.1408196620643139,
      "num_tokens": 24340146.0,
      "step": 11110
    },
    {
      "entropy": 5.62099289894104,
      "epoch": 1.103116315998412,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004886812862524595,
      "loss": 5.402,
      "mean_token_accuracy": 0.14652805924415588,
      "num_tokens": 24349370.0,
      "step": 11115
    },
    {
      "entropy": 5.6466405391693115,
      "epoch": 1.1036125446605796,
      "grad_norm": 1.625,
      "learning_rate": 0.0004886701872551739,
      "loss": 5.3737,
      "mean_token_accuracy": 0.14662513881921768,
      "num_tokens": 24359770.0,
      "step": 11120
    },
    {
      "entropy": 5.592570400238037,
      "epoch": 1.104108773322747,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004886590829593281,
      "loss": 5.4346,
      "mean_token_accuracy": 0.14950930774211885,
      "num_tokens": 24370974.0,
      "step": 11125
    },
    {
      "entropy": 5.598513889312744,
      "epoch": 1.1046050019849147,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004886479733651974,
      "loss": 5.4154,
      "mean_token_accuracy": 0.1502979576587677,
      "num_tokens": 24381771.0,
      "step": 11130
    },
    {
      "entropy": 5.544974899291992,
      "epoch": 1.1051012306470822,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004886368584730575,
      "loss": 5.3869,
      "mean_token_accuracy": 0.14780377969145775,
      "num_tokens": 24393067.0,
      "step": 11135
    },
    {
      "entropy": 5.583708810806274,
      "epoch": 1.1055974593092497,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004886257382831837,
      "loss": 5.4285,
      "mean_token_accuracy": 0.14516044780611992,
      "num_tokens": 24402866.0,
      "step": 11140
    },
    {
      "entropy": 5.62046070098877,
      "epoch": 1.1060936879714172,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000488614612795852,
      "loss": 5.3755,
      "mean_token_accuracy": 0.15019112080335617,
      "num_tokens": 24412956.0,
      "step": 11145
    },
    {
      "entropy": 5.631631374359131,
      "epoch": 1.1065899166335849,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004886034820113382,
      "loss": 5.4299,
      "mean_token_accuracy": 0.14698757976293564,
      "num_tokens": 24424650.0,
      "step": 11150
    },
    {
      "entropy": 5.624796676635742,
      "epoch": 1.1070861452957523,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004885923459299182,
      "loss": 5.4471,
      "mean_token_accuracy": 0.15033173188567162,
      "num_tokens": 24436541.0,
      "step": 11155
    },
    {
      "entropy": 5.633722591400146,
      "epoch": 1.1075823739579198,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004885812045518685,
      "loss": 5.4482,
      "mean_token_accuracy": 0.1440286636352539,
      "num_tokens": 24447588.0,
      "step": 11160
    },
    {
      "entropy": 5.602619361877442,
      "epoch": 1.1080786026200873,
      "grad_norm": 1.796875,
      "learning_rate": 0.000488570057877465,
      "loss": 5.335,
      "mean_token_accuracy": 0.15218466967344285,
      "num_tokens": 24458854.0,
      "step": 11165
    },
    {
      "entropy": 5.56533637046814,
      "epoch": 1.108574831282255,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004885589059069842,
      "loss": 5.4258,
      "mean_token_accuracy": 0.15077076256275176,
      "num_tokens": 24469427.0,
      "step": 11170
    },
    {
      "entropy": 5.533487939834595,
      "epoch": 1.1090710599444225,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004885477486407028,
      "loss": 5.4084,
      "mean_token_accuracy": 0.1451013445854187,
      "num_tokens": 24479352.0,
      "step": 11175
    },
    {
      "entropy": 5.6563255310058596,
      "epoch": 1.10956728860659,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004885365860788974,
      "loss": 5.4142,
      "mean_token_accuracy": 0.14601799696683884,
      "num_tokens": 24489580.0,
      "step": 11180
    },
    {
      "entropy": 5.554216003417968,
      "epoch": 1.1100635172687574,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004885254182218448,
      "loss": 5.3525,
      "mean_token_accuracy": 0.15087575018405913,
      "num_tokens": 24501141.0,
      "step": 11185
    },
    {
      "entropy": 5.482413101196289,
      "epoch": 1.1105597459309249,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004885142450698219,
      "loss": 5.3723,
      "mean_token_accuracy": 0.15374670177698135,
      "num_tokens": 24512079.0,
      "step": 11190
    },
    {
      "entropy": 5.615344476699829,
      "epoch": 1.1110559745930926,
      "grad_norm": 1.625,
      "learning_rate": 0.0004885030666231057,
      "loss": 5.4195,
      "mean_token_accuracy": 0.1487134352326393,
      "num_tokens": 24521839.0,
      "step": 11195
    },
    {
      "entropy": 5.671700716018677,
      "epoch": 1.11155220325526,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004884918828819738,
      "loss": 5.4493,
      "mean_token_accuracy": 0.14406491219997405,
      "num_tokens": 24532704.0,
      "step": 11200
    },
    {
      "entropy": 5.617974328994751,
      "epoch": 1.1120484319174275,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000488480693846703,
      "loss": 5.3682,
      "mean_token_accuracy": 0.1483722858130932,
      "num_tokens": 24543790.0,
      "step": 11205
    },
    {
      "entropy": 5.561884546279908,
      "epoch": 1.112544660579595,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004884694995175711,
      "loss": 5.4188,
      "mean_token_accuracy": 0.1488062098622322,
      "num_tokens": 24554765.0,
      "step": 11210
    },
    {
      "entropy": 5.690768051147461,
      "epoch": 1.1130408892417627,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004884582998948555,
      "loss": 5.4983,
      "mean_token_accuracy": 0.14153558537364005,
      "num_tokens": 24565888.0,
      "step": 11215
    },
    {
      "entropy": 5.551495170593261,
      "epoch": 1.1135371179039302,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004884470949788342,
      "loss": 5.3391,
      "mean_token_accuracy": 0.1546345755457878,
      "num_tokens": 24575482.0,
      "step": 11220
    },
    {
      "entropy": 5.624620962142944,
      "epoch": 1.1140333465660976,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004884358847697847,
      "loss": 5.4705,
      "mean_token_accuracy": 0.1377727948129177,
      "num_tokens": 24588249.0,
      "step": 11225
    },
    {
      "entropy": 5.645045137405395,
      "epoch": 1.1145295752282651,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004884246692679852,
      "loss": 5.4215,
      "mean_token_accuracy": 0.1570864960551262,
      "num_tokens": 24599546.0,
      "step": 11230
    },
    {
      "entropy": 5.5755109786987305,
      "epoch": 1.1150258038904326,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004884134484737139,
      "loss": 5.3539,
      "mean_token_accuracy": 0.14995104670524598,
      "num_tokens": 24611777.0,
      "step": 11235
    },
    {
      "entropy": 5.566615295410156,
      "epoch": 1.1155220325526003,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004884022223872489,
      "loss": 5.3394,
      "mean_token_accuracy": 0.14883611872792243,
      "num_tokens": 24623675.0,
      "step": 11240
    },
    {
      "entropy": 5.576634073257447,
      "epoch": 1.1160182612147678,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004883909910088686,
      "loss": 5.3687,
      "mean_token_accuracy": 0.14955366104841233,
      "num_tokens": 24633929.0,
      "step": 11245
    },
    {
      "entropy": 5.5377113819122314,
      "epoch": 1.1165144898769352,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004883797543388515,
      "loss": 5.3199,
      "mean_token_accuracy": 0.1526401624083519,
      "num_tokens": 24642800.0,
      "step": 11250
    },
    {
      "entropy": 5.54152626991272,
      "epoch": 1.1170107185391027,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004883685123774764,
      "loss": 5.3265,
      "mean_token_accuracy": 0.15583520084619523,
      "num_tokens": 24653780.0,
      "step": 11255
    },
    {
      "entropy": 5.562276411056518,
      "epoch": 1.1175069472012704,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004883572651250218,
      "loss": 5.3531,
      "mean_token_accuracy": 0.15438604205846787,
      "num_tokens": 24663912.0,
      "step": 11260
    },
    {
      "entropy": 5.594871377944946,
      "epoch": 1.118003175863438,
      "grad_norm": 1.375,
      "learning_rate": 0.0004883460125817669,
      "loss": 5.4449,
      "mean_token_accuracy": 0.14818885177373886,
      "num_tokens": 24675222.0,
      "step": 11265
    },
    {
      "entropy": 5.614373207092285,
      "epoch": 1.1184994045256054,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004883347547479905,
      "loss": 5.4603,
      "mean_token_accuracy": 0.1456131726503372,
      "num_tokens": 24686697.0,
      "step": 11270
    },
    {
      "entropy": 5.667960214614868,
      "epoch": 1.1189956331877728,
      "grad_norm": 1.4375,
      "learning_rate": 0.000488323491623972,
      "loss": 5.3697,
      "mean_token_accuracy": 0.1532883659005165,
      "num_tokens": 24697067.0,
      "step": 11275
    },
    {
      "entropy": 5.633295965194702,
      "epoch": 1.1194918618499405,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004883122232099905,
      "loss": 5.3834,
      "mean_token_accuracy": 0.15083587914705276,
      "num_tokens": 24707989.0,
      "step": 11280
    },
    {
      "entropy": 5.504376840591431,
      "epoch": 1.119988090512108,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004883009495063256,
      "loss": 5.3197,
      "mean_token_accuracy": 0.15871001332998275,
      "num_tokens": 24718826.0,
      "step": 11285
    },
    {
      "entropy": 5.581752586364746,
      "epoch": 1.1204843191742755,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004882896705132566,
      "loss": 5.4174,
      "mean_token_accuracy": 0.1424231544137001,
      "num_tokens": 24729913.0,
      "step": 11290
    },
    {
      "entropy": 5.532200574874878,
      "epoch": 1.120980547836443,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048827838623106355,
      "loss": 5.2921,
      "mean_token_accuracy": 0.15641152560710908,
      "num_tokens": 24741621.0,
      "step": 11295
    },
    {
      "entropy": 5.5503137588500975,
      "epoch": 1.1214767764986107,
      "grad_norm": 1.671875,
      "learning_rate": 0.000488267096660026,
      "loss": 5.3515,
      "mean_token_accuracy": 0.15093980580568314,
      "num_tokens": 24753427.0,
      "step": 11300
    },
    {
      "entropy": 5.577446746826172,
      "epoch": 1.1219730051607781,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048825580180042406,
      "loss": 5.3916,
      "mean_token_accuracy": 0.1487269878387451,
      "num_tokens": 24763580.0,
      "step": 11305
    },
    {
      "entropy": 5.555149698257447,
      "epoch": 1.1224692338229456,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004882445016525377,
      "loss": 5.3641,
      "mean_token_accuracy": 0.15085773915052414,
      "num_tokens": 24773955.0,
      "step": 11310
    },
    {
      "entropy": 5.559953594207764,
      "epoch": 1.122965462485113,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048823319621664723,
      "loss": 5.2946,
      "mean_token_accuracy": 0.15657372176647186,
      "num_tokens": 24784039.0,
      "step": 11315
    },
    {
      "entropy": 5.656080293655395,
      "epoch": 1.1234616911472806,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048822188549303296,
      "loss": 5.4021,
      "mean_token_accuracy": 0.1469716176390648,
      "num_tokens": 24794630.0,
      "step": 11320
    },
    {
      "entropy": 5.555363035202026,
      "epoch": 1.1239579198094483,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004882105694819754,
      "loss": 5.4407,
      "mean_token_accuracy": 0.14273337945342063,
      "num_tokens": 24805625.0,
      "step": 11325
    },
    {
      "entropy": 5.57168288230896,
      "epoch": 1.1244541484716157,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004881992481837551,
      "loss": 5.3878,
      "mean_token_accuracy": 0.1549401804804802,
      "num_tokens": 24817356.0,
      "step": 11330
    },
    {
      "entropy": 5.584911441802978,
      "epoch": 1.1249503771337832,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004881879215986529,
      "loss": 5.3519,
      "mean_token_accuracy": 0.1435728743672371,
      "num_tokens": 24828349.0,
      "step": 11335
    },
    {
      "entropy": 5.601088047027588,
      "epoch": 1.1254466057959507,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048817658972694967,
      "loss": 5.3958,
      "mean_token_accuracy": 0.14754419177770614,
      "num_tokens": 24840138.0,
      "step": 11340
    },
    {
      "entropy": 5.580916881561279,
      "epoch": 1.1259428344581184,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004881652525689263,
      "loss": 5.4017,
      "mean_token_accuracy": 0.14724118560552596,
      "num_tokens": 24851570.0,
      "step": 11345
    },
    {
      "entropy": 5.459811401367188,
      "epoch": 1.1264390631202859,
      "grad_norm": 1.375,
      "learning_rate": 0.000488153910124864,
      "loss": 5.2869,
      "mean_token_accuracy": 0.15560160875320433,
      "num_tokens": 24861396.0,
      "step": 11350
    },
    {
      "entropy": 5.582611465454102,
      "epoch": 1.1269352917824533,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000488142562395044,
      "loss": 5.3989,
      "mean_token_accuracy": 0.1454012505710125,
      "num_tokens": 24873519.0,
      "step": 11355
    },
    {
      "entropy": 5.587449455261231,
      "epoch": 1.1274315204446208,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048813120937974774,
      "loss": 5.438,
      "mean_token_accuracy": 0.1463127113878727,
      "num_tokens": 24886856.0,
      "step": 11360
    },
    {
      "entropy": 5.539682912826538,
      "epoch": 1.1279277491067885,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048811985107925684,
      "loss": 5.3729,
      "mean_token_accuracy": 0.14519568681716918,
      "num_tokens": 24897292.0,
      "step": 11365
    },
    {
      "entropy": 5.557141065597534,
      "epoch": 1.128423977768956,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048810848749385286,
      "loss": 5.3816,
      "mean_token_accuracy": 0.14538933336734772,
      "num_tokens": 24907757.0,
      "step": 11370
    },
    {
      "entropy": 5.495296573638916,
      "epoch": 1.1289202064311235,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048809711862381753,
      "loss": 5.3544,
      "mean_token_accuracy": 0.15524989366531372,
      "num_tokens": 24918071.0,
      "step": 11375
    },
    {
      "entropy": 5.5515491485595705,
      "epoch": 1.129416435093291,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048808574446943274,
      "loss": 5.338,
      "mean_token_accuracy": 0.1514362335205078,
      "num_tokens": 24927967.0,
      "step": 11380
    },
    {
      "entropy": 5.6297282695770265,
      "epoch": 1.1299126637554586,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004880743650309808,
      "loss": 5.4128,
      "mean_token_accuracy": 0.15155599117279053,
      "num_tokens": 24939170.0,
      "step": 11385
    },
    {
      "entropy": 5.541307210922241,
      "epoch": 1.130408892417626,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048806298030874373,
      "loss": 5.3993,
      "mean_token_accuracy": 0.15357860848307608,
      "num_tokens": 24950937.0,
      "step": 11390
    },
    {
      "entropy": 5.534867906570435,
      "epoch": 1.1309051210797936,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048805159030300383,
      "loss": 5.3916,
      "mean_token_accuracy": 0.14766136854887008,
      "num_tokens": 24962813.0,
      "step": 11395
    },
    {
      "entropy": 5.589195775985718,
      "epoch": 1.131401349741961,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048804019501404357,
      "loss": 5.2948,
      "mean_token_accuracy": 0.1516301840543747,
      "num_tokens": 24974192.0,
      "step": 11400
    },
    {
      "entropy": 5.555340719223023,
      "epoch": 1.1318975784041285,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048802879444214553,
      "loss": 5.2876,
      "mean_token_accuracy": 0.15271585881710054,
      "num_tokens": 24985353.0,
      "step": 11405
    },
    {
      "entropy": 5.53402009010315,
      "epoch": 1.1323938070662962,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048801738858759243,
      "loss": 5.3871,
      "mean_token_accuracy": 0.1554318904876709,
      "num_tokens": 24997259.0,
      "step": 11410
    },
    {
      "entropy": 5.5836423397064205,
      "epoch": 1.1328900357284637,
      "grad_norm": 1.375,
      "learning_rate": 0.0004880059774506671,
      "loss": 5.3755,
      "mean_token_accuracy": 0.1500416189432144,
      "num_tokens": 25008222.0,
      "step": 11415
    },
    {
      "entropy": 5.598425054550171,
      "epoch": 1.1333862643906312,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004879945610316525,
      "loss": 5.4116,
      "mean_token_accuracy": 0.14301400929689406,
      "num_tokens": 25020267.0,
      "step": 11420
    },
    {
      "entropy": 5.572281789779663,
      "epoch": 1.1338824930527986,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048798313933083175,
      "loss": 5.4013,
      "mean_token_accuracy": 0.1483250617980957,
      "num_tokens": 25031212.0,
      "step": 11425
    },
    {
      "entropy": 5.533501672744751,
      "epoch": 1.1343787217149663,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004879717123484882,
      "loss": 5.3535,
      "mean_token_accuracy": 0.14589081704616547,
      "num_tokens": 25042204.0,
      "step": 11430
    },
    {
      "entropy": 5.594161128997802,
      "epoch": 1.1348749503771338,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000487960280084905,
      "loss": 5.3769,
      "mean_token_accuracy": 0.15346802026033401,
      "num_tokens": 25052010.0,
      "step": 11435
    },
    {
      "entropy": 5.62937216758728,
      "epoch": 1.1353711790393013,
      "grad_norm": 1.25,
      "learning_rate": 0.0004879488425403657,
      "loss": 5.437,
      "mean_token_accuracy": 0.1481371022760868,
      "num_tokens": 25062209.0,
      "step": 11440
    },
    {
      "entropy": 5.576134300231933,
      "epoch": 1.1358674077014688,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048793739971515407,
      "loss": 5.407,
      "mean_token_accuracy": 0.14646496027708053,
      "num_tokens": 25072758.0,
      "step": 11445
    },
    {
      "entropy": 5.5203221321105955,
      "epoch": 1.1363636363636362,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048792595160955374,
      "loss": 5.4622,
      "mean_token_accuracy": 0.14955125972628594,
      "num_tokens": 25084318.0,
      "step": 11450
    },
    {
      "entropy": 5.598471975326538,
      "epoch": 1.136859865025804,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048791449822384856,
      "loss": 5.353,
      "mean_token_accuracy": 0.15065220966935158,
      "num_tokens": 25095715.0,
      "step": 11455
    },
    {
      "entropy": 5.619252872467041,
      "epoch": 1.1373560936879714,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004879030395583226,
      "loss": 5.3781,
      "mean_token_accuracy": 0.1509352833032608,
      "num_tokens": 25106163.0,
      "step": 11460
    },
    {
      "entropy": 5.590196704864502,
      "epoch": 1.137852322350139,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004878915756132601,
      "loss": 5.4547,
      "mean_token_accuracy": 0.14622181802988052,
      "num_tokens": 25118745.0,
      "step": 11465
    },
    {
      "entropy": 5.600633001327514,
      "epoch": 1.1383485510123066,
      "grad_norm": 2.5,
      "learning_rate": 0.00048788010638894523,
      "loss": 5.435,
      "mean_token_accuracy": 0.144600510597229,
      "num_tokens": 25130066.0,
      "step": 11470
    },
    {
      "entropy": 5.553614234924316,
      "epoch": 1.138844779674474,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004878686318856624,
      "loss": 5.2863,
      "mean_token_accuracy": 0.1518443778157234,
      "num_tokens": 25139895.0,
      "step": 11475
    },
    {
      "entropy": 5.585431528091431,
      "epoch": 1.1393410083366415,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048785715210369635,
      "loss": 5.4931,
      "mean_token_accuracy": 0.14274978041648864,
      "num_tokens": 25150644.0,
      "step": 11480
    },
    {
      "entropy": 5.564278936386108,
      "epoch": 1.139837236998809,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004878456670433314,
      "loss": 5.4333,
      "mean_token_accuracy": 0.14598654210567474,
      "num_tokens": 25163038.0,
      "step": 11485
    },
    {
      "entropy": 5.541059589385986,
      "epoch": 1.1403334656609765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004878341767048527,
      "loss": 5.3296,
      "mean_token_accuracy": 0.15456676185131074,
      "num_tokens": 25174154.0,
      "step": 11490
    },
    {
      "entropy": 5.656386804580689,
      "epoch": 1.1408296943231442,
      "grad_norm": 1.5,
      "learning_rate": 0.00048782268108854494,
      "loss": 5.4475,
      "mean_token_accuracy": 0.14694804847240447,
      "num_tokens": 25185369.0,
      "step": 11495
    },
    {
      "entropy": 5.586584281921387,
      "epoch": 1.1413259229853117,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004878111801946933,
      "loss": 5.4458,
      "mean_token_accuracy": 0.14702797830104827,
      "num_tokens": 25197875.0,
      "step": 11500
    },
    {
      "entropy": 5.523218154907227,
      "epoch": 1.1418221516474791,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048779967402358296,
      "loss": 5.3216,
      "mean_token_accuracy": 0.15915649980306626,
      "num_tokens": 25208362.0,
      "step": 11505
    },
    {
      "entropy": 5.636033201217652,
      "epoch": 1.1423183803096466,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004877881625754993,
      "loss": 5.52,
      "mean_token_accuracy": 0.14281244874000548,
      "num_tokens": 25220441.0,
      "step": 11510
    },
    {
      "entropy": 5.547951316833496,
      "epoch": 1.1428146089718143,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048777664585072765,
      "loss": 5.2872,
      "mean_token_accuracy": 0.15461819916963576,
      "num_tokens": 25231134.0,
      "step": 11515
    },
    {
      "entropy": 5.621464729309082,
      "epoch": 1.1433108376339818,
      "grad_norm": 1.375,
      "learning_rate": 0.0004877651238495537,
      "loss": 5.5409,
      "mean_token_accuracy": 0.14054297730326654,
      "num_tokens": 25242327.0,
      "step": 11520
    },
    {
      "entropy": 5.589229154586792,
      "epoch": 1.1438070662961493,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048775359657226316,
      "loss": 5.385,
      "mean_token_accuracy": 0.14980412125587464,
      "num_tokens": 25253571.0,
      "step": 11525
    },
    {
      "entropy": 5.52040376663208,
      "epoch": 1.1443032949583167,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004877420640191418,
      "loss": 5.3344,
      "mean_token_accuracy": 0.15033116042613984,
      "num_tokens": 25264573.0,
      "step": 11530
    },
    {
      "entropy": 5.570263433456421,
      "epoch": 1.1447995236204842,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004877305261904758,
      "loss": 5.4182,
      "mean_token_accuracy": 0.14297931492328644,
      "num_tokens": 25275197.0,
      "step": 11535
    },
    {
      "entropy": 5.551829290390015,
      "epoch": 1.145295752282652,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048771898308655107,
      "loss": 5.3098,
      "mean_token_accuracy": 0.15579285025596618,
      "num_tokens": 25286564.0,
      "step": 11540
    },
    {
      "entropy": 5.563176155090332,
      "epoch": 1.1457919809448194,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004877074347076539,
      "loss": 5.3623,
      "mean_token_accuracy": 0.15205119252204896,
      "num_tokens": 25297874.0,
      "step": 11545
    },
    {
      "entropy": 5.592676019668579,
      "epoch": 1.1462882096069869,
      "grad_norm": 1.75,
      "learning_rate": 0.00048769588105407075,
      "loss": 5.4151,
      "mean_token_accuracy": 0.14948319047689437,
      "num_tokens": 25308047.0,
      "step": 11550
    },
    {
      "entropy": 5.673816394805908,
      "epoch": 1.1467844382691545,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048768432212608807,
      "loss": 5.4674,
      "mean_token_accuracy": 0.14263079687952995,
      "num_tokens": 25318527.0,
      "step": 11555
    },
    {
      "entropy": 5.627174139022827,
      "epoch": 1.147280666931322,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004876727579239925,
      "loss": 5.3576,
      "mean_token_accuracy": 0.1516485631465912,
      "num_tokens": 25328242.0,
      "step": 11560
    },
    {
      "entropy": 5.572349262237549,
      "epoch": 1.1477768955934895,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004876611884480708,
      "loss": 5.3622,
      "mean_token_accuracy": 0.15511356741189958,
      "num_tokens": 25338214.0,
      "step": 11565
    },
    {
      "entropy": 5.593915414810181,
      "epoch": 1.148273124255657,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048764961369860977,
      "loss": 5.397,
      "mean_token_accuracy": 0.14481146261096,
      "num_tokens": 25349762.0,
      "step": 11570
    },
    {
      "entropy": 5.596378517150879,
      "epoch": 1.1487693529178244,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048763803367589663,
      "loss": 5.3969,
      "mean_token_accuracy": 0.14947986975312233,
      "num_tokens": 25361241.0,
      "step": 11575
    },
    {
      "entropy": 5.617586374282837,
      "epoch": 1.1492655815799921,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048762644838021843,
      "loss": 5.432,
      "mean_token_accuracy": 0.14766261354088783,
      "num_tokens": 25373846.0,
      "step": 11580
    },
    {
      "entropy": 5.56834716796875,
      "epoch": 1.1497618102421596,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048761485781186243,
      "loss": 5.3473,
      "mean_token_accuracy": 0.1482924848794937,
      "num_tokens": 25385331.0,
      "step": 11585
    },
    {
      "entropy": 5.544664478302002,
      "epoch": 1.150258038904327,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004876032619711162,
      "loss": 5.3681,
      "mean_token_accuracy": 0.1553872123360634,
      "num_tokens": 25395140.0,
      "step": 11590
    },
    {
      "entropy": 5.63348126411438,
      "epoch": 1.1507542675664946,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048759166085826714,
      "loss": 5.4192,
      "mean_token_accuracy": 0.14311678558588029,
      "num_tokens": 25405455.0,
      "step": 11595
    },
    {
      "entropy": 5.598033142089844,
      "epoch": 1.1512504962286623,
      "grad_norm": 1.375,
      "learning_rate": 0.000487580054473603,
      "loss": 5.4307,
      "mean_token_accuracy": 0.14684400707483292,
      "num_tokens": 25417000.0,
      "step": 11600
    },
    {
      "entropy": 5.559065103530884,
      "epoch": 1.1517467248908297,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048756844281741156,
      "loss": 5.3519,
      "mean_token_accuracy": 0.15599970817565917,
      "num_tokens": 25427808.0,
      "step": 11605
    },
    {
      "entropy": 5.579971694946289,
      "epoch": 1.1522429535529972,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048755682588998085,
      "loss": 5.3659,
      "mean_token_accuracy": 0.14552090615034102,
      "num_tokens": 25438302.0,
      "step": 11610
    },
    {
      "entropy": 5.470318603515625,
      "epoch": 1.1527391822151647,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004875452036915987,
      "loss": 5.2598,
      "mean_token_accuracy": 0.15168599784374237,
      "num_tokens": 25449335.0,
      "step": 11615
    },
    {
      "entropy": 5.682684469223022,
      "epoch": 1.1532354108773322,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004875335762225536,
      "loss": 5.4653,
      "mean_token_accuracy": 0.14342744648456573,
      "num_tokens": 25461173.0,
      "step": 11620
    },
    {
      "entropy": 5.653379011154175,
      "epoch": 1.1537316395394999,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048752194348313383,
      "loss": 5.3482,
      "mean_token_accuracy": 0.14651379808783532,
      "num_tokens": 25472538.0,
      "step": 11625
    },
    {
      "entropy": 5.523631381988525,
      "epoch": 1.1542278682016673,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004875103054736278,
      "loss": 5.3916,
      "mean_token_accuracy": 0.15227073282003403,
      "num_tokens": 25483386.0,
      "step": 11630
    },
    {
      "entropy": 5.586606121063232,
      "epoch": 1.1547240968638348,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004874986621943241,
      "loss": 5.4243,
      "mean_token_accuracy": 0.14751880690455438,
      "num_tokens": 25495505.0,
      "step": 11635
    },
    {
      "entropy": 5.604973316192627,
      "epoch": 1.1552203255260023,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048748701364551147,
      "loss": 5.3808,
      "mean_token_accuracy": 0.1531009167432785,
      "num_tokens": 25506144.0,
      "step": 11640
    },
    {
      "entropy": 5.588434505462646,
      "epoch": 1.15571655418817,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004874753598274787,
      "loss": 5.3828,
      "mean_token_accuracy": 0.1444155365228653,
      "num_tokens": 25516185.0,
      "step": 11645
    },
    {
      "entropy": 5.634286832809448,
      "epoch": 1.1562127828503375,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004874637007405149,
      "loss": 5.467,
      "mean_token_accuracy": 0.14882135838270188,
      "num_tokens": 25526657.0,
      "step": 11650
    },
    {
      "entropy": 5.503026103973388,
      "epoch": 1.156709011512505,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048745203638490917,
      "loss": 5.3171,
      "mean_token_accuracy": 0.15622069090604782,
      "num_tokens": 25537101.0,
      "step": 11655
    },
    {
      "entropy": 5.496335220336914,
      "epoch": 1.1572052401746724,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048744036676095067,
      "loss": 5.4069,
      "mean_token_accuracy": 0.14174350649118422,
      "num_tokens": 25547672.0,
      "step": 11660
    },
    {
      "entropy": 5.6529953956604,
      "epoch": 1.15770146883684,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004874286918689289,
      "loss": 5.474,
      "mean_token_accuracy": 0.14624751657247542,
      "num_tokens": 25558408.0,
      "step": 11665
    },
    {
      "entropy": 5.6682515144348145,
      "epoch": 1.1581976974990076,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00048741701170913326,
      "loss": 5.375,
      "mean_token_accuracy": 0.1485062211751938,
      "num_tokens": 25569823.0,
      "step": 11670
    },
    {
      "entropy": 5.514466905593872,
      "epoch": 1.158693926161175,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048740532628185353,
      "loss": 5.3651,
      "mean_token_accuracy": 0.15548053830862046,
      "num_tokens": 25579705.0,
      "step": 11675
    },
    {
      "entropy": 5.516293525695801,
      "epoch": 1.1591901548233425,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004873936355873793,
      "loss": 5.4492,
      "mean_token_accuracy": 0.14290979132056236,
      "num_tokens": 25590573.0,
      "step": 11680
    },
    {
      "entropy": 5.541019105911255,
      "epoch": 1.1596863834855102,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004873819396260007,
      "loss": 5.3554,
      "mean_token_accuracy": 0.1528298392891884,
      "num_tokens": 25601418.0,
      "step": 11685
    },
    {
      "entropy": 5.595352745056152,
      "epoch": 1.1601826121476777,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048737023839800754,
      "loss": 5.4405,
      "mean_token_accuracy": 0.1474393129348755,
      "num_tokens": 25612409.0,
      "step": 11690
    },
    {
      "entropy": 5.565690183639527,
      "epoch": 1.1606788408098452,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004873585319036901,
      "loss": 5.3345,
      "mean_token_accuracy": 0.15232966244220733,
      "num_tokens": 25623547.0,
      "step": 11695
    },
    {
      "entropy": 5.576412153244019,
      "epoch": 1.1611750694720127,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004873468201433387,
      "loss": 5.5165,
      "mean_token_accuracy": 0.14460463374853133,
      "num_tokens": 25636537.0,
      "step": 11700
    },
    {
      "entropy": 5.588827085494995,
      "epoch": 1.1616712981341801,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004873351031172437,
      "loss": 5.3694,
      "mean_token_accuracy": 0.15230894535779954,
      "num_tokens": 25646734.0,
      "step": 11705
    },
    {
      "entropy": 5.655239248275757,
      "epoch": 1.1621675267963478,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048732338082569564,
      "loss": 5.4707,
      "mean_token_accuracy": 0.15324380844831467,
      "num_tokens": 25658058.0,
      "step": 11710
    },
    {
      "entropy": 5.538433265686035,
      "epoch": 1.1626637554585153,
      "grad_norm": 1.375,
      "learning_rate": 0.0004873116532689853,
      "loss": 5.2949,
      "mean_token_accuracy": 0.15012841373682023,
      "num_tokens": 25669630.0,
      "step": 11715
    },
    {
      "entropy": 5.502696800231933,
      "epoch": 1.1631599841206828,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004872999204474035,
      "loss": 5.3063,
      "mean_token_accuracy": 0.1526948854327202,
      "num_tokens": 25681928.0,
      "step": 11720
    },
    {
      "entropy": 5.500417852401734,
      "epoch": 1.1636562127828503,
      "grad_norm": 1.421875,
      "learning_rate": 0.000487288182361241,
      "loss": 5.2958,
      "mean_token_accuracy": 0.1499994397163391,
      "num_tokens": 25690653.0,
      "step": 11725
    },
    {
      "entropy": 5.59794282913208,
      "epoch": 1.164152441445018,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004872764390107891,
      "loss": 5.4243,
      "mean_token_accuracy": 0.1429888904094696,
      "num_tokens": 25701201.0,
      "step": 11730
    },
    {
      "entropy": 5.516219425201416,
      "epoch": 1.1646486701071854,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048726469039633895,
      "loss": 5.3214,
      "mean_token_accuracy": 0.14950628578662872,
      "num_tokens": 25710927.0,
      "step": 11735
    },
    {
      "entropy": 5.5694129943847654,
      "epoch": 1.165144898769353,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048725293651818183,
      "loss": 5.3734,
      "mean_token_accuracy": 0.1527511715888977,
      "num_tokens": 25722136.0,
      "step": 11740
    },
    {
      "entropy": 5.533573579788208,
      "epoch": 1.1656411274315204,
      "grad_norm": 1.25,
      "learning_rate": 0.0004872411773766093,
      "loss": 5.2517,
      "mean_token_accuracy": 0.16131042093038558,
      "num_tokens": 25733386.0,
      "step": 11745
    },
    {
      "entropy": 5.4848387241363525,
      "epoch": 1.1661373560936878,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048722941297191284,
      "loss": 5.3416,
      "mean_token_accuracy": 0.15170163214206694,
      "num_tokens": 25742801.0,
      "step": 11750
    },
    {
      "entropy": 5.630923080444336,
      "epoch": 1.1666335847558555,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004872176433043843,
      "loss": 5.4556,
      "mean_token_accuracy": 0.14952896758913994,
      "num_tokens": 25755594.0,
      "step": 11755
    },
    {
      "entropy": 5.614663314819336,
      "epoch": 1.167129813418023,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004872058683743155,
      "loss": 5.3858,
      "mean_token_accuracy": 0.15365967899560928,
      "num_tokens": 25766271.0,
      "step": 11760
    },
    {
      "entropy": 5.588860702514649,
      "epoch": 1.1676260420801905,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048719408818199836,
      "loss": 5.4265,
      "mean_token_accuracy": 0.14932061731815338,
      "num_tokens": 25777338.0,
      "step": 11765
    },
    {
      "entropy": 5.550373125076294,
      "epoch": 1.1681222707423582,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048718230272772517,
      "loss": 5.3087,
      "mean_token_accuracy": 0.1521749585866928,
      "num_tokens": 25787968.0,
      "step": 11770
    },
    {
      "entropy": 5.580430746078491,
      "epoch": 1.1686184994045257,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048717051201178804,
      "loss": 5.3954,
      "mean_token_accuracy": 0.15113214254379273,
      "num_tokens": 25798274.0,
      "step": 11775
    },
    {
      "entropy": 5.533266019821167,
      "epoch": 1.1691147280666931,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004871587160344794,
      "loss": 5.3797,
      "mean_token_accuracy": 0.14575085937976837,
      "num_tokens": 25809140.0,
      "step": 11780
    },
    {
      "entropy": 5.576405763626099,
      "epoch": 1.1696109567288606,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004871469147960917,
      "loss": 5.4082,
      "mean_token_accuracy": 0.14604080468416214,
      "num_tokens": 25821502.0,
      "step": 11785
    },
    {
      "entropy": 5.592173147201538,
      "epoch": 1.170107185391028,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004871351082969177,
      "loss": 5.2936,
      "mean_token_accuracy": 0.1577027216553688,
      "num_tokens": 25833231.0,
      "step": 11790
    },
    {
      "entropy": 5.6552649974823,
      "epoch": 1.1706034140531958,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004871232965372501,
      "loss": 5.4696,
      "mean_token_accuracy": 0.14471988752484322,
      "num_tokens": 25844867.0,
      "step": 11795
    },
    {
      "entropy": 5.581682014465332,
      "epoch": 1.1710996427153633,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004871114795173819,
      "loss": 5.4289,
      "mean_token_accuracy": 0.15717205107212068,
      "num_tokens": 25855959.0,
      "step": 11800
    },
    {
      "entropy": 5.551369047164917,
      "epoch": 1.1715958713775307,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004870996572376061,
      "loss": 5.5067,
      "mean_token_accuracy": 0.15130186975002288,
      "num_tokens": 25866926.0,
      "step": 11805
    },
    {
      "entropy": 5.666511344909668,
      "epoch": 1.1720921000396982,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004870878296982157,
      "loss": 5.4432,
      "mean_token_accuracy": 0.14657786190509797,
      "num_tokens": 25877699.0,
      "step": 11810
    },
    {
      "entropy": 5.6513392448425295,
      "epoch": 1.172588328701866,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004870759968995042,
      "loss": 5.3827,
      "mean_token_accuracy": 0.14910238981246948,
      "num_tokens": 25888640.0,
      "step": 11815
    },
    {
      "entropy": 5.543155336380005,
      "epoch": 1.1730845573640334,
      "grad_norm": 1.25,
      "learning_rate": 0.00048706415884176493,
      "loss": 5.3807,
      "mean_token_accuracy": 0.14920878410339355,
      "num_tokens": 25900651.0,
      "step": 11820
    },
    {
      "entropy": 5.5515306949615475,
      "epoch": 1.1735807860262009,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004870523155252915,
      "loss": 5.3595,
      "mean_token_accuracy": 0.15301627963781356,
      "num_tokens": 25911775.0,
      "step": 11825
    },
    {
      "entropy": 5.5304583549499515,
      "epoch": 1.1740770146883683,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004870404669503776,
      "loss": 5.3341,
      "mean_token_accuracy": 0.14804167300462723,
      "num_tokens": 25922126.0,
      "step": 11830
    },
    {
      "entropy": 5.606599378585815,
      "epoch": 1.1745732433505358,
      "grad_norm": 1.578125,
      "learning_rate": 0.000487028613117317,
      "loss": 5.3872,
      "mean_token_accuracy": 0.14983359575271607,
      "num_tokens": 25932945.0,
      "step": 11835
    },
    {
      "entropy": 5.639219903945923,
      "epoch": 1.1750694720127035,
      "grad_norm": 1.625,
      "learning_rate": 0.0004870167540264037,
      "loss": 5.4419,
      "mean_token_accuracy": 0.14347726702690125,
      "num_tokens": 25944277.0,
      "step": 11840
    },
    {
      "entropy": 5.51197566986084,
      "epoch": 1.175565700674871,
      "grad_norm": 4.65625,
      "learning_rate": 0.0004870048896779317,
      "loss": 5.3037,
      "mean_token_accuracy": 0.15606016665697098,
      "num_tokens": 25956838.0,
      "step": 11845
    },
    {
      "entropy": 5.555289363861084,
      "epoch": 1.1760619293370385,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048699302007219533,
      "loss": 5.3631,
      "mean_token_accuracy": 0.15336766541004182,
      "num_tokens": 25966639.0,
      "step": 11850
    },
    {
      "entropy": 5.564569568634033,
      "epoch": 1.1765581579992062,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004869811452094889,
      "loss": 5.4023,
      "mean_token_accuracy": 0.15658990144729615,
      "num_tokens": 25978209.0,
      "step": 11855
    },
    {
      "entropy": 5.524564552307129,
      "epoch": 1.1770543866613736,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048696926509010673,
      "loss": 5.3448,
      "mean_token_accuracy": 0.15160741806030273,
      "num_tokens": 25989306.0,
      "step": 11860
    },
    {
      "entropy": 5.554265308380127,
      "epoch": 1.177550615323541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004869573797143437,
      "loss": 5.4322,
      "mean_token_accuracy": 0.14990700781345367,
      "num_tokens": 26000235.0,
      "step": 11865
    },
    {
      "entropy": 5.5678435325622555,
      "epoch": 1.1780468439857086,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004869454890824942,
      "loss": 5.3348,
      "mean_token_accuracy": 0.15315441489219667,
      "num_tokens": 26010616.0,
      "step": 11870
    },
    {
      "entropy": 5.563578414916992,
      "epoch": 1.178543072647876,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048693359319485337,
      "loss": 5.4043,
      "mean_token_accuracy": 0.14328229501843454,
      "num_tokens": 26021303.0,
      "step": 11875
    },
    {
      "entropy": 5.591171741485596,
      "epoch": 1.1790393013100438,
      "grad_norm": 1.25,
      "learning_rate": 0.0004869216920517161,
      "loss": 5.3446,
      "mean_token_accuracy": 0.15115900933742524,
      "num_tokens": 26032386.0,
      "step": 11880
    },
    {
      "entropy": 5.5014537334442135,
      "epoch": 1.1795355299722112,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048690978565337753,
      "loss": 5.2933,
      "mean_token_accuracy": 0.14972033798694612,
      "num_tokens": 26042889.0,
      "step": 11885
    },
    {
      "entropy": 5.592734861373901,
      "epoch": 1.1800317586343787,
      "grad_norm": 2.28125,
      "learning_rate": 0.00048689787400013287,
      "loss": 5.421,
      "mean_token_accuracy": 0.14785860627889633,
      "num_tokens": 26054299.0,
      "step": 11890
    },
    {
      "entropy": 5.698275232315064,
      "epoch": 1.1805279872965462,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004868859570922775,
      "loss": 5.4462,
      "mean_token_accuracy": 0.14040206521749496,
      "num_tokens": 26064587.0,
      "step": 11895
    },
    {
      "entropy": 5.633814287185669,
      "epoch": 1.1810242159587139,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004868740349301071,
      "loss": 5.3974,
      "mean_token_accuracy": 0.15465076714754106,
      "num_tokens": 26076495.0,
      "step": 11900
    },
    {
      "entropy": 5.561066436767578,
      "epoch": 1.1815204446208813,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000486862107513917,
      "loss": 5.3875,
      "mean_token_accuracy": 0.14858879521489143,
      "num_tokens": 26086496.0,
      "step": 11905
    },
    {
      "entropy": 5.559374189376831,
      "epoch": 1.1820166732830488,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004868501748440033,
      "loss": 5.3617,
      "mean_token_accuracy": 0.14975929856300355,
      "num_tokens": 26097263.0,
      "step": 11910
    },
    {
      "entropy": 5.618248033523559,
      "epoch": 1.1825129019452163,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004868382369206617,
      "loss": 5.381,
      "mean_token_accuracy": 0.14991446137428283,
      "num_tokens": 26108264.0,
      "step": 11915
    },
    {
      "entropy": 5.590499591827393,
      "epoch": 1.1830091306073838,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004868262937441883,
      "loss": 5.4378,
      "mean_token_accuracy": 0.15564050227403642,
      "num_tokens": 26119730.0,
      "step": 11920
    },
    {
      "entropy": 5.577827024459839,
      "epoch": 1.1835053592695515,
      "grad_norm": 1.90625,
      "learning_rate": 0.00048681434531487925,
      "loss": 5.3395,
      "mean_token_accuracy": 0.1498294860124588,
      "num_tokens": 26131055.0,
      "step": 11925
    },
    {
      "entropy": 5.502968168258667,
      "epoch": 1.184001587931719,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048680239163303085,
      "loss": 5.3297,
      "mean_token_accuracy": 0.15462222546339036,
      "num_tokens": 26141577.0,
      "step": 11930
    },
    {
      "entropy": 5.548702907562256,
      "epoch": 1.1844978165938864,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004867904326989396,
      "loss": 5.3936,
      "mean_token_accuracy": 0.14784821420907973,
      "num_tokens": 26152733.0,
      "step": 11935
    },
    {
      "entropy": 5.5944984436035154,
      "epoch": 1.184994045256054,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004867784685129019,
      "loss": 5.4341,
      "mean_token_accuracy": 0.14854197949171066,
      "num_tokens": 26164763.0,
      "step": 11940
    },
    {
      "entropy": 5.567025136947632,
      "epoch": 1.1854902739182216,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048676649907521455,
      "loss": 5.2893,
      "mean_token_accuracy": 0.1543369024991989,
      "num_tokens": 26174391.0,
      "step": 11945
    },
    {
      "entropy": 5.497171068191529,
      "epoch": 1.185986502580389,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048675452438617434,
      "loss": 5.2757,
      "mean_token_accuracy": 0.1590538203716278,
      "num_tokens": 26184772.0,
      "step": 11950
    },
    {
      "entropy": 5.610151243209839,
      "epoch": 1.1864827312425565,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004867425444460782,
      "loss": 5.4258,
      "mean_token_accuracy": 0.1406768336892128,
      "num_tokens": 26195411.0,
      "step": 11955
    },
    {
      "entropy": 5.555606698989868,
      "epoch": 1.186978959904724,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004867305592552233,
      "loss": 5.3498,
      "mean_token_accuracy": 0.15423523485660554,
      "num_tokens": 26206719.0,
      "step": 11960
    },
    {
      "entropy": 5.544403505325318,
      "epoch": 1.1874751885668917,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004867185688139067,
      "loss": 5.2969,
      "mean_token_accuracy": 0.15522740483283998,
      "num_tokens": 26217469.0,
      "step": 11965
    },
    {
      "entropy": 5.486196660995484,
      "epoch": 1.1879714172290592,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004867065731224258,
      "loss": 5.3233,
      "mean_token_accuracy": 0.14972129017114638,
      "num_tokens": 26228556.0,
      "step": 11970
    },
    {
      "entropy": 5.577236890792847,
      "epoch": 1.1884676458912267,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004866945721810781,
      "loss": 5.4341,
      "mean_token_accuracy": 0.15076203942298888,
      "num_tokens": 26239891.0,
      "step": 11975
    },
    {
      "entropy": 5.5940775871276855,
      "epoch": 1.1889638745533941,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004866825659901612,
      "loss": 5.3759,
      "mean_token_accuracy": 0.15318043529987335,
      "num_tokens": 26251793.0,
      "step": 11980
    },
    {
      "entropy": 5.528149223327636,
      "epoch": 1.1894601032155618,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048667055454997276,
      "loss": 5.2716,
      "mean_token_accuracy": 0.15696460008621216,
      "num_tokens": 26262367.0,
      "step": 11985
    },
    {
      "entropy": 5.539589357376099,
      "epoch": 1.1899563318777293,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048665853786081063,
      "loss": 5.3693,
      "mean_token_accuracy": 0.14914708137512206,
      "num_tokens": 26272681.0,
      "step": 11990
    },
    {
      "entropy": 5.562288904190064,
      "epoch": 1.1904525605398968,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048664651592297294,
      "loss": 5.4227,
      "mean_token_accuracy": 0.14779862761497498,
      "num_tokens": 26281194.0,
      "step": 11995
    },
    {
      "entropy": 5.576343584060669,
      "epoch": 1.1909487892020643,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048663448873675766,
      "loss": 5.3219,
      "mean_token_accuracy": 0.14904801100492476,
      "num_tokens": 26292682.0,
      "step": 12000
    },
    {
      "epoch": 1.1909487892020643,
      "eval_entropy": 5.477951667801052,
      "eval_loss": 5.515817642211914,
      "eval_mean_token_accuracy": 0.14880079638684665,
      "eval_num_tokens": 26292682.0,
      "eval_runtime": 27.1971,
      "eval_samples_per_second": 1237.965,
      "eval_steps_per_second": 154.759,
      "step": 12000
    },
    {
      "entropy": 5.587806463241577,
      "epoch": 1.1914450178642317,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004866224563024631,
      "loss": 5.4154,
      "mean_token_accuracy": 0.14618875682353974,
      "num_tokens": 26305500.0,
      "step": 12005
    },
    {
      "entropy": 5.610121965408325,
      "epoch": 1.1919412465263994,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004866104186203877,
      "loss": 5.3567,
      "mean_token_accuracy": 0.14806900471448897,
      "num_tokens": 26317207.0,
      "step": 12010
    },
    {
      "entropy": 5.592868804931641,
      "epoch": 1.192437475188567,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048659837569082977,
      "loss": 5.3385,
      "mean_token_accuracy": 0.14877532720565795,
      "num_tokens": 26328728.0,
      "step": 12015
    },
    {
      "entropy": 5.60077223777771,
      "epoch": 1.1929337038507344,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004865863275140882,
      "loss": 5.3799,
      "mean_token_accuracy": 0.14791335463523864,
      "num_tokens": 26338622.0,
      "step": 12020
    },
    {
      "entropy": 5.534156656265258,
      "epoch": 1.1934299325129019,
      "grad_norm": 1.5,
      "learning_rate": 0.0004865742740904615,
      "loss": 5.3729,
      "mean_token_accuracy": 0.15156215131282808,
      "num_tokens": 26348807.0,
      "step": 12025
    },
    {
      "entropy": 5.530178642272949,
      "epoch": 1.1939261611750696,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004865622154202488,
      "loss": 5.3151,
      "mean_token_accuracy": 0.15594715774059295,
      "num_tokens": 26358481.0,
      "step": 12030
    },
    {
      "entropy": 5.597234058380127,
      "epoch": 1.194422389837237,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048655015150374904,
      "loss": 5.4021,
      "mean_token_accuracy": 0.14681665897369384,
      "num_tokens": 26370519.0,
      "step": 12035
    },
    {
      "entropy": 5.62939190864563,
      "epoch": 1.1949186184994045,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004865380823412613,
      "loss": 5.4041,
      "mean_token_accuracy": 0.15180511474609376,
      "num_tokens": 26380988.0,
      "step": 12040
    },
    {
      "entropy": 5.501454591751099,
      "epoch": 1.195414847161572,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000486526007933085,
      "loss": 5.302,
      "mean_token_accuracy": 0.1565579429268837,
      "num_tokens": 26390308.0,
      "step": 12045
    },
    {
      "entropy": 5.583420896530152,
      "epoch": 1.1959110758237395,
      "grad_norm": 1.375,
      "learning_rate": 0.00048651392827951944,
      "loss": 5.4158,
      "mean_token_accuracy": 0.14179497361183166,
      "num_tokens": 26401145.0,
      "step": 12050
    },
    {
      "entropy": 5.616935968399048,
      "epoch": 1.1964073044859072,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048650184338086436,
      "loss": 5.3447,
      "mean_token_accuracy": 0.15172840356826783,
      "num_tokens": 26411930.0,
      "step": 12055
    },
    {
      "entropy": 5.544000673294067,
      "epoch": 1.1969035331480746,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048648975323741916,
      "loss": 5.2949,
      "mean_token_accuracy": 0.14887586906552314,
      "num_tokens": 26422320.0,
      "step": 12060
    },
    {
      "entropy": 5.683514976501465,
      "epoch": 1.197399761810242,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048647765784948395,
      "loss": 5.4423,
      "mean_token_accuracy": 0.14140122830867768,
      "num_tokens": 26433760.0,
      "step": 12065
    },
    {
      "entropy": 5.649495410919189,
      "epoch": 1.1978959904724098,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004864655572173584,
      "loss": 5.3369,
      "mean_token_accuracy": 0.1514887697994709,
      "num_tokens": 26444256.0,
      "step": 12070
    },
    {
      "entropy": 5.55250391960144,
      "epoch": 1.1983922191345773,
      "grad_norm": 1.84375,
      "learning_rate": 0.00048645345134134274,
      "loss": 5.4372,
      "mean_token_accuracy": 0.1470012217760086,
      "num_tokens": 26456433.0,
      "step": 12075
    },
    {
      "entropy": 5.609834384918213,
      "epoch": 1.1988884477967447,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048644134022173707,
      "loss": 5.4135,
      "mean_token_accuracy": 0.1520857498049736,
      "num_tokens": 26466909.0,
      "step": 12080
    },
    {
      "entropy": 5.622609376907349,
      "epoch": 1.1993846764589122,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004864292238588419,
      "loss": 5.4035,
      "mean_token_accuracy": 0.1532752737402916,
      "num_tokens": 26478785.0,
      "step": 12085
    },
    {
      "entropy": 5.547451066970825,
      "epoch": 1.1998809051210797,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048641710225295746,
      "loss": 5.4128,
      "mean_token_accuracy": 0.15343670696020126,
      "num_tokens": 26489596.0,
      "step": 12090
    },
    {
      "entropy": 5.553436326980591,
      "epoch": 1.2003771337832474,
      "grad_norm": 1.5,
      "learning_rate": 0.00048640497540438455,
      "loss": 5.3484,
      "mean_token_accuracy": 0.14914562851190566,
      "num_tokens": 26500376.0,
      "step": 12095
    },
    {
      "entropy": 5.570246315002441,
      "epoch": 1.2008733624454149,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004863928433134237,
      "loss": 5.2775,
      "mean_token_accuracy": 0.1534338116645813,
      "num_tokens": 26510339.0,
      "step": 12100
    },
    {
      "entropy": 5.589764976501465,
      "epoch": 1.2013695911075823,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048638070598037585,
      "loss": 5.3693,
      "mean_token_accuracy": 0.14770288169384002,
      "num_tokens": 26521359.0,
      "step": 12105
    },
    {
      "entropy": 5.590108203887939,
      "epoch": 1.2018658197697498,
      "grad_norm": 1.640625,
      "learning_rate": 0.000486368563405542,
      "loss": 5.4543,
      "mean_token_accuracy": 0.14502543061971665,
      "num_tokens": 26532716.0,
      "step": 12110
    },
    {
      "entropy": 5.543052864074707,
      "epoch": 1.2023620484319175,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004863564155892232,
      "loss": 5.2917,
      "mean_token_accuracy": 0.15484083592891693,
      "num_tokens": 26542624.0,
      "step": 12115
    },
    {
      "entropy": 5.601221323013306,
      "epoch": 1.202858277094085,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004863442625317208,
      "loss": 5.409,
      "mean_token_accuracy": 0.15234583914279937,
      "num_tokens": 26552907.0,
      "step": 12120
    },
    {
      "entropy": 5.523341703414917,
      "epoch": 1.2033545057562525,
      "grad_norm": 1.828125,
      "learning_rate": 0.00048633210423333607,
      "loss": 5.3447,
      "mean_token_accuracy": 0.14861688166856765,
      "num_tokens": 26564871.0,
      "step": 12125
    },
    {
      "entropy": 5.571046972274781,
      "epoch": 1.20385073441842,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004863199406943705,
      "loss": 5.3478,
      "mean_token_accuracy": 0.1541406601667404,
      "num_tokens": 26576254.0,
      "step": 12130
    },
    {
      "entropy": 5.590707206726075,
      "epoch": 1.2043469630805874,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048630777191512584,
      "loss": 5.3996,
      "mean_token_accuracy": 0.15505435764789582,
      "num_tokens": 26587518.0,
      "step": 12135
    },
    {
      "entropy": 5.535691642761231,
      "epoch": 1.2048431917427551,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048629559789590366,
      "loss": 5.2948,
      "mean_token_accuracy": 0.15322870016098022,
      "num_tokens": 26597485.0,
      "step": 12140
    },
    {
      "entropy": 5.553314208984375,
      "epoch": 1.2053394204049226,
      "grad_norm": 3.21875,
      "learning_rate": 0.000486283418637006,
      "loss": 5.3348,
      "mean_token_accuracy": 0.15412704795598983,
      "num_tokens": 26607696.0,
      "step": 12145
    },
    {
      "entropy": 5.594108581542969,
      "epoch": 1.20583564906709,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004862712341387348,
      "loss": 5.3656,
      "mean_token_accuracy": 0.15496738106012345,
      "num_tokens": 26618974.0,
      "step": 12150
    },
    {
      "entropy": 5.626499128341675,
      "epoch": 1.2063318777292578,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004862590444013923,
      "loss": 5.4138,
      "mean_token_accuracy": 0.1505058452486992,
      "num_tokens": 26629949.0,
      "step": 12155
    },
    {
      "entropy": 5.631276178359985,
      "epoch": 1.2068281063914252,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004862468494252806,
      "loss": 5.3569,
      "mean_token_accuracy": 0.1449028730392456,
      "num_tokens": 26641460.0,
      "step": 12160
    },
    {
      "entropy": 5.640768718719483,
      "epoch": 1.2073243350535927,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004862346492107024,
      "loss": 5.4532,
      "mean_token_accuracy": 0.14615924656391144,
      "num_tokens": 26652553.0,
      "step": 12165
    },
    {
      "entropy": 5.518619775772095,
      "epoch": 1.2078205637157602,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00048622244375796,
      "loss": 5.352,
      "mean_token_accuracy": 0.14533724561333655,
      "num_tokens": 26662479.0,
      "step": 12170
    },
    {
      "entropy": 5.587811136245728,
      "epoch": 1.2083167923779277,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004862102330673561,
      "loss": 5.386,
      "mean_token_accuracy": 0.14768699705600738,
      "num_tokens": 26673326.0,
      "step": 12175
    },
    {
      "entropy": 5.626950788497925,
      "epoch": 1.2088130210400954,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048619801713919357,
      "loss": 5.4589,
      "mean_token_accuracy": 0.14376690834760666,
      "num_tokens": 26684189.0,
      "step": 12180
    },
    {
      "entropy": 5.651653623580932,
      "epoch": 1.2093092497022628,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004861857959737753,
      "loss": 5.4459,
      "mean_token_accuracy": 0.14807091727852822,
      "num_tokens": 26695399.0,
      "step": 12185
    },
    {
      "entropy": 5.652006149291992,
      "epoch": 1.2098054783644303,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004861735695714044,
      "loss": 5.4168,
      "mean_token_accuracy": 0.15000323355197906,
      "num_tokens": 26705993.0,
      "step": 12190
    },
    {
      "entropy": 5.619342565536499,
      "epoch": 1.2103017070265978,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048616133793238394,
      "loss": 5.3672,
      "mean_token_accuracy": 0.15637729018926622,
      "num_tokens": 26716120.0,
      "step": 12195
    },
    {
      "entropy": 5.607169675827026,
      "epoch": 1.2107979356887655,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048614910105701737,
      "loss": 5.4937,
      "mean_token_accuracy": 0.14124025106430055,
      "num_tokens": 26727736.0,
      "step": 12200
    },
    {
      "entropy": 5.657632350921631,
      "epoch": 1.211294164350933,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004861368589456081,
      "loss": 5.3536,
      "mean_token_accuracy": 0.14946447759866716,
      "num_tokens": 26737825.0,
      "step": 12205
    },
    {
      "entropy": 5.70032148361206,
      "epoch": 1.2117903930131004,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004861246115984596,
      "loss": 5.4297,
      "mean_token_accuracy": 0.1422755479812622,
      "num_tokens": 26748035.0,
      "step": 12210
    },
    {
      "entropy": 5.590206670761108,
      "epoch": 1.212286621675268,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048611235901587575,
      "loss": 5.3538,
      "mean_token_accuracy": 0.15279785841703414,
      "num_tokens": 26759899.0,
      "step": 12215
    },
    {
      "entropy": 5.508137512207031,
      "epoch": 1.2127828503374354,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004861001011981602,
      "loss": 5.2892,
      "mean_token_accuracy": 0.1552995726466179,
      "num_tokens": 26771562.0,
      "step": 12220
    },
    {
      "entropy": 5.632553005218506,
      "epoch": 1.213279078999603,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004860878381456171,
      "loss": 5.4515,
      "mean_token_accuracy": 0.15011319816112517,
      "num_tokens": 26783002.0,
      "step": 12225
    },
    {
      "entropy": 5.611722421646118,
      "epoch": 1.2137753076617706,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048607556985855047,
      "loss": 5.4221,
      "mean_token_accuracy": 0.14730474203824998,
      "num_tokens": 26794136.0,
      "step": 12230
    },
    {
      "entropy": 5.522471809387207,
      "epoch": 1.214271536323938,
      "grad_norm": 2.015625,
      "learning_rate": 0.00048606329633726447,
      "loss": 5.3906,
      "mean_token_accuracy": 0.14802951663732528,
      "num_tokens": 26804776.0,
      "step": 12235
    },
    {
      "entropy": 5.548605585098267,
      "epoch": 1.2147677649861055,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004860510175820636,
      "loss": 5.324,
      "mean_token_accuracy": 0.15403158813714982,
      "num_tokens": 26816696.0,
      "step": 12240
    },
    {
      "entropy": 5.5903205394744875,
      "epoch": 1.2152639936482732,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004860387335932522,
      "loss": 5.3595,
      "mean_token_accuracy": 0.15180681943893432,
      "num_tokens": 26827789.0,
      "step": 12245
    },
    {
      "entropy": 5.597366571426392,
      "epoch": 1.2157602223104407,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004860264443711351,
      "loss": 5.2997,
      "mean_token_accuracy": 0.15480596274137498,
      "num_tokens": 26838763.0,
      "step": 12250
    },
    {
      "entropy": 5.483070611953735,
      "epoch": 1.2162564509726081,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004860141499160167,
      "loss": 5.3519,
      "mean_token_accuracy": 0.1431449681520462,
      "num_tokens": 26849304.0,
      "step": 12255
    },
    {
      "entropy": 5.562550401687622,
      "epoch": 1.2167526796347756,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004860018502282022,
      "loss": 5.4157,
      "mean_token_accuracy": 0.15046968311071396,
      "num_tokens": 26858777.0,
      "step": 12260
    },
    {
      "entropy": 5.648476648330688,
      "epoch": 1.2172489082969433,
      "grad_norm": 2.0,
      "learning_rate": 0.00048598954530799644,
      "loss": 5.3702,
      "mean_token_accuracy": 0.14573246613144875,
      "num_tokens": 26870286.0,
      "step": 12265
    },
    {
      "entropy": 5.622893047332764,
      "epoch": 1.2177451369591108,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048597723515570465,
      "loss": 5.3485,
      "mean_token_accuracy": 0.16040409356355667,
      "num_tokens": 26880485.0,
      "step": 12270
    },
    {
      "entropy": 5.532984113693237,
      "epoch": 1.2182413656212783,
      "grad_norm": 1.25,
      "learning_rate": 0.000485964919771632,
      "loss": 5.2847,
      "mean_token_accuracy": 0.1575680509209633,
      "num_tokens": 26893556.0,
      "step": 12275
    },
    {
      "entropy": 5.528851175308228,
      "epoch": 1.2187375942834457,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048595259915608396,
      "loss": 5.3425,
      "mean_token_accuracy": 0.14584220945835114,
      "num_tokens": 26905699.0,
      "step": 12280
    },
    {
      "entropy": 5.547272157669068,
      "epoch": 1.2192338229456134,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048594027330936594,
      "loss": 5.3684,
      "mean_token_accuracy": 0.14541950225830078,
      "num_tokens": 26917188.0,
      "step": 12285
    },
    {
      "entropy": 5.581481266021728,
      "epoch": 1.219730051607781,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048592794223178374,
      "loss": 5.3554,
      "mean_token_accuracy": 0.15071195960044861,
      "num_tokens": 26927905.0,
      "step": 12290
    },
    {
      "entropy": 5.503611612319946,
      "epoch": 1.2202262802699484,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048591560592364314,
      "loss": 5.3187,
      "mean_token_accuracy": 0.15694567561149597,
      "num_tokens": 26938665.0,
      "step": 12295
    },
    {
      "entropy": 5.566227912902832,
      "epoch": 1.2207225089321159,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004859032643852499,
      "loss": 5.4407,
      "mean_token_accuracy": 0.14990585446357726,
      "num_tokens": 26950516.0,
      "step": 12300
    },
    {
      "entropy": 5.670513916015625,
      "epoch": 1.2212187375942833,
      "grad_norm": 1.375,
      "learning_rate": 0.0004858909176169102,
      "loss": 5.4275,
      "mean_token_accuracy": 0.15341316908597946,
      "num_tokens": 26960018.0,
      "step": 12305
    },
    {
      "entropy": 5.581910896301269,
      "epoch": 1.221714966256451,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048587856561893014,
      "loss": 5.3784,
      "mean_token_accuracy": 0.1519193947315216,
      "num_tokens": 26969825.0,
      "step": 12310
    },
    {
      "entropy": 5.589643478393555,
      "epoch": 1.2222111949186185,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004858662083916161,
      "loss": 5.4759,
      "mean_token_accuracy": 0.14708292856812477,
      "num_tokens": 26979873.0,
      "step": 12315
    },
    {
      "entropy": 5.634253454208374,
      "epoch": 1.222707423580786,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004858538459352744,
      "loss": 5.4526,
      "mean_token_accuracy": 0.1475948765873909,
      "num_tokens": 26990577.0,
      "step": 12320
    },
    {
      "entropy": 5.569654130935669,
      "epoch": 1.2232036522429535,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004858414782502117,
      "loss": 5.281,
      "mean_token_accuracy": 0.15585556179285048,
      "num_tokens": 27001598.0,
      "step": 12325
    },
    {
      "entropy": 5.536151504516601,
      "epoch": 1.2236998809051212,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048582910533673474,
      "loss": 5.303,
      "mean_token_accuracy": 0.1553460881114006,
      "num_tokens": 27012535.0,
      "step": 12330
    },
    {
      "entropy": 5.533866548538208,
      "epoch": 1.2241961095672886,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004858167271951502,
      "loss": 5.3669,
      "mean_token_accuracy": 0.1421804942190647,
      "num_tokens": 27021867.0,
      "step": 12335
    },
    {
      "entropy": 5.500246238708496,
      "epoch": 1.2246923382294561,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004858043438257651,
      "loss": 5.3392,
      "mean_token_accuracy": 0.15250529795885087,
      "num_tokens": 27032540.0,
      "step": 12340
    },
    {
      "entropy": 5.491591691970825,
      "epoch": 1.2251885668916236,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004857919552288865,
      "loss": 5.326,
      "mean_token_accuracy": 0.15734471678733825,
      "num_tokens": 27044008.0,
      "step": 12345
    },
    {
      "entropy": 5.621547031402588,
      "epoch": 1.225684795553791,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004857795614048217,
      "loss": 5.3313,
      "mean_token_accuracy": 0.15750225335359574,
      "num_tokens": 27053646.0,
      "step": 12350
    },
    {
      "entropy": 5.6558550834655765,
      "epoch": 1.2261810242159588,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048576716235387786,
      "loss": 5.4849,
      "mean_token_accuracy": 0.14320140779018403,
      "num_tokens": 27065834.0,
      "step": 12355
    },
    {
      "entropy": 5.552185964584351,
      "epoch": 1.2266772528781262,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048575475807636267,
      "loss": 5.416,
      "mean_token_accuracy": 0.14511745572090148,
      "num_tokens": 27077544.0,
      "step": 12360
    },
    {
      "entropy": 5.649219942092896,
      "epoch": 1.2271734815402937,
      "grad_norm": 2.015625,
      "learning_rate": 0.00048574234857258356,
      "loss": 5.4507,
      "mean_token_accuracy": 0.15687849074602128,
      "num_tokens": 27088294.0,
      "step": 12365
    },
    {
      "entropy": 5.629151821136475,
      "epoch": 1.2276697102024614,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00048572993384284833,
      "loss": 5.3963,
      "mean_token_accuracy": 0.1429994560778141,
      "num_tokens": 27098848.0,
      "step": 12370
    },
    {
      "entropy": 5.536941242218018,
      "epoch": 1.2281659388646289,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004857175138874648,
      "loss": 5.3588,
      "mean_token_accuracy": 0.15086302012205124,
      "num_tokens": 27112319.0,
      "step": 12375
    },
    {
      "entropy": 5.5927516460418705,
      "epoch": 1.2286621675267964,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048570508870674097,
      "loss": 5.3989,
      "mean_token_accuracy": 0.15069908946752547,
      "num_tokens": 27121840.0,
      "step": 12380
    },
    {
      "entropy": 5.5176843166351315,
      "epoch": 1.2291583961889638,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048569265830098493,
      "loss": 5.407,
      "mean_token_accuracy": 0.14835759103298188,
      "num_tokens": 27133619.0,
      "step": 12385
    },
    {
      "entropy": 5.587302207946777,
      "epoch": 1.2296546248511313,
      "grad_norm": 1.296875,
      "learning_rate": 0.000485680222670505,
      "loss": 5.3844,
      "mean_token_accuracy": 0.14568580985069274,
      "num_tokens": 27144161.0,
      "step": 12390
    },
    {
      "entropy": 5.556478261947632,
      "epoch": 1.230150853513299,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004856677818156096,
      "loss": 5.3138,
      "mean_token_accuracy": 0.15167154297232627,
      "num_tokens": 27154906.0,
      "step": 12395
    },
    {
      "entropy": 5.5445434093475345,
      "epoch": 1.2306470821754665,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048565533573660707,
      "loss": 5.3406,
      "mean_token_accuracy": 0.1518824502825737,
      "num_tokens": 27165746.0,
      "step": 12400
    },
    {
      "entropy": 5.540196323394776,
      "epoch": 1.231143310837634,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004856428844338062,
      "loss": 5.3283,
      "mean_token_accuracy": 0.15534317418932914,
      "num_tokens": 27176108.0,
      "step": 12405
    },
    {
      "entropy": 5.574157428741455,
      "epoch": 1.2316395394998014,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004856304279075156,
      "loss": 5.4039,
      "mean_token_accuracy": 0.146202152967453,
      "num_tokens": 27186721.0,
      "step": 12410
    },
    {
      "entropy": 5.5774956226348875,
      "epoch": 1.2321357681619691,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004856179661580444,
      "loss": 5.2752,
      "mean_token_accuracy": 0.15339501947164536,
      "num_tokens": 27196120.0,
      "step": 12415
    },
    {
      "entropy": 5.603684425354004,
      "epoch": 1.2326319968241366,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004856054991857013,
      "loss": 5.4042,
      "mean_token_accuracy": 0.14328193068504333,
      "num_tokens": 27207431.0,
      "step": 12420
    },
    {
      "entropy": 5.5260881900787355,
      "epoch": 1.233128225486304,
      "grad_norm": 1.625,
      "learning_rate": 0.00048559302699079576,
      "loss": 5.2844,
      "mean_token_accuracy": 0.1602466180920601,
      "num_tokens": 27218489.0,
      "step": 12425
    },
    {
      "entropy": 5.611947250366211,
      "epoch": 1.2336244541484715,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004855805495736369,
      "loss": 5.4263,
      "mean_token_accuracy": 0.1477315753698349,
      "num_tokens": 27230849.0,
      "step": 12430
    },
    {
      "entropy": 5.600693464279175,
      "epoch": 1.234120682810639,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004855680669345342,
      "loss": 5.3684,
      "mean_token_accuracy": 0.1476411372423172,
      "num_tokens": 27242838.0,
      "step": 12435
    },
    {
      "entropy": 5.582112216949463,
      "epoch": 1.2346169114728067,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048555557907379714,
      "loss": 5.3974,
      "mean_token_accuracy": 0.15531554520130159,
      "num_tokens": 27253945.0,
      "step": 12440
    },
    {
      "entropy": 5.5815016746521,
      "epoch": 1.2351131401349742,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048554308599173544,
      "loss": 5.4576,
      "mean_token_accuracy": 0.1404146134853363,
      "num_tokens": 27263999.0,
      "step": 12445
    },
    {
      "entropy": 5.6622405529022215,
      "epoch": 1.2356093687971417,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004855305876886589,
      "loss": 5.3005,
      "mean_token_accuracy": 0.15401808097958564,
      "num_tokens": 27274677.0,
      "step": 12450
    },
    {
      "entropy": 5.570734930038452,
      "epoch": 1.2361055974593094,
      "grad_norm": 1.6875,
      "learning_rate": 0.00048551808416487743,
      "loss": 5.4008,
      "mean_token_accuracy": 0.14744462072849274,
      "num_tokens": 27286512.0,
      "step": 12455
    },
    {
      "entropy": 5.534234857559204,
      "epoch": 1.2366018261214768,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004855055754207011,
      "loss": 5.3817,
      "mean_token_accuracy": 0.15207858383655548,
      "num_tokens": 27297620.0,
      "step": 12460
    },
    {
      "entropy": 5.615413236618042,
      "epoch": 1.2370980547836443,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004854930614564401,
      "loss": 5.3964,
      "mean_token_accuracy": 0.1485779881477356,
      "num_tokens": 27309963.0,
      "step": 12465
    },
    {
      "entropy": 5.6588225841522215,
      "epoch": 1.2375942834458118,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004854805422724047,
      "loss": 5.4929,
      "mean_token_accuracy": 0.14764803946018218,
      "num_tokens": 27320740.0,
      "step": 12470
    },
    {
      "entropy": 5.60316276550293,
      "epoch": 1.2380905121079793,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004854680178689055,
      "loss": 5.3169,
      "mean_token_accuracy": 0.1568467453122139,
      "num_tokens": 27331543.0,
      "step": 12475
    },
    {
      "entropy": 5.550124645233154,
      "epoch": 1.238586740770147,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048545548824625287,
      "loss": 5.31,
      "mean_token_accuracy": 0.14904602468013764,
      "num_tokens": 27343222.0,
      "step": 12480
    },
    {
      "entropy": 5.5475719451904295,
      "epoch": 1.2390829694323144,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004854429534047577,
      "loss": 5.3733,
      "mean_token_accuracy": 0.150028857588768,
      "num_tokens": 27354694.0,
      "step": 12485
    },
    {
      "entropy": 5.513445472717285,
      "epoch": 1.239579198094482,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004854304133447307,
      "loss": 5.3399,
      "mean_token_accuracy": 0.1475573718547821,
      "num_tokens": 27365462.0,
      "step": 12490
    },
    {
      "entropy": 5.54975643157959,
      "epoch": 1.2400754267566494,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048541786806648293,
      "loss": 5.2629,
      "mean_token_accuracy": 0.15528396368026734,
      "num_tokens": 27377548.0,
      "step": 12495
    },
    {
      "entropy": 5.608302402496338,
      "epoch": 1.240571655418817,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004854053175703254,
      "loss": 5.3925,
      "mean_token_accuracy": 0.1487332336604595,
      "num_tokens": 27388221.0,
      "step": 12500
    },
    {
      "entropy": 5.5642454624176025,
      "epoch": 1.2410678840809846,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004853927618565694,
      "loss": 5.394,
      "mean_token_accuracy": 0.14612849056720734,
      "num_tokens": 27400497.0,
      "step": 12505
    },
    {
      "entropy": 5.544233322143555,
      "epoch": 1.241564112743152,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048538020092552626,
      "loss": 5.2941,
      "mean_token_accuracy": 0.1519656203687191,
      "num_tokens": 27410700.0,
      "step": 12510
    },
    {
      "entropy": 5.6218372821807865,
      "epoch": 1.2420603414053195,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048536763477750747,
      "loss": 5.411,
      "mean_token_accuracy": 0.14697907269001007,
      "num_tokens": 27421286.0,
      "step": 12515
    },
    {
      "entropy": 5.623745918273926,
      "epoch": 1.242556570067487,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004853550634128246,
      "loss": 5.4406,
      "mean_token_accuracy": 0.1489515095949173,
      "num_tokens": 27433922.0,
      "step": 12520
    },
    {
      "entropy": 5.547319984436035,
      "epoch": 1.2430527987296547,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004853424868317895,
      "loss": 5.3522,
      "mean_token_accuracy": 0.14944472014904023,
      "num_tokens": 27444684.0,
      "step": 12525
    },
    {
      "entropy": 5.550248289108277,
      "epoch": 1.2435490273918222,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004853299050347139,
      "loss": 5.3279,
      "mean_token_accuracy": 0.15573048144578933,
      "num_tokens": 27455494.0,
      "step": 12530
    },
    {
      "entropy": 5.612189912796021,
      "epoch": 1.2440452560539896,
      "grad_norm": 1.625,
      "learning_rate": 0.00048531731802190994,
      "loss": 5.3794,
      "mean_token_accuracy": 0.14386983811855317,
      "num_tokens": 27464927.0,
      "step": 12535
    },
    {
      "entropy": 5.62739896774292,
      "epoch": 1.244541484716157,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004853047257936897,
      "loss": 5.3843,
      "mean_token_accuracy": 0.15132649689912797,
      "num_tokens": 27476228.0,
      "step": 12540
    },
    {
      "entropy": 5.583481121063232,
      "epoch": 1.2450377133783248,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004852921283503653,
      "loss": 5.3838,
      "mean_token_accuracy": 0.1507735699415207,
      "num_tokens": 27488835.0,
      "step": 12545
    },
    {
      "entropy": 5.550186681747436,
      "epoch": 1.2455339420404923,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004852795256922494,
      "loss": 5.3576,
      "mean_token_accuracy": 0.14997979253530502,
      "num_tokens": 27501481.0,
      "step": 12550
    },
    {
      "entropy": 5.5780397891998295,
      "epoch": 1.2460301707026598,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048526691781965425,
      "loss": 5.3739,
      "mean_token_accuracy": 0.1497849017381668,
      "num_tokens": 27512234.0,
      "step": 12555
    },
    {
      "entropy": 5.588928508758545,
      "epoch": 1.2465263993648272,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048525430473289274,
      "loss": 5.4201,
      "mean_token_accuracy": 0.14770554825663568,
      "num_tokens": 27523756.0,
      "step": 12560
    },
    {
      "entropy": 5.597650671005249,
      "epoch": 1.247022628026995,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048524168643227747,
      "loss": 5.4122,
      "mean_token_accuracy": 0.14830739721655845,
      "num_tokens": 27534579.0,
      "step": 12565
    },
    {
      "entropy": 5.625125312805176,
      "epoch": 1.2475188566891624,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048522906291812145,
      "loss": 5.4633,
      "mean_token_accuracy": 0.14512896537780762,
      "num_tokens": 27546420.0,
      "step": 12570
    },
    {
      "entropy": 5.5710217475891115,
      "epoch": 1.2480150853513299,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048521643419073763,
      "loss": 5.3333,
      "mean_token_accuracy": 0.15216598510742188,
      "num_tokens": 27557701.0,
      "step": 12575
    },
    {
      "entropy": 5.482281160354614,
      "epoch": 1.2485113140134974,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004852038002504392,
      "loss": 5.3421,
      "mean_token_accuracy": 0.1576241783797741,
      "num_tokens": 27568592.0,
      "step": 12580
    },
    {
      "entropy": 5.57126293182373,
      "epoch": 1.249007542675665,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048519116109753955,
      "loss": 5.385,
      "mean_token_accuracy": 0.15510101318359376,
      "num_tokens": 27579286.0,
      "step": 12585
    },
    {
      "entropy": 5.582639455795288,
      "epoch": 1.2495037713378325,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048517851673235196,
      "loss": 5.334,
      "mean_token_accuracy": 0.1521566703915596,
      "num_tokens": 27590031.0,
      "step": 12590
    },
    {
      "entropy": 5.5643918991088865,
      "epoch": 1.25,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004851658671551901,
      "loss": 5.3728,
      "mean_token_accuracy": 0.14801832139492035,
      "num_tokens": 27600999.0,
      "step": 12595
    },
    {
      "entropy": 5.532628917694092,
      "epoch": 1.2504962286621675,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048515321236636754,
      "loss": 5.3543,
      "mean_token_accuracy": 0.1470926284790039,
      "num_tokens": 27613394.0,
      "step": 12600
    },
    {
      "entropy": 5.592876863479614,
      "epoch": 1.250992457324335,
      "grad_norm": 2.0,
      "learning_rate": 0.0004851405523661982,
      "loss": 5.3338,
      "mean_token_accuracy": 0.15926629453897476,
      "num_tokens": 27624819.0,
      "step": 12605
    },
    {
      "entropy": 5.512431812286377,
      "epoch": 1.2514886859865026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048512788715499597,
      "loss": 5.3659,
      "mean_token_accuracy": 0.155659019947052,
      "num_tokens": 27635993.0,
      "step": 12610
    },
    {
      "entropy": 5.565937137603759,
      "epoch": 1.2519849146486701,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004851152167330749,
      "loss": 5.4211,
      "mean_token_accuracy": 0.14991864636540414,
      "num_tokens": 27647298.0,
      "step": 12615
    },
    {
      "entropy": 5.636095094680786,
      "epoch": 1.2524811433108376,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048510254110074923,
      "loss": 5.3996,
      "mean_token_accuracy": 0.14807143062353134,
      "num_tokens": 27659381.0,
      "step": 12620
    },
    {
      "entropy": 5.604834985733032,
      "epoch": 1.2529773719730053,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004850898602583333,
      "loss": 5.39,
      "mean_token_accuracy": 0.14930998384952546,
      "num_tokens": 27671122.0,
      "step": 12625
    },
    {
      "entropy": 5.625185823440551,
      "epoch": 1.2534736006351728,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048507717420614153,
      "loss": 5.3823,
      "mean_token_accuracy": 0.15360722541809083,
      "num_tokens": 27683258.0,
      "step": 12630
    },
    {
      "entropy": 5.476237058639526,
      "epoch": 1.2539698292973402,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004850644829444884,
      "loss": 5.2433,
      "mean_token_accuracy": 0.15188805013895035,
      "num_tokens": 27693116.0,
      "step": 12635
    },
    {
      "entropy": 5.509416103363037,
      "epoch": 1.2544660579595077,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048505178647368887,
      "loss": 5.3513,
      "mean_token_accuracy": 0.15534612387418748,
      "num_tokens": 27704859.0,
      "step": 12640
    },
    {
      "entropy": 5.545328140258789,
      "epoch": 1.2549622866216752,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004850390847940576,
      "loss": 5.2851,
      "mean_token_accuracy": 0.15771727859973908,
      "num_tokens": 27715854.0,
      "step": 12645
    },
    {
      "entropy": 5.5508997440338135,
      "epoch": 1.2554585152838427,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004850263779059097,
      "loss": 5.3262,
      "mean_token_accuracy": 0.1470785692334175,
      "num_tokens": 27726398.0,
      "step": 12650
    },
    {
      "entropy": 5.568870878219604,
      "epoch": 1.2559547439460104,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004850136658095601,
      "loss": 5.3573,
      "mean_token_accuracy": 0.15258002579212188,
      "num_tokens": 27737337.0,
      "step": 12655
    },
    {
      "entropy": 5.603633832931519,
      "epoch": 1.2564509726081778,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048500094850532417,
      "loss": 5.3965,
      "mean_token_accuracy": 0.1516628161072731,
      "num_tokens": 27748593.0,
      "step": 12660
    },
    {
      "entropy": 5.584478712081909,
      "epoch": 1.2569472012703453,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004849882259935172,
      "loss": 5.3674,
      "mean_token_accuracy": 0.1502511218190193,
      "num_tokens": 27760908.0,
      "step": 12665
    },
    {
      "entropy": 5.577605247497559,
      "epoch": 1.257443429932513,
      "grad_norm": 1.5,
      "learning_rate": 0.0004849754982744548,
      "loss": 5.44,
      "mean_token_accuracy": 0.14887894988059996,
      "num_tokens": 27772104.0,
      "step": 12670
    },
    {
      "entropy": 5.53955569267273,
      "epoch": 1.2579396585946805,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004849627653484524,
      "loss": 5.3547,
      "mean_token_accuracy": 0.154725394397974,
      "num_tokens": 27781654.0,
      "step": 12675
    },
    {
      "entropy": 5.588595533370972,
      "epoch": 1.258435887256848,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004849500272158259,
      "loss": 5.3654,
      "mean_token_accuracy": 0.15242195650935172,
      "num_tokens": 27793026.0,
      "step": 12680
    },
    {
      "entropy": 5.591387462615967,
      "epoch": 1.2589321159190154,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004849372838768911,
      "loss": 5.3325,
      "mean_token_accuracy": 0.15427452176809311,
      "num_tokens": 27803447.0,
      "step": 12685
    },
    {
      "entropy": 5.638734149932861,
      "epoch": 1.259428344581183,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000484924535331964,
      "loss": 5.3501,
      "mean_token_accuracy": 0.15045448541641235,
      "num_tokens": 27814632.0,
      "step": 12690
    },
    {
      "entropy": 5.58971905708313,
      "epoch": 1.2599245732433506,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048491178158136086,
      "loss": 5.4333,
      "mean_token_accuracy": 0.14772067219018936,
      "num_tokens": 27826582.0,
      "step": 12695
    },
    {
      "entropy": 5.550472354888916,
      "epoch": 1.260420801905518,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004848990226253978,
      "loss": 5.2914,
      "mean_token_accuracy": 0.15210264027118683,
      "num_tokens": 27838065.0,
      "step": 12700
    },
    {
      "entropy": 5.490083456039429,
      "epoch": 1.2609170305676856,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048488625846439135,
      "loss": 5.3276,
      "mean_token_accuracy": 0.14947343468666077,
      "num_tokens": 27848182.0,
      "step": 12705
    },
    {
      "entropy": 5.613709783554077,
      "epoch": 1.261413259229853,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004848734890986578,
      "loss": 5.3583,
      "mean_token_accuracy": 0.14490787014365197,
      "num_tokens": 27858524.0,
      "step": 12710
    },
    {
      "entropy": 5.509617137908935,
      "epoch": 1.2619094878920207,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004848607145285141,
      "loss": 5.2609,
      "mean_token_accuracy": 0.15548285990953445,
      "num_tokens": 27868588.0,
      "step": 12715
    },
    {
      "entropy": 5.491306161880493,
      "epoch": 1.2624057165541882,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004848479347542768,
      "loss": 5.2816,
      "mean_token_accuracy": 0.15655839145183564,
      "num_tokens": 27878829.0,
      "step": 12720
    },
    {
      "entropy": 5.55536298751831,
      "epoch": 1.2629019452163557,
      "grad_norm": 1.21875,
      "learning_rate": 0.000484835149776263,
      "loss": 5.3579,
      "mean_token_accuracy": 0.1579335570335388,
      "num_tokens": 27889872.0,
      "step": 12725
    },
    {
      "entropy": 5.591164493560791,
      "epoch": 1.2633981738785232,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048482235959478953,
      "loss": 5.4294,
      "mean_token_accuracy": 0.15196538120508193,
      "num_tokens": 27900567.0,
      "step": 12730
    },
    {
      "entropy": 5.6184333801269535,
      "epoch": 1.2638944025406906,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004848095642101737,
      "loss": 5.3001,
      "mean_token_accuracy": 0.15098789483308792,
      "num_tokens": 27911594.0,
      "step": 12735
    },
    {
      "entropy": 5.616883754730225,
      "epoch": 1.2643906312028583,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00048479676362273277,
      "loss": 5.4113,
      "mean_token_accuracy": 0.1427378311753273,
      "num_tokens": 27923694.0,
      "step": 12740
    },
    {
      "entropy": 5.553993844985962,
      "epoch": 1.2648868598650258,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048478395783278425,
      "loss": 5.4746,
      "mean_token_accuracy": 0.13807003796100617,
      "num_tokens": 27934168.0,
      "step": 12745
    },
    {
      "entropy": 5.619164896011353,
      "epoch": 1.2653830885271933,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004847711468406455,
      "loss": 5.375,
      "mean_token_accuracy": 0.1494707092642784,
      "num_tokens": 27944401.0,
      "step": 12750
    },
    {
      "entropy": 5.568649816513061,
      "epoch": 1.265879317189361,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048475833064663443,
      "loss": 5.265,
      "mean_token_accuracy": 0.158959199488163,
      "num_tokens": 27956219.0,
      "step": 12755
    },
    {
      "entropy": 5.56947603225708,
      "epoch": 1.2663755458515285,
      "grad_norm": 4.3125,
      "learning_rate": 0.0004847455092510687,
      "loss": 5.369,
      "mean_token_accuracy": 0.15354631021618842,
      "num_tokens": 27967457.0,
      "step": 12760
    },
    {
      "entropy": 5.5270898818969725,
      "epoch": 1.266871774513696,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004847326826542663,
      "loss": 5.3734,
      "mean_token_accuracy": 0.14961889535188674,
      "num_tokens": 27978287.0,
      "step": 12765
    },
    {
      "entropy": 5.603542995452881,
      "epoch": 1.2673680031758634,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048471985085654537,
      "loss": 5.3919,
      "mean_token_accuracy": 0.15767520517110825,
      "num_tokens": 27989703.0,
      "step": 12770
    },
    {
      "entropy": 5.572586631774902,
      "epoch": 1.2678642318380309,
      "grad_norm": 1.84375,
      "learning_rate": 0.00048470701385822394,
      "loss": 5.3635,
      "mean_token_accuracy": 0.15158511102199554,
      "num_tokens": 28002071.0,
      "step": 12775
    },
    {
      "entropy": 5.606336069107056,
      "epoch": 1.2683604605001986,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048469417165962047,
      "loss": 5.4325,
      "mean_token_accuracy": 0.1479561969637871,
      "num_tokens": 28014517.0,
      "step": 12780
    },
    {
      "entropy": 5.574123907089233,
      "epoch": 1.268856689162366,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004846813242610535,
      "loss": 5.38,
      "mean_token_accuracy": 0.1551190659403801,
      "num_tokens": 28025135.0,
      "step": 12785
    },
    {
      "entropy": 5.539662218093872,
      "epoch": 1.2693529178245335,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004846684716628414,
      "loss": 5.3009,
      "mean_token_accuracy": 0.15886031836271286,
      "num_tokens": 28036105.0,
      "step": 12790
    },
    {
      "entropy": 5.554424667358399,
      "epoch": 1.269849146486701,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000484655613865303,
      "loss": 5.3679,
      "mean_token_accuracy": 0.1523560330271721,
      "num_tokens": 28048825.0,
      "step": 12795
    },
    {
      "entropy": 5.5978978157043455,
      "epoch": 1.2703453751488687,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048464275086875717,
      "loss": 5.3178,
      "mean_token_accuracy": 0.15005845725536346,
      "num_tokens": 28060408.0,
      "step": 12800
    },
    {
      "entropy": 5.593704175949097,
      "epoch": 1.2708416038110362,
      "grad_norm": 1.375,
      "learning_rate": 0.0004846298826735229,
      "loss": 5.3889,
      "mean_token_accuracy": 0.14791881889104844,
      "num_tokens": 28072179.0,
      "step": 12805
    },
    {
      "entropy": 5.569877481460571,
      "epoch": 1.2713378324732036,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004846170092799192,
      "loss": 5.3552,
      "mean_token_accuracy": 0.14534294605255127,
      "num_tokens": 28083825.0,
      "step": 12810
    },
    {
      "entropy": 5.741600751876831,
      "epoch": 1.2718340611353711,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048460413068826536,
      "loss": 5.508,
      "mean_token_accuracy": 0.14574243500828743,
      "num_tokens": 28095136.0,
      "step": 12815
    },
    {
      "entropy": 5.599288702011108,
      "epoch": 1.2723302897975386,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004845912468988807,
      "loss": 5.3261,
      "mean_token_accuracy": 0.14950667023658754,
      "num_tokens": 28104763.0,
      "step": 12820
    },
    {
      "entropy": 5.552009057998657,
      "epoch": 1.2728265184597063,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004845783579120849,
      "loss": 5.3856,
      "mean_token_accuracy": 0.15191560685634614,
      "num_tokens": 28115771.0,
      "step": 12825
    },
    {
      "entropy": 5.501192665100097,
      "epoch": 1.2733227471218738,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004845654637281973,
      "loss": 5.3166,
      "mean_token_accuracy": 0.14955545514822005,
      "num_tokens": 28128039.0,
      "step": 12830
    },
    {
      "entropy": 5.5299300193786625,
      "epoch": 1.2738189757840412,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048455256434753774,
      "loss": 5.325,
      "mean_token_accuracy": 0.15650611370801926,
      "num_tokens": 28139163.0,
      "step": 12835
    },
    {
      "entropy": 5.555418539047241,
      "epoch": 1.274315204446209,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004845396597704262,
      "loss": 5.3599,
      "mean_token_accuracy": 0.14783878922462462,
      "num_tokens": 28150625.0,
      "step": 12840
    },
    {
      "entropy": 5.602498579025268,
      "epoch": 1.2748114331083764,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048452674999718257,
      "loss": 5.4475,
      "mean_token_accuracy": 0.149952432513237,
      "num_tokens": 28163072.0,
      "step": 12845
    },
    {
      "entropy": 5.640607500076294,
      "epoch": 1.2753076617705439,
      "grad_norm": 1.375,
      "learning_rate": 0.00048451383502812705,
      "loss": 5.4596,
      "mean_token_accuracy": 0.14492754265666008,
      "num_tokens": 28174192.0,
      "step": 12850
    },
    {
      "entropy": 5.531475019454956,
      "epoch": 1.2758038904327114,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048450091486357987,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1590728521347046,
      "num_tokens": 28184477.0,
      "step": 12855
    },
    {
      "entropy": 5.524982738494873,
      "epoch": 1.2763001190948788,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048448798950386147,
      "loss": 5.2436,
      "mean_token_accuracy": 0.15408096313476563,
      "num_tokens": 28194552.0,
      "step": 12860
    },
    {
      "entropy": 5.5430745601654055,
      "epoch": 1.2767963477570463,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048447505894929233,
      "loss": 5.4101,
      "mean_token_accuracy": 0.15018145143985748,
      "num_tokens": 28205712.0,
      "step": 12865
    },
    {
      "entropy": 5.605202913284302,
      "epoch": 1.277292576419214,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048446212320019306,
      "loss": 5.348,
      "mean_token_accuracy": 0.1521398663520813,
      "num_tokens": 28217246.0,
      "step": 12870
    },
    {
      "entropy": 5.579190254211426,
      "epoch": 1.2777888050813815,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004844491822568846,
      "loss": 5.3587,
      "mean_token_accuracy": 0.1473693385720253,
      "num_tokens": 28227933.0,
      "step": 12875
    },
    {
      "entropy": 5.549760150909424,
      "epoch": 1.278285033743549,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048443623611968764,
      "loss": 5.3805,
      "mean_token_accuracy": 0.15235921144485473,
      "num_tokens": 28239692.0,
      "step": 12880
    },
    {
      "entropy": 5.527844047546386,
      "epoch": 1.2787812624057167,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048442328478892334,
      "loss": 5.3414,
      "mean_token_accuracy": 0.1490297019481659,
      "num_tokens": 28249815.0,
      "step": 12885
    },
    {
      "entropy": 5.618090105056763,
      "epoch": 1.2792774910678841,
      "grad_norm": 1.375,
      "learning_rate": 0.0004844103282649128,
      "loss": 5.3755,
      "mean_token_accuracy": 0.14933249354362488,
      "num_tokens": 28260385.0,
      "step": 12890
    },
    {
      "entropy": 5.526648473739624,
      "epoch": 1.2797737197300516,
      "grad_norm": 1.25,
      "learning_rate": 0.0004843973665479774,
      "loss": 5.3784,
      "mean_token_accuracy": 0.1443515494465828,
      "num_tokens": 28271858.0,
      "step": 12895
    },
    {
      "entropy": 5.606471300125122,
      "epoch": 1.280269948392219,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048438439963843853,
      "loss": 5.3824,
      "mean_token_accuracy": 0.14972061216831206,
      "num_tokens": 28283988.0,
      "step": 12900
    },
    {
      "entropy": 5.58480920791626,
      "epoch": 1.2807661770543866,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048437142753661775,
      "loss": 5.406,
      "mean_token_accuracy": 0.1487390622496605,
      "num_tokens": 28293281.0,
      "step": 12905
    },
    {
      "entropy": 5.602814435958862,
      "epoch": 1.2812624057165543,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004843584502428367,
      "loss": 5.3609,
      "mean_token_accuracy": 0.14878564327955246,
      "num_tokens": 28304468.0,
      "step": 12910
    },
    {
      "entropy": 5.5257039070129395,
      "epoch": 1.2817586343787217,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004843454677574173,
      "loss": 5.3212,
      "mean_token_accuracy": 0.15448291152715682,
      "num_tokens": 28315864.0,
      "step": 12915
    },
    {
      "entropy": 5.509317684173584,
      "epoch": 1.2822548630408892,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048433248008068123,
      "loss": 5.3805,
      "mean_token_accuracy": 0.15584477931261062,
      "num_tokens": 28327212.0,
      "step": 12920
    },
    {
      "entropy": 5.569231367111206,
      "epoch": 1.282751091703057,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048431948721295085,
      "loss": 5.3924,
      "mean_token_accuracy": 0.15438440442085266,
      "num_tokens": 28339256.0,
      "step": 12925
    },
    {
      "entropy": 5.640489673614502,
      "epoch": 1.2832473203652244,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004843064891545482,
      "loss": 5.4091,
      "mean_token_accuracy": 0.14625887870788573,
      "num_tokens": 28350196.0,
      "step": 12930
    },
    {
      "entropy": 5.553873062133789,
      "epoch": 1.2837435490273919,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004842934859057956,
      "loss": 5.3435,
      "mean_token_accuracy": 0.15176190584897994,
      "num_tokens": 28361241.0,
      "step": 12935
    },
    {
      "entropy": 5.633756303787232,
      "epoch": 1.2842397776895593,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004842804774670156,
      "loss": 5.4925,
      "mean_token_accuracy": 0.14367921203374862,
      "num_tokens": 28373369.0,
      "step": 12940
    },
    {
      "entropy": 5.58552188873291,
      "epoch": 1.2847360063517268,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048426746383853066,
      "loss": 5.2981,
      "mean_token_accuracy": 0.15410922914743425,
      "num_tokens": 28384970.0,
      "step": 12945
    },
    {
      "entropy": 5.595558738708496,
      "epoch": 1.2852322350138943,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048425444502066354,
      "loss": 5.3299,
      "mean_token_accuracy": 0.15125948339700698,
      "num_tokens": 28396469.0,
      "step": 12950
    },
    {
      "entropy": 5.534946441650391,
      "epoch": 1.285728463676062,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048424142101373706,
      "loss": 5.3857,
      "mean_token_accuracy": 0.15493735671043396,
      "num_tokens": 28406975.0,
      "step": 12955
    },
    {
      "entropy": 5.506554317474365,
      "epoch": 1.2862246923382294,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048422839181807426,
      "loss": 5.2392,
      "mean_token_accuracy": 0.16755890995264053,
      "num_tokens": 28418285.0,
      "step": 12960
    },
    {
      "entropy": 5.550190496444702,
      "epoch": 1.286720921000397,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004842153574339982,
      "loss": 5.4042,
      "mean_token_accuracy": 0.14861235693097113,
      "num_tokens": 28429769.0,
      "step": 12965
    },
    {
      "entropy": 5.527503108978271,
      "epoch": 1.2872171496625646,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048420231786183216,
      "loss": 5.3598,
      "mean_token_accuracy": 0.15536876767873764,
      "num_tokens": 28441188.0,
      "step": 12970
    },
    {
      "entropy": 5.6512003421783445,
      "epoch": 1.287713378324732,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004841892731018994,
      "loss": 5.4334,
      "mean_token_accuracy": 0.14364735558629035,
      "num_tokens": 28453504.0,
      "step": 12975
    },
    {
      "entropy": 5.577330875396728,
      "epoch": 1.2882096069868996,
      "grad_norm": 1.875,
      "learning_rate": 0.0004841762231545233,
      "loss": 5.4169,
      "mean_token_accuracy": 0.1548916831612587,
      "num_tokens": 28465930.0,
      "step": 12980
    },
    {
      "entropy": 5.542925024032593,
      "epoch": 1.288705835649067,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048416316802002776,
      "loss": 5.3453,
      "mean_token_accuracy": 0.14989717453718185,
      "num_tokens": 28477803.0,
      "step": 12985
    },
    {
      "entropy": 5.638417100906372,
      "epoch": 1.2892020643112345,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048415010769873625,
      "loss": 5.4134,
      "mean_token_accuracy": 0.14915625452995301,
      "num_tokens": 28488556.0,
      "step": 12990
    },
    {
      "entropy": 5.569115781784058,
      "epoch": 1.2896982929734022,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004841370421909728,
      "loss": 5.3207,
      "mean_token_accuracy": 0.15568244010210036,
      "num_tokens": 28499275.0,
      "step": 12995
    },
    {
      "entropy": 5.586712265014649,
      "epoch": 1.2901945216355697,
      "grad_norm": 1.796875,
      "learning_rate": 0.00048412397149706133,
      "loss": 5.3461,
      "mean_token_accuracy": 0.1543872371315956,
      "num_tokens": 28511327.0,
      "step": 13000
    },
    {
      "entropy": 5.5323258399963375,
      "epoch": 1.2906907502977372,
      "grad_norm": 1.453125,
      "learning_rate": 0.000484110895617326,
      "loss": 5.3778,
      "mean_token_accuracy": 0.1535607770085335,
      "num_tokens": 28523937.0,
      "step": 13005
    },
    {
      "entropy": 5.492031097412109,
      "epoch": 1.2911869789599046,
      "grad_norm": 1.828125,
      "learning_rate": 0.00048409781455209104,
      "loss": 5.2861,
      "mean_token_accuracy": 0.15117571055889129,
      "num_tokens": 28534028.0,
      "step": 13010
    },
    {
      "entropy": 5.6284003257751465,
      "epoch": 1.2916832076220723,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004840847283016809,
      "loss": 5.3809,
      "mean_token_accuracy": 0.14275674819946288,
      "num_tokens": 28545480.0,
      "step": 13015
    },
    {
      "entropy": 5.603991413116455,
      "epoch": 1.2921794362842398,
      "grad_norm": 1.375,
      "learning_rate": 0.00048407163686642,
      "loss": 5.3276,
      "mean_token_accuracy": 0.159427011013031,
      "num_tokens": 28555401.0,
      "step": 13020
    },
    {
      "entropy": 5.576207780838013,
      "epoch": 1.2926756649464073,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000484058540246633,
      "loss": 5.3822,
      "mean_token_accuracy": 0.1527725487947464,
      "num_tokens": 28565497.0,
      "step": 13025
    },
    {
      "entropy": 5.554252099990845,
      "epoch": 1.2931718936085748,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004840454384426447,
      "loss": 5.4869,
      "mean_token_accuracy": 0.14387504905462264,
      "num_tokens": 28576186.0,
      "step": 13030
    },
    {
      "entropy": 5.591579914093018,
      "epoch": 1.2936681222707422,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004840323314547799,
      "loss": 5.3406,
      "mean_token_accuracy": 0.15421019345521927,
      "num_tokens": 28586202.0,
      "step": 13035
    },
    {
      "entropy": 5.572428846359253,
      "epoch": 1.29416435093291,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048401921928336385,
      "loss": 5.32,
      "mean_token_accuracy": 0.15363701060414314,
      "num_tokens": 28596486.0,
      "step": 13040
    },
    {
      "entropy": 5.60749773979187,
      "epoch": 1.2946605795950774,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004840061019287215,
      "loss": 5.3613,
      "mean_token_accuracy": 0.15421962589025498,
      "num_tokens": 28607934.0,
      "step": 13045
    },
    {
      "entropy": 5.625794553756714,
      "epoch": 1.2951568082572449,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048399297939117817,
      "loss": 5.3935,
      "mean_token_accuracy": 0.1478491336107254,
      "num_tokens": 28617673.0,
      "step": 13050
    },
    {
      "entropy": 5.640213489532471,
      "epoch": 1.2956530369194126,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004839798516710593,
      "loss": 5.3954,
      "mean_token_accuracy": 0.14870392829179763,
      "num_tokens": 28629811.0,
      "step": 13055
    },
    {
      "entropy": 5.658992671966553,
      "epoch": 1.29614926558158,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004839667187686904,
      "loss": 5.4612,
      "mean_token_accuracy": 0.15079117566347122,
      "num_tokens": 28640371.0,
      "step": 13060
    },
    {
      "entropy": 5.603936624526978,
      "epoch": 1.2966454942437475,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004839535806843972,
      "loss": 5.3833,
      "mean_token_accuracy": 0.15697976499795913,
      "num_tokens": 28651904.0,
      "step": 13065
    },
    {
      "entropy": 5.529257917404175,
      "epoch": 1.297141722905915,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004839404374185054,
      "loss": 5.3092,
      "mean_token_accuracy": 0.15195778235793114,
      "num_tokens": 28662241.0,
      "step": 13070
    },
    {
      "entropy": 5.526766777038574,
      "epoch": 1.2976379515680825,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000483927288971341,
      "loss": 5.2947,
      "mean_token_accuracy": 0.15698595643043517,
      "num_tokens": 28672791.0,
      "step": 13075
    },
    {
      "entropy": 5.624585199356079,
      "epoch": 1.2981341802302502,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004839141353432301,
      "loss": 5.4226,
      "mean_token_accuracy": 0.14754125028848647,
      "num_tokens": 28684218.0,
      "step": 13080
    },
    {
      "entropy": 5.561284828186035,
      "epoch": 1.2986304088924177,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004839009765344987,
      "loss": 5.3639,
      "mean_token_accuracy": 0.149440698325634,
      "num_tokens": 28694690.0,
      "step": 13085
    },
    {
      "entropy": 5.56783127784729,
      "epoch": 1.2991266375545851,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00048388781254547324,
      "loss": 5.3655,
      "mean_token_accuracy": 0.1562001071870327,
      "num_tokens": 28705890.0,
      "step": 13090
    },
    {
      "entropy": 5.638034009933472,
      "epoch": 1.2996228662167526,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048387464337648026,
      "loss": 5.3832,
      "mean_token_accuracy": 0.151560540497303,
      "num_tokens": 28716935.0,
      "step": 13095
    },
    {
      "entropy": 5.593366384506226,
      "epoch": 1.3001190948789203,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048386146902784606,
      "loss": 5.3771,
      "mean_token_accuracy": 0.15100871920585632,
      "num_tokens": 28728053.0,
      "step": 13100
    },
    {
      "entropy": 5.52743124961853,
      "epoch": 1.3006153235410878,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048384828949989757,
      "loss": 5.3288,
      "mean_token_accuracy": 0.15141219049692153,
      "num_tokens": 28738940.0,
      "step": 13105
    },
    {
      "entropy": 5.6083296775817875,
      "epoch": 1.3011115522032553,
      "grad_norm": 1.375,
      "learning_rate": 0.00048383510479296153,
      "loss": 5.3937,
      "mean_token_accuracy": 0.14763452857732773,
      "num_tokens": 28750259.0,
      "step": 13110
    },
    {
      "entropy": 5.599593257904052,
      "epoch": 1.3016077808654227,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004838219149073649,
      "loss": 5.338,
      "mean_token_accuracy": 0.15788400620222093,
      "num_tokens": 28760119.0,
      "step": 13115
    },
    {
      "entropy": 5.602177381515503,
      "epoch": 1.3021040095275902,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004838087198434347,
      "loss": 5.3942,
      "mean_token_accuracy": 0.1470458671450615,
      "num_tokens": 28772022.0,
      "step": 13120
    },
    {
      "entropy": 5.6040068626403805,
      "epoch": 1.302600238189758,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004837955196014982,
      "loss": 5.3393,
      "mean_token_accuracy": 0.14665279388427735,
      "num_tokens": 28782935.0,
      "step": 13125
    },
    {
      "entropy": 5.659886455535888,
      "epoch": 1.3030964668519254,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048378231418188275,
      "loss": 5.4528,
      "mean_token_accuracy": 0.15408275723457338,
      "num_tokens": 28793604.0,
      "step": 13130
    },
    {
      "entropy": 5.612650918960571,
      "epoch": 1.3035926955140928,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048376910358491576,
      "loss": 5.4019,
      "mean_token_accuracy": 0.14457454681396484,
      "num_tokens": 28803494.0,
      "step": 13135
    },
    {
      "entropy": 5.547444248199463,
      "epoch": 1.3040889241762605,
      "grad_norm": 1.25,
      "learning_rate": 0.0004837558878109249,
      "loss": 5.3913,
      "mean_token_accuracy": 0.1452224887907505,
      "num_tokens": 28816232.0,
      "step": 13140
    },
    {
      "entropy": 5.588050174713135,
      "epoch": 1.304585152838428,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004837426668602378,
      "loss": 5.3415,
      "mean_token_accuracy": 0.15535330325365065,
      "num_tokens": 28827056.0,
      "step": 13145
    },
    {
      "entropy": 5.558589839935303,
      "epoch": 1.3050813815005955,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004837294407331825,
      "loss": 5.3195,
      "mean_token_accuracy": 0.1550546258687973,
      "num_tokens": 28837217.0,
      "step": 13150
    },
    {
      "entropy": 5.488318061828613,
      "epoch": 1.305577610162763,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004837162094300867,
      "loss": 5.2379,
      "mean_token_accuracy": 0.15384215265512466,
      "num_tokens": 28846838.0,
      "step": 13155
    },
    {
      "entropy": 5.580583286285401,
      "epoch": 1.3060738388249304,
      "grad_norm": 1.5,
      "learning_rate": 0.0004837029729512787,
      "loss": 5.2817,
      "mean_token_accuracy": 0.16084386855363847,
      "num_tokens": 28856269.0,
      "step": 13160
    },
    {
      "entropy": 5.589701223373413,
      "epoch": 1.306570067487098,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00048368973129708667,
      "loss": 5.337,
      "mean_token_accuracy": 0.1564713880419731,
      "num_tokens": 28867142.0,
      "step": 13165
    },
    {
      "entropy": 5.561997175216675,
      "epoch": 1.3070662961492656,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048367648446783894,
      "loss": 5.2996,
      "mean_token_accuracy": 0.14592718929052353,
      "num_tokens": 28877346.0,
      "step": 13170
    },
    {
      "entropy": 5.54708161354065,
      "epoch": 1.307562524811433,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004836632324638642,
      "loss": 5.2924,
      "mean_token_accuracy": 0.15328410118818284,
      "num_tokens": 28888425.0,
      "step": 13175
    },
    {
      "entropy": 5.593842458724976,
      "epoch": 1.3080587534736006,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048364997528549077,
      "loss": 5.3331,
      "mean_token_accuracy": 0.15295085087418556,
      "num_tokens": 28899575.0,
      "step": 13180
    },
    {
      "entropy": 5.595189189910888,
      "epoch": 1.3085549821357683,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048363671293304763,
      "loss": 5.3161,
      "mean_token_accuracy": 0.15440469682216645,
      "num_tokens": 28910769.0,
      "step": 13185
    },
    {
      "entropy": 5.566035604476928,
      "epoch": 1.3090512107979357,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004836234454068636,
      "loss": 5.3426,
      "mean_token_accuracy": 0.15950863361358641,
      "num_tokens": 28921224.0,
      "step": 13190
    },
    {
      "entropy": 5.5573341846466064,
      "epoch": 1.3095474394601032,
      "grad_norm": 1.90625,
      "learning_rate": 0.00048361017270726764,
      "loss": 5.3073,
      "mean_token_accuracy": 0.15121300145983696,
      "num_tokens": 28932585.0,
      "step": 13195
    },
    {
      "entropy": 5.496748399734497,
      "epoch": 1.3100436681222707,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004835968948345889,
      "loss": 5.3721,
      "mean_token_accuracy": 0.149286749958992,
      "num_tokens": 28943471.0,
      "step": 13200
    },
    {
      "entropy": 5.534739875793457,
      "epoch": 1.3105398967844382,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048358361178915666,
      "loss": 5.2969,
      "mean_token_accuracy": 0.1617283880710602,
      "num_tokens": 28956336.0,
      "step": 13205
    },
    {
      "entropy": 5.6661313533782955,
      "epoch": 1.3110361254466059,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004835703235713003,
      "loss": 5.4057,
      "mean_token_accuracy": 0.15103012025356294,
      "num_tokens": 28966689.0,
      "step": 13210
    },
    {
      "entropy": 5.650691509246826,
      "epoch": 1.3115323541087733,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048355703018134935,
      "loss": 5.549,
      "mean_token_accuracy": 0.15192205160856248,
      "num_tokens": 28978548.0,
      "step": 13215
    },
    {
      "entropy": 5.679022264480591,
      "epoch": 1.3120285827709408,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004835437316196335,
      "loss": 5.421,
      "mean_token_accuracy": 0.15069350004196166,
      "num_tokens": 28988899.0,
      "step": 13220
    },
    {
      "entropy": 5.5571788311004635,
      "epoch": 1.3125248114331085,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004835304278864824,
      "loss": 5.3306,
      "mean_token_accuracy": 0.15821077525615693,
      "num_tokens": 28999666.0,
      "step": 13225
    },
    {
      "entropy": 5.574817228317261,
      "epoch": 1.313021040095276,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004835171189822261,
      "loss": 5.3478,
      "mean_token_accuracy": 0.1499969929456711,
      "num_tokens": 29010185.0,
      "step": 13230
    },
    {
      "entropy": 5.5349036693573,
      "epoch": 1.3135172687574435,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004835038049071945,
      "loss": 5.3652,
      "mean_token_accuracy": 0.15304857045412062,
      "num_tokens": 29020761.0,
      "step": 13235
    },
    {
      "entropy": 5.656700181961059,
      "epoch": 1.314013497419611,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004834904856617178,
      "loss": 5.5017,
      "mean_token_accuracy": 0.1512613207101822,
      "num_tokens": 29031844.0,
      "step": 13240
    },
    {
      "entropy": 5.726397943496704,
      "epoch": 1.3145097260817784,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048347716124612643,
      "loss": 5.4697,
      "mean_token_accuracy": 0.1430974066257477,
      "num_tokens": 29043101.0,
      "step": 13245
    },
    {
      "entropy": 5.627517223358154,
      "epoch": 1.3150059547439459,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004834638316607506,
      "loss": 5.3929,
      "mean_token_accuracy": 0.15277172178030013,
      "num_tokens": 29054016.0,
      "step": 13250
    },
    {
      "entropy": 5.537429237365723,
      "epoch": 1.3155021834061136,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048345049690592105,
      "loss": 5.3572,
      "mean_token_accuracy": 0.147917278110981,
      "num_tokens": 29065127.0,
      "step": 13255
    },
    {
      "entropy": 5.673310852050781,
      "epoch": 1.315998412068281,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048343715698196814,
      "loss": 5.323,
      "mean_token_accuracy": 0.15524778962135316,
      "num_tokens": 29075130.0,
      "step": 13260
    },
    {
      "entropy": 5.684695816040039,
      "epoch": 1.3164946407304485,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004834238118892231,
      "loss": 5.4538,
      "mean_token_accuracy": 0.14309114068746567,
      "num_tokens": 29086405.0,
      "step": 13265
    },
    {
      "entropy": 5.627870321273804,
      "epoch": 1.3169908693926162,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004834104616280165,
      "loss": 5.419,
      "mean_token_accuracy": 0.14766013324260713,
      "num_tokens": 29096945.0,
      "step": 13270
    },
    {
      "entropy": 5.6671790599823,
      "epoch": 1.3174870980547837,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004833971061986796,
      "loss": 5.4653,
      "mean_token_accuracy": 0.1524786800146103,
      "num_tokens": 29108275.0,
      "step": 13275
    },
    {
      "entropy": 5.572581624984741,
      "epoch": 1.3179833267169512,
      "grad_norm": 1.375,
      "learning_rate": 0.00048338374560154344,
      "loss": 5.3194,
      "mean_token_accuracy": 0.15354548543691635,
      "num_tokens": 29118832.0,
      "step": 13280
    },
    {
      "entropy": 5.548596000671386,
      "epoch": 1.3184795553791187,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004833703798369395,
      "loss": 5.3458,
      "mean_token_accuracy": 0.15526807606220244,
      "num_tokens": 29130720.0,
      "step": 13285
    },
    {
      "entropy": 5.6260065078735355,
      "epoch": 1.3189757840412861,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004833570089051992,
      "loss": 5.3199,
      "mean_token_accuracy": 0.15394648015499116,
      "num_tokens": 29140441.0,
      "step": 13290
    },
    {
      "entropy": 5.611601638793945,
      "epoch": 1.3194720127034538,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048334363280665393,
      "loss": 5.3924,
      "mean_token_accuracy": 0.14619396924972533,
      "num_tokens": 29151544.0,
      "step": 13295
    },
    {
      "entropy": 5.6447371482849125,
      "epoch": 1.3199682413656213,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004833302515416355,
      "loss": 5.4751,
      "mean_token_accuracy": 0.1516022652387619,
      "num_tokens": 29165475.0,
      "step": 13300
    },
    {
      "entropy": 5.534718322753906,
      "epoch": 1.3204644700277888,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048331686511047584,
      "loss": 5.2961,
      "mean_token_accuracy": 0.14893410205841065,
      "num_tokens": 29177139.0,
      "step": 13305
    },
    {
      "entropy": 5.604742860794067,
      "epoch": 1.3209606986899565,
      "grad_norm": 1.25,
      "learning_rate": 0.00048330347351350676,
      "loss": 5.3691,
      "mean_token_accuracy": 0.15456988513469697,
      "num_tokens": 29187812.0,
      "step": 13310
    },
    {
      "entropy": 5.513737154006958,
      "epoch": 1.321456927352124,
      "grad_norm": 1.5,
      "learning_rate": 0.00048329007675106036,
      "loss": 5.2682,
      "mean_token_accuracy": 0.15255578160285949,
      "num_tokens": 29197932.0,
      "step": 13315
    },
    {
      "entropy": 5.562999534606933,
      "epoch": 1.3219531560142914,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048327667482346895,
      "loss": 5.2752,
      "mean_token_accuracy": 0.15911680161952974,
      "num_tokens": 29209529.0,
      "step": 13320
    },
    {
      "entropy": 5.540583562850952,
      "epoch": 1.322449384676459,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004832632677310647,
      "loss": 5.3338,
      "mean_token_accuracy": 0.15652776062488555,
      "num_tokens": 29219027.0,
      "step": 13325
    },
    {
      "entropy": 5.5445887565612795,
      "epoch": 1.3229456133386264,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048324985547418026,
      "loss": 5.3811,
      "mean_token_accuracy": 0.14986519515514374,
      "num_tokens": 29229647.0,
      "step": 13330
    },
    {
      "entropy": 5.621123743057251,
      "epoch": 1.3234418420007938,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004832364380531482,
      "loss": 5.4319,
      "mean_token_accuracy": 0.15309610664844514,
      "num_tokens": 29240263.0,
      "step": 13335
    },
    {
      "entropy": 5.652767753601074,
      "epoch": 1.3239380706629615,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048322301546830103,
      "loss": 5.4117,
      "mean_token_accuracy": 0.1465339258313179,
      "num_tokens": 29251115.0,
      "step": 13340
    },
    {
      "entropy": 5.579132890701294,
      "epoch": 1.324434299325129,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004832095877199718,
      "loss": 5.3905,
      "mean_token_accuracy": 0.14945088773965837,
      "num_tokens": 29261888.0,
      "step": 13345
    },
    {
      "entropy": 5.488748025894165,
      "epoch": 1.3249305279872965,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048319615480849347,
      "loss": 5.2721,
      "mean_token_accuracy": 0.1607190862298012,
      "num_tokens": 29271948.0,
      "step": 13350
    },
    {
      "entropy": 5.5756189823150635,
      "epoch": 1.3254267566494642,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048318271673419914,
      "loss": 5.3288,
      "mean_token_accuracy": 0.1518791064620018,
      "num_tokens": 29282449.0,
      "step": 13355
    },
    {
      "entropy": 5.624840974807739,
      "epoch": 1.3259229853116317,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048316927349742207,
      "loss": 5.378,
      "mean_token_accuracy": 0.14904360622167587,
      "num_tokens": 29294280.0,
      "step": 13360
    },
    {
      "entropy": 5.607909488677978,
      "epoch": 1.3264192139737991,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004831558250984955,
      "loss": 5.3399,
      "mean_token_accuracy": 0.15486692041158676,
      "num_tokens": 29305143.0,
      "step": 13365
    },
    {
      "entropy": 5.531710433959961,
      "epoch": 1.3269154426359666,
      "grad_norm": 2.09375,
      "learning_rate": 0.00048314237153775303,
      "loss": 5.2711,
      "mean_token_accuracy": 0.15018926858901976,
      "num_tokens": 29314697.0,
      "step": 13370
    },
    {
      "entropy": 5.484866809844971,
      "epoch": 1.327411671298134,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004831289128155283,
      "loss": 5.2942,
      "mean_token_accuracy": 0.15260104984045028,
      "num_tokens": 29325257.0,
      "step": 13375
    },
    {
      "entropy": 5.557030725479126,
      "epoch": 1.3279078999603018,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048311544893215496,
      "loss": 5.3926,
      "mean_token_accuracy": 0.1459612399339676,
      "num_tokens": 29335449.0,
      "step": 13380
    },
    {
      "entropy": 5.587840366363525,
      "epoch": 1.3284041286224693,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000483101979887967,
      "loss": 5.3792,
      "mean_token_accuracy": 0.1483737587928772,
      "num_tokens": 29346898.0,
      "step": 13385
    },
    {
      "entropy": 5.655613946914673,
      "epoch": 1.3289003572846367,
      "grad_norm": 1.953125,
      "learning_rate": 0.00048308850568329825,
      "loss": 5.3532,
      "mean_token_accuracy": 0.14797112941741944,
      "num_tokens": 29356118.0,
      "step": 13390
    },
    {
      "entropy": 5.599618148803711,
      "epoch": 1.3293965859468042,
      "grad_norm": 1.625,
      "learning_rate": 0.00048307502631848306,
      "loss": 5.4307,
      "mean_token_accuracy": 0.14288025051355363,
      "num_tokens": 29367597.0,
      "step": 13395
    },
    {
      "entropy": 5.567211484909057,
      "epoch": 1.329892814608972,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048306154179385546,
      "loss": 5.2901,
      "mean_token_accuracy": 0.1498557835817337,
      "num_tokens": 29377829.0,
      "step": 13400
    },
    {
      "entropy": 5.555682516098022,
      "epoch": 1.3303890432711394,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004830480521097501,
      "loss": 5.2451,
      "mean_token_accuracy": 0.16279564201831817,
      "num_tokens": 29387945.0,
      "step": 13405
    },
    {
      "entropy": 5.4644063949584964,
      "epoch": 1.3308852719333069,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004830345572665012,
      "loss": 5.228,
      "mean_token_accuracy": 0.1595796063542366,
      "num_tokens": 29398269.0,
      "step": 13410
    },
    {
      "entropy": 5.5952788352966305,
      "epoch": 1.3313815005954743,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048302105726444366,
      "loss": 5.4295,
      "mean_token_accuracy": 0.1447591334581375,
      "num_tokens": 29410946.0,
      "step": 13415
    },
    {
      "entropy": 5.585665035247803,
      "epoch": 1.3318777292576418,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00048300755210391205,
      "loss": 5.2604,
      "mean_token_accuracy": 0.15266963094472885,
      "num_tokens": 29421770.0,
      "step": 13420
    },
    {
      "entropy": 5.651066589355469,
      "epoch": 1.3323739579198095,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004829940417852415,
      "loss": 5.4427,
      "mean_token_accuracy": 0.14749201238155366,
      "num_tokens": 29433717.0,
      "step": 13425
    },
    {
      "entropy": 5.522225713729858,
      "epoch": 1.332870186581977,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004829805263087668,
      "loss": 5.2568,
      "mean_token_accuracy": 0.1599634602665901,
      "num_tokens": 29444382.0,
      "step": 13430
    },
    {
      "entropy": 5.584672498703003,
      "epoch": 1.3333664152441445,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048296700567482324,
      "loss": 5.3808,
      "mean_token_accuracy": 0.1493477627635002,
      "num_tokens": 29454680.0,
      "step": 13435
    },
    {
      "entropy": 5.60670018196106,
      "epoch": 1.3338626439063122,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004829534798837461,
      "loss": 5.3973,
      "mean_token_accuracy": 0.1513490632176399,
      "num_tokens": 29465065.0,
      "step": 13440
    },
    {
      "entropy": 5.626371192932129,
      "epoch": 1.3343588725684796,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004829399489358707,
      "loss": 5.3239,
      "mean_token_accuracy": 0.15114783942699433,
      "num_tokens": 29475704.0,
      "step": 13445
    },
    {
      "entropy": 5.561667203903198,
      "epoch": 1.334855101230647,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004829264128315326,
      "loss": 5.3674,
      "mean_token_accuracy": 0.15148624330759047,
      "num_tokens": 29486418.0,
      "step": 13450
    },
    {
      "entropy": 5.6249549865722654,
      "epoch": 1.3353513298928146,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004829128715710676,
      "loss": 5.4478,
      "mean_token_accuracy": 0.1438056670129299,
      "num_tokens": 29498939.0,
      "step": 13455
    },
    {
      "entropy": 5.5718285083770756,
      "epoch": 1.335847558554982,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004828993251548114,
      "loss": 5.2479,
      "mean_token_accuracy": 0.16082777380943297,
      "num_tokens": 29508525.0,
      "step": 13460
    },
    {
      "entropy": 5.603612470626831,
      "epoch": 1.3363437872171495,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004828857735830999,
      "loss": 5.3803,
      "mean_token_accuracy": 0.15381963104009627,
      "num_tokens": 29519962.0,
      "step": 13465
    },
    {
      "entropy": 5.570424127578735,
      "epoch": 1.3368400158793172,
      "grad_norm": 2.125,
      "learning_rate": 0.0004828722168562691,
      "loss": 5.3362,
      "mean_token_accuracy": 0.1524523004889488,
      "num_tokens": 29529586.0,
      "step": 13470
    },
    {
      "entropy": 5.591733455657959,
      "epoch": 1.3373362445414847,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004828586549746553,
      "loss": 5.4066,
      "mean_token_accuracy": 0.1448112055659294,
      "num_tokens": 29540670.0,
      "step": 13475
    },
    {
      "entropy": 5.6166322231292725,
      "epoch": 1.3378324732036522,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004828450879385947,
      "loss": 5.3804,
      "mean_token_accuracy": 0.14908209443092346,
      "num_tokens": 29551484.0,
      "step": 13480
    },
    {
      "entropy": 5.610360336303711,
      "epoch": 1.3383287018658199,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004828315157484239,
      "loss": 5.298,
      "mean_token_accuracy": 0.15646330267190933,
      "num_tokens": 29561596.0,
      "step": 13485
    },
    {
      "entropy": 5.596626138687133,
      "epoch": 1.3388249305279873,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004828179384044793,
      "loss": 5.3367,
      "mean_token_accuracy": 0.15386511534452438,
      "num_tokens": 29572837.0,
      "step": 13490
    },
    {
      "entropy": 5.526793146133423,
      "epoch": 1.3393211591901548,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048280435590709756,
      "loss": 5.297,
      "mean_token_accuracy": 0.15412984639406205,
      "num_tokens": 29584095.0,
      "step": 13495
    },
    {
      "entropy": 5.526909494400025,
      "epoch": 1.3398173878523223,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004827907682566157,
      "loss": 5.3521,
      "mean_token_accuracy": 0.1491902768611908,
      "num_tokens": 29594903.0,
      "step": 13500
    },
    {
      "entropy": 5.566684913635254,
      "epoch": 1.3403136165144898,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004827771754533704,
      "loss": 5.3141,
      "mean_token_accuracy": 0.15712702125310898,
      "num_tokens": 29605199.0,
      "step": 13505
    },
    {
      "entropy": 5.639581871032715,
      "epoch": 1.3408098451766575,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048276357749769895,
      "loss": 5.4995,
      "mean_token_accuracy": 0.1437547467648983,
      "num_tokens": 29617535.0,
      "step": 13510
    },
    {
      "entropy": 5.550845956802368,
      "epoch": 1.341306073838825,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048274997438993845,
      "loss": 5.2944,
      "mean_token_accuracy": 0.1543087601661682,
      "num_tokens": 29628227.0,
      "step": 13515
    },
    {
      "entropy": 5.586678314208984,
      "epoch": 1.3418023025009924,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048273636613042626,
      "loss": 5.4285,
      "mean_token_accuracy": 0.1478160500526428,
      "num_tokens": 29639112.0,
      "step": 13520
    },
    {
      "entropy": 5.560399293899536,
      "epoch": 1.3422985311631601,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004827227527194998,
      "loss": 5.2891,
      "mean_token_accuracy": 0.15110662281513215,
      "num_tokens": 29650768.0,
      "step": 13525
    },
    {
      "entropy": 5.595443153381348,
      "epoch": 1.3427947598253276,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004827091341574967,
      "loss": 5.415,
      "mean_token_accuracy": 0.15476094037294388,
      "num_tokens": 29662388.0,
      "step": 13530
    },
    {
      "entropy": 5.58453106880188,
      "epoch": 1.343290988487495,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004826955104447546,
      "loss": 5.2899,
      "mean_token_accuracy": 0.15947073847055435,
      "num_tokens": 29672231.0,
      "step": 13535
    },
    {
      "entropy": 5.620069932937622,
      "epoch": 1.3437872171496625,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004826818815816114,
      "loss": 5.377,
      "mean_token_accuracy": 0.15510047376155853,
      "num_tokens": 29683019.0,
      "step": 13540
    },
    {
      "entropy": 5.613540506362915,
      "epoch": 1.34428344581183,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048266824756840513,
      "loss": 5.3636,
      "mean_token_accuracy": 0.15577624142169952,
      "num_tokens": 29694233.0,
      "step": 13545
    },
    {
      "entropy": 5.612598562240601,
      "epoch": 1.3447796744739975,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004826546084054737,
      "loss": 5.3841,
      "mean_token_accuracy": 0.15341310873627662,
      "num_tokens": 29704023.0,
      "step": 13550
    },
    {
      "entropy": 5.663148832321167,
      "epoch": 1.3452759031361652,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004826409640931555,
      "loss": 5.4831,
      "mean_token_accuracy": 0.1463843271136284,
      "num_tokens": 29715407.0,
      "step": 13555
    },
    {
      "entropy": 5.552429628372193,
      "epoch": 1.3457721317983327,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048262731463178884,
      "loss": 5.322,
      "mean_token_accuracy": 0.15814246833324433,
      "num_tokens": 29726223.0,
      "step": 13560
    },
    {
      "entropy": 5.528812074661255,
      "epoch": 1.3462683604605001,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004826136600217121,
      "loss": 5.2441,
      "mean_token_accuracy": 0.15609188228845597,
      "num_tokens": 29735850.0,
      "step": 13565
    },
    {
      "entropy": 5.597129201889038,
      "epoch": 1.3467645891226678,
      "grad_norm": 1.40625,
      "learning_rate": 0.000482600000263264,
      "loss": 5.3836,
      "mean_token_accuracy": 0.14905630946159362,
      "num_tokens": 29747381.0,
      "step": 13570
    },
    {
      "entropy": 5.600694227218628,
      "epoch": 1.3472608177848353,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004825863353567832,
      "loss": 5.4453,
      "mean_token_accuracy": 0.14750638529658316,
      "num_tokens": 29758440.0,
      "step": 13575
    },
    {
      "entropy": 5.547540473937988,
      "epoch": 1.3477570464470028,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048257266530260865,
      "loss": 5.3068,
      "mean_token_accuracy": 0.15175359845161437,
      "num_tokens": 29770606.0,
      "step": 13580
    },
    {
      "entropy": 5.560246181488037,
      "epoch": 1.3482532751091703,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004825589901010792,
      "loss": 5.3526,
      "mean_token_accuracy": 0.15345120579004287,
      "num_tokens": 29781469.0,
      "step": 13585
    },
    {
      "entropy": 5.592626857757568,
      "epoch": 1.3487495037713377,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004825453097525341,
      "loss": 5.4129,
      "mean_token_accuracy": 0.15098052620887756,
      "num_tokens": 29792847.0,
      "step": 13590
    },
    {
      "entropy": 5.567552471160889,
      "epoch": 1.3492457324335054,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048253162425731253,
      "loss": 5.3273,
      "mean_token_accuracy": 0.15475840121507645,
      "num_tokens": 29802784.0,
      "step": 13595
    },
    {
      "entropy": 5.651505374908448,
      "epoch": 1.349741961095673,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004825179336157539,
      "loss": 5.331,
      "mean_token_accuracy": 0.1578106701374054,
      "num_tokens": 29812849.0,
      "step": 13600
    },
    {
      "entropy": 5.580387449264526,
      "epoch": 1.3502381897578404,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048250423782819765,
      "loss": 5.3319,
      "mean_token_accuracy": 0.1515057496726513,
      "num_tokens": 29823495.0,
      "step": 13605
    },
    {
      "entropy": 5.543574380874634,
      "epoch": 1.350734418420008,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004824905368949834,
      "loss": 5.3712,
      "mean_token_accuracy": 0.1500635877251625,
      "num_tokens": 29835295.0,
      "step": 13610
    },
    {
      "entropy": 5.550374364852905,
      "epoch": 1.3512306470821756,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004824768308164509,
      "loss": 5.3137,
      "mean_token_accuracy": 0.15903579890727998,
      "num_tokens": 29845747.0,
      "step": 13615
    },
    {
      "entropy": 5.603609895706176,
      "epoch": 1.351726875744343,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004824631195929402,
      "loss": 5.3636,
      "mean_token_accuracy": 0.15750412195920943,
      "num_tokens": 29857111.0,
      "step": 13620
    },
    {
      "entropy": 5.52392954826355,
      "epoch": 1.3522231044065105,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004824494032247911,
      "loss": 5.352,
      "mean_token_accuracy": 0.15662529617547988,
      "num_tokens": 29866919.0,
      "step": 13625
    },
    {
      "entropy": 5.583753442764282,
      "epoch": 1.352719333068678,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004824356817123439,
      "loss": 5.3005,
      "mean_token_accuracy": 0.15229649543762208,
      "num_tokens": 29878631.0,
      "step": 13630
    },
    {
      "entropy": 5.580267143249512,
      "epoch": 1.3532155617308455,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048242195505593865,
      "loss": 5.3342,
      "mean_token_accuracy": 0.15470725744962693,
      "num_tokens": 29890809.0,
      "step": 13635
    },
    {
      "entropy": 5.591029500961303,
      "epoch": 1.3537117903930131,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048240822325591596,
      "loss": 5.4233,
      "mean_token_accuracy": 0.15383840277791022,
      "num_tokens": 29901789.0,
      "step": 13640
    },
    {
      "entropy": 5.567114543914795,
      "epoch": 1.3542080190551806,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004823944863126162,
      "loss": 5.329,
      "mean_token_accuracy": 0.14945722818374635,
      "num_tokens": 29912576.0,
      "step": 13645
    },
    {
      "entropy": 5.621696329116821,
      "epoch": 1.354704247717348,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004823807442263801,
      "loss": 5.4316,
      "mean_token_accuracy": 0.1425947666168213,
      "num_tokens": 29924076.0,
      "step": 13650
    },
    {
      "entropy": 5.554328489303589,
      "epoch": 1.3552004763795158,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048236699699754845,
      "loss": 5.339,
      "mean_token_accuracy": 0.1507178619503975,
      "num_tokens": 29935854.0,
      "step": 13655
    },
    {
      "entropy": 5.603705549240113,
      "epoch": 1.3556967050416833,
      "grad_norm": 1.265625,
      "learning_rate": 0.000482353244626462,
      "loss": 5.2917,
      "mean_token_accuracy": 0.15207396149635316,
      "num_tokens": 29946389.0,
      "step": 13660
    },
    {
      "entropy": 5.595849227905274,
      "epoch": 1.3561929337038507,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048233948711346196,
      "loss": 5.3548,
      "mean_token_accuracy": 0.14622425585985183,
      "num_tokens": 29958170.0,
      "step": 13665
    },
    {
      "entropy": 5.492078685760498,
      "epoch": 1.3566891623660182,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048232572445888945,
      "loss": 5.2593,
      "mean_token_accuracy": 0.1548987239599228,
      "num_tokens": 29969067.0,
      "step": 13670
    },
    {
      "entropy": 5.5916577816009525,
      "epoch": 1.3571853910281857,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004823119566630857,
      "loss": 5.3915,
      "mean_token_accuracy": 0.14677224159240723,
      "num_tokens": 29979727.0,
      "step": 13675
    },
    {
      "entropy": 5.651455307006836,
      "epoch": 1.3576816196903534,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048229818372639216,
      "loss": 5.3591,
      "mean_token_accuracy": 0.1546714648604393,
      "num_tokens": 29989397.0,
      "step": 13680
    },
    {
      "entropy": 5.475157833099365,
      "epoch": 1.3581778483525209,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048228440564915026,
      "loss": 5.2824,
      "mean_token_accuracy": 0.15312213599681854,
      "num_tokens": 29999415.0,
      "step": 13685
    },
    {
      "entropy": 5.546486759185791,
      "epoch": 1.3586740770146883,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048227062243170186,
      "loss": 5.3112,
      "mean_token_accuracy": 0.15573932230472565,
      "num_tokens": 30010277.0,
      "step": 13690
    },
    {
      "entropy": 5.543376159667969,
      "epoch": 1.3591703056768558,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004822568340743886,
      "loss": 5.221,
      "mean_token_accuracy": 0.16213708370923996,
      "num_tokens": 30020684.0,
      "step": 13695
    },
    {
      "entropy": 5.613788318634033,
      "epoch": 1.3596665343390235,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004822430405775525,
      "loss": 5.4314,
      "mean_token_accuracy": 0.15010374933481216,
      "num_tokens": 30030916.0,
      "step": 13700
    },
    {
      "entropy": 5.631393098831177,
      "epoch": 1.360162763001191,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004822292419415355,
      "loss": 5.3586,
      "mean_token_accuracy": 0.14909230768680573,
      "num_tokens": 30041140.0,
      "step": 13705
    },
    {
      "entropy": 5.534036540985108,
      "epoch": 1.3606589916633585,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048221543816667984,
      "loss": 5.3074,
      "mean_token_accuracy": 0.16173945665359496,
      "num_tokens": 30052765.0,
      "step": 13710
    },
    {
      "entropy": 5.517458105087281,
      "epoch": 1.361155220325526,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048220162925332786,
      "loss": 5.3325,
      "mean_token_accuracy": 0.15192051827907563,
      "num_tokens": 30062511.0,
      "step": 13715
    },
    {
      "entropy": 5.689849853515625,
      "epoch": 1.3616514489876934,
      "grad_norm": 1.9453125,
      "learning_rate": 0.000482187815201822,
      "loss": 5.4442,
      "mean_token_accuracy": 0.14851779937744142,
      "num_tokens": 30073404.0,
      "step": 13720
    },
    {
      "entropy": 5.517990827560425,
      "epoch": 1.362147677649861,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048217399601250464,
      "loss": 5.2833,
      "mean_token_accuracy": 0.15863201171159744,
      "num_tokens": 30084978.0,
      "step": 13725
    },
    {
      "entropy": 5.511034870147705,
      "epoch": 1.3626439063120286,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004821601716857187,
      "loss": 5.3433,
      "mean_token_accuracy": 0.15185248404741286,
      "num_tokens": 30094498.0,
      "step": 13730
    },
    {
      "entropy": 5.551659631729126,
      "epoch": 1.363140134974196,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048214634222180686,
      "loss": 5.3627,
      "mean_token_accuracy": 0.15748399794101714,
      "num_tokens": 30104545.0,
      "step": 13735
    },
    {
      "entropy": 5.646594285964966,
      "epoch": 1.3636363636363638,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004821325076211121,
      "loss": 5.426,
      "mean_token_accuracy": 0.15163007378578186,
      "num_tokens": 30114863.0,
      "step": 13740
    },
    {
      "entropy": 5.646785116195678,
      "epoch": 1.3641325922985312,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048211866788397753,
      "loss": 5.3594,
      "mean_token_accuracy": 0.1478980928659439,
      "num_tokens": 30125101.0,
      "step": 13745
    },
    {
      "entropy": 5.5665367603302,
      "epoch": 1.3646288209606987,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004821048230107462,
      "loss": 5.4242,
      "mean_token_accuracy": 0.15588893592357636,
      "num_tokens": 30135081.0,
      "step": 13750
    },
    {
      "entropy": 5.714246416091919,
      "epoch": 1.3651250496228662,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004820909730017616,
      "loss": 5.4341,
      "mean_token_accuracy": 0.14383077919483184,
      "num_tokens": 30147169.0,
      "step": 13755
    },
    {
      "entropy": 5.620228242874146,
      "epoch": 1.3656212782850337,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048207711785736707,
      "loss": 5.2502,
      "mean_token_accuracy": 0.156867878139019,
      "num_tokens": 30157523.0,
      "step": 13760
    },
    {
      "entropy": 5.529635095596314,
      "epoch": 1.3661175069472011,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048206325757790624,
      "loss": 5.3267,
      "mean_token_accuracy": 0.15431780964136124,
      "num_tokens": 30168313.0,
      "step": 13765
    },
    {
      "entropy": 5.4616471290588375,
      "epoch": 1.3666137356093688,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048204939216372285,
      "loss": 5.281,
      "mean_token_accuracy": 0.1535404846072197,
      "num_tokens": 30179173.0,
      "step": 13770
    },
    {
      "entropy": 5.586549425125122,
      "epoch": 1.3671099642715363,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004820355216151607,
      "loss": 5.3771,
      "mean_token_accuracy": 0.1506172940135002,
      "num_tokens": 30191666.0,
      "step": 13775
    },
    {
      "entropy": 5.644274520874023,
      "epoch": 1.3676061929337038,
      "grad_norm": 2.59375,
      "learning_rate": 0.00048202164593256363,
      "loss": 5.3396,
      "mean_token_accuracy": 0.15419528633356094,
      "num_tokens": 30202752.0,
      "step": 13780
    },
    {
      "entropy": 5.583611392974854,
      "epoch": 1.3681024215958715,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048200776511627585,
      "loss": 5.3599,
      "mean_token_accuracy": 0.15034971833229066,
      "num_tokens": 30213127.0,
      "step": 13785
    },
    {
      "entropy": 5.621376466751099,
      "epoch": 1.368598650258039,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004819938791666416,
      "loss": 5.3805,
      "mean_token_accuracy": 0.15239813774824143,
      "num_tokens": 30224782.0,
      "step": 13790
    },
    {
      "entropy": 5.569597244262695,
      "epoch": 1.3690948789202064,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048197998808400524,
      "loss": 5.4011,
      "mean_token_accuracy": 0.15346841961145402,
      "num_tokens": 30235129.0,
      "step": 13795
    },
    {
      "entropy": 5.569791269302368,
      "epoch": 1.369591107582374,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004819660918687111,
      "loss": 5.4063,
      "mean_token_accuracy": 0.14484800398349762,
      "num_tokens": 30247304.0,
      "step": 13800
    },
    {
      "entropy": 5.542311763763427,
      "epoch": 1.3700873362445414,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004819521905211039,
      "loss": 5.2819,
      "mean_token_accuracy": 0.15623597502708436,
      "num_tokens": 30257900.0,
      "step": 13805
    },
    {
      "entropy": 5.627724552154541,
      "epoch": 1.370583564906709,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00048193828404152826,
      "loss": 5.4136,
      "mean_token_accuracy": 0.1583661690354347,
      "num_tokens": 30269741.0,
      "step": 13810
    },
    {
      "entropy": 5.616897869110107,
      "epoch": 1.3710797935688765,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004819243724303291,
      "loss": 5.3921,
      "mean_token_accuracy": 0.14639653638005257,
      "num_tokens": 30279382.0,
      "step": 13815
    },
    {
      "entropy": 5.602604866027832,
      "epoch": 1.371576022231044,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048191045568785144,
      "loss": 5.3649,
      "mean_token_accuracy": 0.15298472642898558,
      "num_tokens": 30289837.0,
      "step": 13820
    },
    {
      "entropy": 5.557691669464111,
      "epoch": 1.3720722508932117,
      "grad_norm": 1.78125,
      "learning_rate": 0.00048189653381444024,
      "loss": 5.337,
      "mean_token_accuracy": 0.15470173358917236,
      "num_tokens": 30302138.0,
      "step": 13825
    },
    {
      "entropy": 5.587356758117676,
      "epoch": 1.3725684795553792,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004818826068104409,
      "loss": 5.4085,
      "mean_token_accuracy": 0.15479329824447632,
      "num_tokens": 30313623.0,
      "step": 13830
    },
    {
      "entropy": 5.605863857269287,
      "epoch": 1.3730647082175467,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004818686746761987,
      "loss": 5.3493,
      "mean_token_accuracy": 0.14714564830064775,
      "num_tokens": 30324533.0,
      "step": 13835
    },
    {
      "entropy": 5.662343597412109,
      "epoch": 1.3735609368797141,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004818547374120591,
      "loss": 5.4643,
      "mean_token_accuracy": 0.14264634847640992,
      "num_tokens": 30335819.0,
      "step": 13840
    },
    {
      "entropy": 5.573651695251465,
      "epoch": 1.3740571655418816,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004818407950183678,
      "loss": 5.3422,
      "mean_token_accuracy": 0.1550154507160187,
      "num_tokens": 30345947.0,
      "step": 13845
    },
    {
      "entropy": 5.497993659973145,
      "epoch": 1.374553394204049,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004818268474954704,
      "loss": 5.3163,
      "mean_token_accuracy": 0.1511186257004738,
      "num_tokens": 30356972.0,
      "step": 13850
    },
    {
      "entropy": 5.615346527099609,
      "epoch": 1.3750496228662168,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004818128948437129,
      "loss": 5.3826,
      "mean_token_accuracy": 0.14888860434293746,
      "num_tokens": 30368414.0,
      "step": 13855
    },
    {
      "entropy": 5.6329758644104,
      "epoch": 1.3755458515283843,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004817989370634412,
      "loss": 5.3374,
      "mean_token_accuracy": 0.15524144545197488,
      "num_tokens": 30379141.0,
      "step": 13860
    },
    {
      "entropy": 5.566586303710937,
      "epoch": 1.3760420801905517,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048178497415500153,
      "loss": 5.297,
      "mean_token_accuracy": 0.1547309748828411,
      "num_tokens": 30388985.0,
      "step": 13865
    },
    {
      "entropy": 5.530737209320068,
      "epoch": 1.3765383088527194,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004817710061187401,
      "loss": 5.311,
      "mean_token_accuracy": 0.16068000346422195,
      "num_tokens": 30399599.0,
      "step": 13870
    },
    {
      "entropy": 5.5858160018920895,
      "epoch": 1.377034537514887,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048175703295500323,
      "loss": 5.4066,
      "mean_token_accuracy": 0.15034065246582032,
      "num_tokens": 30411143.0,
      "step": 13875
    },
    {
      "entropy": 5.66247410774231,
      "epoch": 1.3775307661770544,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048174305466413755,
      "loss": 5.3824,
      "mean_token_accuracy": 0.15205209702253342,
      "num_tokens": 30422339.0,
      "step": 13880
    },
    {
      "entropy": 5.579911756515503,
      "epoch": 1.3780269948392219,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048172907124648947,
      "loss": 5.3216,
      "mean_token_accuracy": 0.15470529943704606,
      "num_tokens": 30432588.0,
      "step": 13885
    },
    {
      "entropy": 5.60498423576355,
      "epoch": 1.3785232235013893,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004817150827024059,
      "loss": 5.3793,
      "mean_token_accuracy": 0.1551692932844162,
      "num_tokens": 30442272.0,
      "step": 13890
    },
    {
      "entropy": 5.546994733810425,
      "epoch": 1.379019452163557,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004817010890322338,
      "loss": 5.306,
      "mean_token_accuracy": 0.15733089596033095,
      "num_tokens": 30453142.0,
      "step": 13895
    },
    {
      "entropy": 5.556444072723389,
      "epoch": 1.3795156808257245,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048168709023632006,
      "loss": 5.3916,
      "mean_token_accuracy": 0.1506037786602974,
      "num_tokens": 30465136.0,
      "step": 13900
    },
    {
      "entropy": 5.6148961067199705,
      "epoch": 1.380011909487892,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048167308631501184,
      "loss": 5.3172,
      "mean_token_accuracy": 0.15355598628520967,
      "num_tokens": 30476511.0,
      "step": 13905
    },
    {
      "entropy": 5.602449512481689,
      "epoch": 1.3805081381500597,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048165907726865643,
      "loss": 5.427,
      "mean_token_accuracy": 0.14917163848876952,
      "num_tokens": 30486547.0,
      "step": 13910
    },
    {
      "entropy": 5.610246992111206,
      "epoch": 1.3810043668122272,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004816450630976012,
      "loss": 5.3842,
      "mean_token_accuracy": 0.15460807234048843,
      "num_tokens": 30496163.0,
      "step": 13915
    },
    {
      "entropy": 5.599671173095703,
      "epoch": 1.3815005954743946,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004816310438021937,
      "loss": 5.383,
      "mean_token_accuracy": 0.15015150755643844,
      "num_tokens": 30506979.0,
      "step": 13920
    },
    {
      "entropy": 5.62280650138855,
      "epoch": 1.381996824136562,
      "grad_norm": 1.9765625,
      "learning_rate": 0.00048161701938278156,
      "loss": 5.4122,
      "mean_token_accuracy": 0.1414540193974972,
      "num_tokens": 30517492.0,
      "step": 13925
    },
    {
      "entropy": 5.552213668823242,
      "epoch": 1.3824930527987296,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048160298983971256,
      "loss": 5.293,
      "mean_token_accuracy": 0.1595747575163841,
      "num_tokens": 30528614.0,
      "step": 13930
    },
    {
      "entropy": 5.5525267124176025,
      "epoch": 1.382989281460897,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004815889551733346,
      "loss": 5.3831,
      "mean_token_accuracy": 0.1514338254928589,
      "num_tokens": 30539479.0,
      "step": 13935
    },
    {
      "entropy": 5.62359094619751,
      "epoch": 1.3834855101230648,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00048157491538399573,
      "loss": 5.4153,
      "mean_token_accuracy": 0.14745342880487441,
      "num_tokens": 30549909.0,
      "step": 13940
    },
    {
      "entropy": 5.626338863372803,
      "epoch": 1.3839817387852322,
      "grad_norm": 2.34375,
      "learning_rate": 0.00048156087047204407,
      "loss": 5.3178,
      "mean_token_accuracy": 0.15208786129951476,
      "num_tokens": 30562265.0,
      "step": 13945
    },
    {
      "entropy": 5.596759891510009,
      "epoch": 1.3844779674473997,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004815468204378279,
      "loss": 5.3194,
      "mean_token_accuracy": 0.15172715932130815,
      "num_tokens": 30572622.0,
      "step": 13950
    },
    {
      "entropy": 5.555500173568726,
      "epoch": 1.3849741961095674,
      "grad_norm": 2.140625,
      "learning_rate": 0.0004815327652816957,
      "loss": 5.3118,
      "mean_token_accuracy": 0.15657808482646943,
      "num_tokens": 30583300.0,
      "step": 13955
    },
    {
      "entropy": 5.546760988235474,
      "epoch": 1.3854704247717349,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004815187050039959,
      "loss": 5.2754,
      "mean_token_accuracy": 0.15877688974142073,
      "num_tokens": 30594460.0,
      "step": 13960
    },
    {
      "entropy": 5.579444074630738,
      "epoch": 1.3859666534339024,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048150463960507724,
      "loss": 5.3131,
      "mean_token_accuracy": 0.15467251390218734,
      "num_tokens": 30604658.0,
      "step": 13965
    },
    {
      "entropy": 5.605820655822754,
      "epoch": 1.3864628820960698,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048149056908528855,
      "loss": 5.3072,
      "mean_token_accuracy": 0.1514147028326988,
      "num_tokens": 30616151.0,
      "step": 13970
    },
    {
      "entropy": 5.60298113822937,
      "epoch": 1.3869591107582373,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004814764934449786,
      "loss": 5.3722,
      "mean_token_accuracy": 0.15086500644683837,
      "num_tokens": 30627560.0,
      "step": 13975
    },
    {
      "entropy": 5.532004499435425,
      "epoch": 1.387455339420405,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048146241268449665,
      "loss": 5.3554,
      "mean_token_accuracy": 0.15673008412122727,
      "num_tokens": 30638490.0,
      "step": 13980
    },
    {
      "entropy": 5.584368181228638,
      "epoch": 1.3879515680825725,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004814483268041917,
      "loss": 5.4665,
      "mean_token_accuracy": 0.1469423547387123,
      "num_tokens": 30649501.0,
      "step": 13985
    },
    {
      "entropy": 5.582320785522461,
      "epoch": 1.38844779674474,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004814342358044131,
      "loss": 5.3284,
      "mean_token_accuracy": 0.1516440838575363,
      "num_tokens": 30660479.0,
      "step": 13990
    },
    {
      "entropy": 5.683701372146606,
      "epoch": 1.3889440254069074,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048142013968551026,
      "loss": 5.517,
      "mean_token_accuracy": 0.14488700032234192,
      "num_tokens": 30672015.0,
      "step": 13995
    },
    {
      "entropy": 5.538846635818482,
      "epoch": 1.3894402540690751,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048140603844783273,
      "loss": 5.3679,
      "mean_token_accuracy": 0.15511614084243774,
      "num_tokens": 30682612.0,
      "step": 14000
    },
    {
      "entropy": 5.531894302368164,
      "epoch": 1.3899364827312426,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004813919320917303,
      "loss": 5.2944,
      "mean_token_accuracy": 0.1530350148677826,
      "num_tokens": 30693289.0,
      "step": 14005
    },
    {
      "entropy": 5.568963098526001,
      "epoch": 1.39043271139341,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048137782061755265,
      "loss": 5.3367,
      "mean_token_accuracy": 0.15533661097288132,
      "num_tokens": 30705346.0,
      "step": 14010
    },
    {
      "entropy": 5.536048412322998,
      "epoch": 1.3909289400555775,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048136370402564974,
      "loss": 5.3449,
      "mean_token_accuracy": 0.15245912820100785,
      "num_tokens": 30715968.0,
      "step": 14015
    },
    {
      "entropy": 5.601094675064087,
      "epoch": 1.391425168717745,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004813495823163716,
      "loss": 5.3004,
      "mean_token_accuracy": 0.15722040235996246,
      "num_tokens": 30726225.0,
      "step": 14020
    },
    {
      "entropy": 5.553696775436402,
      "epoch": 1.3919213973799127,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004813354554900685,
      "loss": 5.2565,
      "mean_token_accuracy": 0.1596837356686592,
      "num_tokens": 30735603.0,
      "step": 14025
    },
    {
      "entropy": 5.528075790405273,
      "epoch": 1.3924176260420802,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004813213235470907,
      "loss": 5.3552,
      "mean_token_accuracy": 0.14728132858872414,
      "num_tokens": 30747414.0,
      "step": 14030
    },
    {
      "entropy": 5.645507526397705,
      "epoch": 1.3929138547042477,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004813071864877887,
      "loss": 5.3828,
      "mean_token_accuracy": 0.150152388215065,
      "num_tokens": 30758389.0,
      "step": 14035
    },
    {
      "entropy": 5.5901727199554445,
      "epoch": 1.3934100833664154,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000481293044312513,
      "loss": 5.406,
      "mean_token_accuracy": 0.14988000318408012,
      "num_tokens": 30769484.0,
      "step": 14040
    },
    {
      "entropy": 5.528121519088745,
      "epoch": 1.3939063120285828,
      "grad_norm": 1.625,
      "learning_rate": 0.00048127889702161435,
      "loss": 5.3126,
      "mean_token_accuracy": 0.1548627197742462,
      "num_tokens": 30779775.0,
      "step": 14045
    },
    {
      "entropy": 5.61658034324646,
      "epoch": 1.3944025406907503,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004812647446154435,
      "loss": 5.4101,
      "mean_token_accuracy": 0.14839381501078605,
      "num_tokens": 30791456.0,
      "step": 14050
    },
    {
      "entropy": 5.596114206314087,
      "epoch": 1.3948987693529178,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004812505870943515,
      "loss": 5.2368,
      "mean_token_accuracy": 0.15571902990341185,
      "num_tokens": 30801659.0,
      "step": 14055
    },
    {
      "entropy": 5.666314888000488,
      "epoch": 1.3953949980150853,
      "grad_norm": 1.625,
      "learning_rate": 0.0004812364244586893,
      "loss": 5.476,
      "mean_token_accuracy": 0.1522856757044792,
      "num_tokens": 30813389.0,
      "step": 14060
    },
    {
      "entropy": 5.572531795501709,
      "epoch": 1.395891226677253,
      "grad_norm": 1.375,
      "learning_rate": 0.0004812222567088082,
      "loss": 5.248,
      "mean_token_accuracy": 0.1504622831940651,
      "num_tokens": 30823725.0,
      "step": 14065
    },
    {
      "entropy": 5.549576473236084,
      "epoch": 1.3963874553394204,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004812080838450596,
      "loss": 5.2678,
      "mean_token_accuracy": 0.15489615947008134,
      "num_tokens": 30834448.0,
      "step": 14070
    },
    {
      "entropy": 5.587962484359741,
      "epoch": 1.396883684001588,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004811939058677947,
      "loss": 5.3066,
      "mean_token_accuracy": 0.15410882383584976,
      "num_tokens": 30845891.0,
      "step": 14075
    },
    {
      "entropy": 5.489449071884155,
      "epoch": 1.3973799126637554,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004811797227773654,
      "loss": 5.2329,
      "mean_token_accuracy": 0.15933302640914918,
      "num_tokens": 30855170.0,
      "step": 14080
    },
    {
      "entropy": 5.486531448364258,
      "epoch": 1.397876141325923,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048116553457412317,
      "loss": 5.3303,
      "mean_token_accuracy": 0.150497505068779,
      "num_tokens": 30866527.0,
      "step": 14085
    },
    {
      "entropy": 5.531078052520752,
      "epoch": 1.3983723699880906,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048115134125841993,
      "loss": 5.2946,
      "mean_token_accuracy": 0.15762494057416915,
      "num_tokens": 30876292.0,
      "step": 14090
    },
    {
      "entropy": 5.5675843238830565,
      "epoch": 1.398868598650258,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004811371428306076,
      "loss": 5.2867,
      "mean_token_accuracy": 0.15398220270872115,
      "num_tokens": 30887440.0,
      "step": 14095
    },
    {
      "entropy": 5.558921575546265,
      "epoch": 1.3993648273124255,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004811229392910384,
      "loss": 5.3235,
      "mean_token_accuracy": 0.15423514693975449,
      "num_tokens": 30897925.0,
      "step": 14100
    },
    {
      "entropy": 5.463570022583008,
      "epoch": 1.399861055974593,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004811087306400644,
      "loss": 5.3353,
      "mean_token_accuracy": 0.16134523153305053,
      "num_tokens": 30908413.0,
      "step": 14105
    },
    {
      "entropy": 5.613696908950805,
      "epoch": 1.4003572846367607,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048109451687803804,
      "loss": 5.3825,
      "mean_token_accuracy": 0.15281220376491547,
      "num_tokens": 30919149.0,
      "step": 14110
    },
    {
      "entropy": 5.628803539276123,
      "epoch": 1.4008535132989282,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004810802980053117,
      "loss": 5.3956,
      "mean_token_accuracy": 0.15678383260965348,
      "num_tokens": 30929883.0,
      "step": 14115
    },
    {
      "entropy": 5.558426809310913,
      "epoch": 1.4013497419610956,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048106607402223815,
      "loss": 5.3185,
      "mean_token_accuracy": 0.1513163909316063,
      "num_tokens": 30941358.0,
      "step": 14120
    },
    {
      "entropy": 5.593570423126221,
      "epoch": 1.4018459706232633,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048105184492916995,
      "loss": 5.3268,
      "mean_token_accuracy": 0.15071033090353012,
      "num_tokens": 30953289.0,
      "step": 14125
    },
    {
      "entropy": 5.545131874084473,
      "epoch": 1.4023421992854308,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048103761072645993,
      "loss": 5.2479,
      "mean_token_accuracy": 0.1568477965891361,
      "num_tokens": 30964304.0,
      "step": 14130
    },
    {
      "entropy": 5.5043998718261715,
      "epoch": 1.4028384279475983,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048102337141446116,
      "loss": 5.2729,
      "mean_token_accuracy": 0.15719584077596666,
      "num_tokens": 30974840.0,
      "step": 14135
    },
    {
      "entropy": 5.549101734161377,
      "epoch": 1.4033346566097658,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00048100912699352666,
      "loss": 5.3514,
      "mean_token_accuracy": 0.15358481854200362,
      "num_tokens": 30985805.0,
      "step": 14140
    },
    {
      "entropy": 5.584851837158203,
      "epoch": 1.4038308852719332,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004809948774640098,
      "loss": 5.3525,
      "mean_token_accuracy": 0.15098376870155333,
      "num_tokens": 30998174.0,
      "step": 14145
    },
    {
      "entropy": 5.569659471511841,
      "epoch": 1.4043271139341007,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004809806228262639,
      "loss": 5.3598,
      "mean_token_accuracy": 0.1506463333964348,
      "num_tokens": 31010050.0,
      "step": 14150
    },
    {
      "entropy": 5.661315393447876,
      "epoch": 1.4048233425962684,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00048096636308064236,
      "loss": 5.4786,
      "mean_token_accuracy": 0.14449890106916427,
      "num_tokens": 31023069.0,
      "step": 14155
    },
    {
      "entropy": 5.600331830978393,
      "epoch": 1.4053195712584359,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004809520982274989,
      "loss": 5.2677,
      "mean_token_accuracy": 0.15784199237823487,
      "num_tokens": 31033512.0,
      "step": 14160
    },
    {
      "entropy": 5.546965837478638,
      "epoch": 1.4058157999206033,
      "grad_norm": 1.375,
      "learning_rate": 0.00048093782826718706,
      "loss": 5.3305,
      "mean_token_accuracy": 0.14564902931451798,
      "num_tokens": 31045731.0,
      "step": 14165
    },
    {
      "entropy": 5.502837228775024,
      "epoch": 1.406312028582771,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048092355320006086,
      "loss": 5.2403,
      "mean_token_accuracy": 0.16150530129671098,
      "num_tokens": 31055443.0,
      "step": 14170
    },
    {
      "entropy": 5.530621194839478,
      "epoch": 1.4068082572449385,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004809092730264744,
      "loss": 5.3409,
      "mean_token_accuracy": 0.15543268471956254,
      "num_tokens": 31067575.0,
      "step": 14175
    },
    {
      "entropy": 5.558857011795044,
      "epoch": 1.407304485907106,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048089498774678155,
      "loss": 5.4059,
      "mean_token_accuracy": 0.15019888877868653,
      "num_tokens": 31078198.0,
      "step": 14180
    },
    {
      "entropy": 5.506910896301269,
      "epoch": 1.4078007145692735,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048088069736133664,
      "loss": 5.3159,
      "mean_token_accuracy": 0.15462663173675537,
      "num_tokens": 31089961.0,
      "step": 14185
    },
    {
      "entropy": 5.541952753067017,
      "epoch": 1.408296943231441,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004808664018704941,
      "loss": 5.3306,
      "mean_token_accuracy": 0.15529872626066207,
      "num_tokens": 31100534.0,
      "step": 14190
    },
    {
      "entropy": 5.685495471954345,
      "epoch": 1.4087931718936086,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004808521012746085,
      "loss": 5.3687,
      "mean_token_accuracy": 0.15970780998468398,
      "num_tokens": 31112016.0,
      "step": 14195
    },
    {
      "entropy": 5.5436633110046385,
      "epoch": 1.4092894005557761,
      "grad_norm": 2.109375,
      "learning_rate": 0.00048083779557403424,
      "loss": 5.3843,
      "mean_token_accuracy": 0.1505210742354393,
      "num_tokens": 31123659.0,
      "step": 14200
    },
    {
      "entropy": 5.526986837387085,
      "epoch": 1.4097856292179436,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004808234847691262,
      "loss": 5.3162,
      "mean_token_accuracy": 0.14779610335826873,
      "num_tokens": 31133872.0,
      "step": 14205
    },
    {
      "entropy": 5.606729030609131,
      "epoch": 1.4102818578801113,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004808091688602393,
      "loss": 5.3586,
      "mean_token_accuracy": 0.15057032108306884,
      "num_tokens": 31145834.0,
      "step": 14210
    },
    {
      "entropy": 5.628369283676148,
      "epoch": 1.4107780865422788,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048079484784772854,
      "loss": 5.4396,
      "mean_token_accuracy": 0.1483740448951721,
      "num_tokens": 31157037.0,
      "step": 14215
    },
    {
      "entropy": 5.547802448272705,
      "epoch": 1.4112743152044462,
      "grad_norm": 1.75,
      "learning_rate": 0.00048078052173194886,
      "loss": 5.3204,
      "mean_token_accuracy": 0.15650108605623245,
      "num_tokens": 31166812.0,
      "step": 14220
    },
    {
      "entropy": 5.532660436630249,
      "epoch": 1.4117705438666137,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004807661905132557,
      "loss": 5.3103,
      "mean_token_accuracy": 0.1539248913526535,
      "num_tokens": 31177162.0,
      "step": 14225
    },
    {
      "entropy": 5.577081298828125,
      "epoch": 1.4122667725287812,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004807518541920045,
      "loss": 5.333,
      "mean_token_accuracy": 0.15340053290128708,
      "num_tokens": 31187333.0,
      "step": 14230
    },
    {
      "entropy": 5.6078712940216064,
      "epoch": 1.4127630011909487,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048073751276855056,
      "loss": 5.3628,
      "mean_token_accuracy": 0.15015761703252792,
      "num_tokens": 31197869.0,
      "step": 14235
    },
    {
      "entropy": 5.546566963195801,
      "epoch": 1.4132592298531164,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004807231662432497,
      "loss": 5.3738,
      "mean_token_accuracy": 0.1499041885137558,
      "num_tokens": 31208199.0,
      "step": 14240
    },
    {
      "entropy": 5.555455207824707,
      "epoch": 1.4137554585152838,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004807088146164575,
      "loss": 5.3414,
      "mean_token_accuracy": 0.15058445557951927,
      "num_tokens": 31218903.0,
      "step": 14245
    },
    {
      "entropy": 5.582147026062012,
      "epoch": 1.4142516871774513,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004806944578885301,
      "loss": 5.3561,
      "mean_token_accuracy": 0.14838663637638091,
      "num_tokens": 31231146.0,
      "step": 14250
    },
    {
      "entropy": 5.586977863311768,
      "epoch": 1.414747915839619,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048068009605982333,
      "loss": 5.2843,
      "mean_token_accuracy": 0.15814200788736343,
      "num_tokens": 31240980.0,
      "step": 14255
    },
    {
      "entropy": 5.5635847568511965,
      "epoch": 1.4152441445017865,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048066572913069334,
      "loss": 5.3325,
      "mean_token_accuracy": 0.15335536003112793,
      "num_tokens": 31252458.0,
      "step": 14260
    },
    {
      "entropy": 5.492844295501709,
      "epoch": 1.415740373163954,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004806513571014965,
      "loss": 5.3063,
      "mean_token_accuracy": 0.15073698610067368,
      "num_tokens": 31262979.0,
      "step": 14265
    },
    {
      "entropy": 5.525537443161011,
      "epoch": 1.4162366018261214,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004806369799725891,
      "loss": 5.3097,
      "mean_token_accuracy": 0.1508309468626976,
      "num_tokens": 31274811.0,
      "step": 14270
    },
    {
      "entropy": 5.559125232696533,
      "epoch": 1.416732830488289,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004806225977443277,
      "loss": 5.2776,
      "mean_token_accuracy": 0.15782931447029114,
      "num_tokens": 31285573.0,
      "step": 14275
    },
    {
      "entropy": 5.599280309677124,
      "epoch": 1.4172290591504566,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004806082104170689,
      "loss": 5.4482,
      "mean_token_accuracy": 0.1449013575911522,
      "num_tokens": 31296993.0,
      "step": 14280
    },
    {
      "entropy": 5.514555168151856,
      "epoch": 1.417725287812624,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00048059381799116967,
      "loss": 5.3117,
      "mean_token_accuracy": 0.15488311201333999,
      "num_tokens": 31307217.0,
      "step": 14285
    },
    {
      "entropy": 5.535370492935181,
      "epoch": 1.4182215164747916,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004805794204669867,
      "loss": 5.3538,
      "mean_token_accuracy": 0.15366009324789048,
      "num_tokens": 31317939.0,
      "step": 14290
    },
    {
      "entropy": 5.637701463699341,
      "epoch": 1.418717745136959,
      "grad_norm": 1.5,
      "learning_rate": 0.00048056501784487707,
      "loss": 5.3283,
      "mean_token_accuracy": 0.15120181441307068,
      "num_tokens": 31327731.0,
      "step": 14295
    },
    {
      "entropy": 5.700837087631226,
      "epoch": 1.4192139737991267,
      "grad_norm": 1.640625,
      "learning_rate": 0.000480550610125198,
      "loss": 5.5185,
      "mean_token_accuracy": 0.13965979143977164,
      "num_tokens": 31339388.0,
      "step": 14300
    },
    {
      "entropy": 5.5222681045532225,
      "epoch": 1.4197102024612942,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004805361973083066,
      "loss": 5.2003,
      "mean_token_accuracy": 0.15811712741851808,
      "num_tokens": 31350186.0,
      "step": 14305
    },
    {
      "entropy": 5.525717639923096,
      "epoch": 1.4202064311234617,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004805217793945605,
      "loss": 5.3528,
      "mean_token_accuracy": 0.15049481391906738,
      "num_tokens": 31361219.0,
      "step": 14310
    },
    {
      "entropy": 5.540582275390625,
      "epoch": 1.4207026597856292,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004805073563843171,
      "loss": 5.3884,
      "mean_token_accuracy": 0.1503108784556389,
      "num_tokens": 31372820.0,
      "step": 14315
    },
    {
      "entropy": 5.57339653968811,
      "epoch": 1.4211988884477966,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048049292827793413,
      "loss": 5.2491,
      "mean_token_accuracy": 0.15929470360279083,
      "num_tokens": 31382779.0,
      "step": 14320
    },
    {
      "entropy": 5.619552946090698,
      "epoch": 1.4216951171099643,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004804784950757693,
      "loss": 5.486,
      "mean_token_accuracy": 0.14985369071364402,
      "num_tokens": 31394872.0,
      "step": 14325
    },
    {
      "entropy": 5.577064847946167,
      "epoch": 1.4221913457721318,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048046405677818057,
      "loss": 5.3773,
      "mean_token_accuracy": 0.1566290318965912,
      "num_tokens": 31405767.0,
      "step": 14330
    },
    {
      "entropy": 5.623222017288208,
      "epoch": 1.4226875744342993,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048044961338552595,
      "loss": 5.3833,
      "mean_token_accuracy": 0.15162235349416733,
      "num_tokens": 31417255.0,
      "step": 14335
    },
    {
      "entropy": 5.5749755859375,
      "epoch": 1.423183803096467,
      "grad_norm": 1.375,
      "learning_rate": 0.0004804351648981636,
      "loss": 5.3256,
      "mean_token_accuracy": 0.1577051267027855,
      "num_tokens": 31428494.0,
      "step": 14340
    },
    {
      "entropy": 5.538766098022461,
      "epoch": 1.4236800317586344,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004804207113164519,
      "loss": 5.2372,
      "mean_token_accuracy": 0.15720331519842148,
      "num_tokens": 31439221.0,
      "step": 14345
    },
    {
      "entropy": 5.568926858901977,
      "epoch": 1.424176260420802,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048040625264074915,
      "loss": 5.3117,
      "mean_token_accuracy": 0.1559128239750862,
      "num_tokens": 31450575.0,
      "step": 14350
    },
    {
      "entropy": 5.444109535217285,
      "epoch": 1.4246724890829694,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048039178887141404,
      "loss": 5.3232,
      "mean_token_accuracy": 0.15457070469856263,
      "num_tokens": 31462168.0,
      "step": 14355
    },
    {
      "entropy": 5.583235406875611,
      "epoch": 1.4251687177451369,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048037732000880503,
      "loss": 5.334,
      "mean_token_accuracy": 0.153232441842556,
      "num_tokens": 31472619.0,
      "step": 14360
    },
    {
      "entropy": 5.660704135894775,
      "epoch": 1.4256649464073046,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048036284605328106,
      "loss": 5.4196,
      "mean_token_accuracy": 0.15342248231172562,
      "num_tokens": 31485246.0,
      "step": 14365
    },
    {
      "entropy": 5.5711097240448,
      "epoch": 1.426161175069472,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004803483670052011,
      "loss": 5.3285,
      "mean_token_accuracy": 0.14840276837348937,
      "num_tokens": 31496258.0,
      "step": 14370
    },
    {
      "entropy": 5.532622575759888,
      "epoch": 1.4266574037316395,
      "grad_norm": 2.09375,
      "learning_rate": 0.000480333882864924,
      "loss": 5.3119,
      "mean_token_accuracy": 0.15068431794643403,
      "num_tokens": 31506671.0,
      "step": 14375
    },
    {
      "entropy": 5.561853075027466,
      "epoch": 1.427153632393807,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004803193936328092,
      "loss": 5.3499,
      "mean_token_accuracy": 0.14743599593639373,
      "num_tokens": 31517001.0,
      "step": 14380
    },
    {
      "entropy": 5.588726043701172,
      "epoch": 1.4276498610559747,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004803048993092158,
      "loss": 5.3477,
      "mean_token_accuracy": 0.15741432160139085,
      "num_tokens": 31529649.0,
      "step": 14385
    },
    {
      "entropy": 5.5168187618255615,
      "epoch": 1.4281460897181422,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048029039989450317,
      "loss": 5.2826,
      "mean_token_accuracy": 0.1582947239279747,
      "num_tokens": 31540083.0,
      "step": 14390
    },
    {
      "entropy": 5.608189249038697,
      "epoch": 1.4286423183803096,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048027589538903115,
      "loss": 5.3902,
      "mean_token_accuracy": 0.15284068435430526,
      "num_tokens": 31551400.0,
      "step": 14395
    },
    {
      "entropy": 5.59791841506958,
      "epoch": 1.4291385470424771,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004802613857931592,
      "loss": 5.3739,
      "mean_token_accuracy": 0.14777264446020127,
      "num_tokens": 31562430.0,
      "step": 14400
    },
    {
      "entropy": 5.58103380203247,
      "epoch": 1.4296347757046446,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004802468711072472,
      "loss": 5.4061,
      "mean_token_accuracy": 0.15546592473983764,
      "num_tokens": 31573173.0,
      "step": 14405
    },
    {
      "entropy": 5.615566396713257,
      "epoch": 1.4301310043668123,
      "grad_norm": 1.265625,
      "learning_rate": 0.000480232351331655,
      "loss": 5.3621,
      "mean_token_accuracy": 0.15319690853357315,
      "num_tokens": 31584072.0,
      "step": 14410
    },
    {
      "entropy": 5.569519805908203,
      "epoch": 1.4306272330289798,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048021782646674277,
      "loss": 5.2873,
      "mean_token_accuracy": 0.15449349731206893,
      "num_tokens": 31594993.0,
      "step": 14415
    },
    {
      "entropy": 5.5183515548706055,
      "epoch": 1.4311234616911472,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048020329651287064,
      "loss": 5.3089,
      "mean_token_accuracy": 0.15945848375558852,
      "num_tokens": 31604284.0,
      "step": 14420
    },
    {
      "entropy": 5.538793277740479,
      "epoch": 1.431619690353315,
      "grad_norm": 1.5,
      "learning_rate": 0.000480188761470399,
      "loss": 5.3855,
      "mean_token_accuracy": 0.15272359400987626,
      "num_tokens": 31616188.0,
      "step": 14425
    },
    {
      "entropy": 5.679145956039429,
      "epoch": 1.4321159190154824,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004801742213396881,
      "loss": 5.3408,
      "mean_token_accuracy": 0.15573089122772216,
      "num_tokens": 31626606.0,
      "step": 14430
    },
    {
      "entropy": 5.522057914733887,
      "epoch": 1.4326121476776499,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00048015967612109875,
      "loss": 5.279,
      "mean_token_accuracy": 0.16273462027311325,
      "num_tokens": 31638050.0,
      "step": 14435
    },
    {
      "entropy": 5.525677824020386,
      "epoch": 1.4331083763398174,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048014512581499136,
      "loss": 5.2791,
      "mean_token_accuracy": 0.15344303101301193,
      "num_tokens": 31648451.0,
      "step": 14440
    },
    {
      "entropy": 5.565181541442871,
      "epoch": 1.4336046050019848,
      "grad_norm": 1.28125,
      "learning_rate": 0.000480130570421727,
      "loss": 5.3137,
      "mean_token_accuracy": 0.14858073592185975,
      "num_tokens": 31659310.0,
      "step": 14445
    },
    {
      "entropy": 5.5714751243591305,
      "epoch": 1.4341008336641523,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004801160099416665,
      "loss": 5.3477,
      "mean_token_accuracy": 0.15222630202770232,
      "num_tokens": 31669594.0,
      "step": 14450
    },
    {
      "entropy": 5.6040528297424315,
      "epoch": 1.43459706232632,
      "grad_norm": 1.5,
      "learning_rate": 0.0004801014443751709,
      "loss": 5.3454,
      "mean_token_accuracy": 0.1464107558131218,
      "num_tokens": 31682034.0,
      "step": 14455
    },
    {
      "entropy": 5.628639125823975,
      "epoch": 1.4350932909884875,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004800868737226015,
      "loss": 5.3687,
      "mean_token_accuracy": 0.1508548989892006,
      "num_tokens": 31692390.0,
      "step": 14460
    },
    {
      "entropy": 5.5761363983154295,
      "epoch": 1.435589519650655,
      "grad_norm": 1.5,
      "learning_rate": 0.00048007229798431946,
      "loss": 5.3428,
      "mean_token_accuracy": 0.15368054956197738,
      "num_tokens": 31703690.0,
      "step": 14465
    },
    {
      "entropy": 5.597110033035278,
      "epoch": 1.4360857483128227,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004800577171606864,
      "loss": 5.3605,
      "mean_token_accuracy": 0.15428978353738784,
      "num_tokens": 31714655.0,
      "step": 14470
    },
    {
      "entropy": 5.521809387207031,
      "epoch": 1.4365819769749901,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004800431312520638,
      "loss": 5.2674,
      "mean_token_accuracy": 0.15878042280673982,
      "num_tokens": 31727022.0,
      "step": 14475
    },
    {
      "entropy": 5.558440685272217,
      "epoch": 1.4370782056371576,
      "grad_norm": 3.375,
      "learning_rate": 0.0004800285402588134,
      "loss": 5.2871,
      "mean_token_accuracy": 0.16187798827886582,
      "num_tokens": 31738071.0,
      "step": 14480
    },
    {
      "entropy": 5.618892812728882,
      "epoch": 1.437574434299325,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004800139441812969,
      "loss": 5.4155,
      "mean_token_accuracy": 0.14760954827070236,
      "num_tokens": 31749342.0,
      "step": 14485
    },
    {
      "entropy": 5.64426212310791,
      "epoch": 1.4380706629614926,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047999934301987644,
      "loss": 5.3933,
      "mean_token_accuracy": 0.15260061323642732,
      "num_tokens": 31761218.0,
      "step": 14490
    },
    {
      "entropy": 5.588413095474243,
      "epoch": 1.4385668916236602,
      "grad_norm": 1.453125,
      "learning_rate": 0.000479984736774914,
      "loss": 5.4126,
      "mean_token_accuracy": 0.15265292376279832,
      "num_tokens": 31773111.0,
      "step": 14495
    },
    {
      "entropy": 5.602659130096436,
      "epoch": 1.4390631202858277,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047997012544677176,
      "loss": 5.3723,
      "mean_token_accuracy": 0.15174679160118104,
      "num_tokens": 31782948.0,
      "step": 14500
    },
    {
      "entropy": 5.647914600372315,
      "epoch": 1.4395593489479952,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004799555090358121,
      "loss": 5.3518,
      "mean_token_accuracy": 0.15148211866617203,
      "num_tokens": 31793992.0,
      "step": 14505
    },
    {
      "entropy": 5.646610832214355,
      "epoch": 1.440055577610163,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004799408875423975,
      "loss": 5.3617,
      "mean_token_accuracy": 0.14598882794380189,
      "num_tokens": 31804830.0,
      "step": 14510
    },
    {
      "entropy": 5.555121755599975,
      "epoch": 1.4405518062723304,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004799262609668904,
      "loss": 5.3469,
      "mean_token_accuracy": 0.15870327800512313,
      "num_tokens": 31817200.0,
      "step": 14515
    },
    {
      "entropy": 5.566103601455689,
      "epoch": 1.4410480349344978,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004799116293096537,
      "loss": 5.3877,
      "mean_token_accuracy": 0.15093349739909173,
      "num_tokens": 31828226.0,
      "step": 14520
    },
    {
      "entropy": 5.56671462059021,
      "epoch": 1.4415442635966653,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004798969925710501,
      "loss": 5.272,
      "mean_token_accuracy": 0.1600637137889862,
      "num_tokens": 31839635.0,
      "step": 14525
    },
    {
      "entropy": 5.552546167373658,
      "epoch": 1.4420404922588328,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047988235075144254,
      "loss": 5.3017,
      "mean_token_accuracy": 0.15718430727720262,
      "num_tokens": 31850333.0,
      "step": 14530
    },
    {
      "entropy": 5.5687305450439455,
      "epoch": 1.4425367209210003,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004798677038511942,
      "loss": 5.2852,
      "mean_token_accuracy": 0.1552279844880104,
      "num_tokens": 31861002.0,
      "step": 14535
    },
    {
      "entropy": 5.609229946136475,
      "epoch": 1.443032949583168,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004798530518706683,
      "loss": 5.376,
      "mean_token_accuracy": 0.14877621084451675,
      "num_tokens": 31871713.0,
      "step": 14540
    },
    {
      "entropy": 5.579813051223755,
      "epoch": 1.4435291782453354,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004798383948102281,
      "loss": 5.3037,
      "mean_token_accuracy": 0.15603711009025573,
      "num_tokens": 31882381.0,
      "step": 14545
    },
    {
      "entropy": 5.638813114166259,
      "epoch": 1.444025406907503,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004798237326702371,
      "loss": 5.3401,
      "mean_token_accuracy": 0.15507427379488944,
      "num_tokens": 31892347.0,
      "step": 14550
    },
    {
      "entropy": 5.661196279525757,
      "epoch": 1.4445216355696706,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047980906545105884,
      "loss": 5.4468,
      "mean_token_accuracy": 0.14440320283174515,
      "num_tokens": 31902751.0,
      "step": 14555
    },
    {
      "entropy": 5.537927961349487,
      "epoch": 1.445017864231838,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004797943931530571,
      "loss": 5.312,
      "mean_token_accuracy": 0.15986493676900865,
      "num_tokens": 31913294.0,
      "step": 14560
    },
    {
      "entropy": 5.559374713897705,
      "epoch": 1.4455140928940056,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004797797157765958,
      "loss": 5.4234,
      "mean_token_accuracy": 0.15450235903263093,
      "num_tokens": 31925678.0,
      "step": 14565
    },
    {
      "entropy": 5.543670177459717,
      "epoch": 1.446010321556173,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004797650333220387,
      "loss": 5.242,
      "mean_token_accuracy": 0.1535148486495018,
      "num_tokens": 31934578.0,
      "step": 14570
    },
    {
      "entropy": 5.550230646133423,
      "epoch": 1.4465065502183405,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047975034578975006,
      "loss": 5.2961,
      "mean_token_accuracy": 0.15362483859062195,
      "num_tokens": 31945482.0,
      "step": 14575
    },
    {
      "entropy": 5.5912388324737545,
      "epoch": 1.4470027788805082,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000479735653180094,
      "loss": 5.427,
      "mean_token_accuracy": 0.156046162545681,
      "num_tokens": 31957303.0,
      "step": 14580
    },
    {
      "entropy": 5.590869808197022,
      "epoch": 1.4474990075426757,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004797209554934349,
      "loss": 5.3347,
      "mean_token_accuracy": 0.15141113698482514,
      "num_tokens": 31968601.0,
      "step": 14585
    },
    {
      "entropy": 5.5783453464508055,
      "epoch": 1.4479952362048432,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004797062527301373,
      "loss": 5.3483,
      "mean_token_accuracy": 0.15204426944255828,
      "num_tokens": 31978987.0,
      "step": 14590
    },
    {
      "entropy": 5.559156322479248,
      "epoch": 1.4484914648670106,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004796915448905657,
      "loss": 5.3493,
      "mean_token_accuracy": 0.15294792056083678,
      "num_tokens": 31990950.0,
      "step": 14595
    },
    {
      "entropy": 5.660120344161987,
      "epoch": 1.4489876935291783,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047967683197508486,
      "loss": 5.3654,
      "mean_token_accuracy": 0.15034189820289612,
      "num_tokens": 32001795.0,
      "step": 14600
    },
    {
      "entropy": 5.617771100997925,
      "epoch": 1.4494839221913458,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047966211398405966,
      "loss": 5.3766,
      "mean_token_accuracy": 0.143355805426836,
      "num_tokens": 32011775.0,
      "step": 14605
    },
    {
      "entropy": 5.686474180221557,
      "epoch": 1.4499801508535133,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000479647390917855,
      "loss": 5.4674,
      "mean_token_accuracy": 0.1400725245475769,
      "num_tokens": 32022674.0,
      "step": 14610
    },
    {
      "entropy": 5.623770666122437,
      "epoch": 1.4504763795156808,
      "grad_norm": 2.015625,
      "learning_rate": 0.000479632662776836,
      "loss": 5.4115,
      "mean_token_accuracy": 0.15417221039533616,
      "num_tokens": 32033383.0,
      "step": 14615
    },
    {
      "entropy": 5.5864729404449465,
      "epoch": 1.4509726081778482,
      "grad_norm": 1.375,
      "learning_rate": 0.000479617929561368,
      "loss": 5.2994,
      "mean_token_accuracy": 0.15602629780769348,
      "num_tokens": 32044066.0,
      "step": 14620
    },
    {
      "entropy": 5.5603358268737795,
      "epoch": 1.451468836840016,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004796031912718162,
      "loss": 5.3772,
      "mean_token_accuracy": 0.14848840981721878,
      "num_tokens": 32054650.0,
      "step": 14625
    },
    {
      "entropy": 5.667237377166748,
      "epoch": 1.4519650655021834,
      "grad_norm": 2.0625,
      "learning_rate": 0.00047958844790854606,
      "loss": 5.388,
      "mean_token_accuracy": 0.1493011385202408,
      "num_tokens": 32065786.0,
      "step": 14630
    },
    {
      "entropy": 5.618560028076172,
      "epoch": 1.4524612941643509,
      "grad_norm": 1.375,
      "learning_rate": 0.00047957369947192337,
      "loss": 5.3306,
      "mean_token_accuracy": 0.15982178449630738,
      "num_tokens": 32076326.0,
      "step": 14635
    },
    {
      "entropy": 5.503454065322876,
      "epoch": 1.4529575228265186,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004795589459623136,
      "loss": 5.2959,
      "mean_token_accuracy": 0.14781116396188737,
      "num_tokens": 32086915.0,
      "step": 14640
    },
    {
      "entropy": 5.558610725402832,
      "epoch": 1.453453751488686,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047954418738008295,
      "loss": 5.2463,
      "mean_token_accuracy": 0.15749150216579438,
      "num_tokens": 32096614.0,
      "step": 14645
    },
    {
      "entropy": 5.6073802471160885,
      "epoch": 1.4539499801508535,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004795294237255971,
      "loss": 5.3555,
      "mean_token_accuracy": 0.15531847327947618,
      "num_tokens": 32107721.0,
      "step": 14650
    },
    {
      "entropy": 5.5569833278656,
      "epoch": 1.454446208813021,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047951465499922234,
      "loss": 5.3394,
      "mean_token_accuracy": 0.14927127063274384,
      "num_tokens": 32116621.0,
      "step": 14655
    },
    {
      "entropy": 5.577189970016479,
      "epoch": 1.4549424374751885,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004794998812013248,
      "loss": 5.3754,
      "mean_token_accuracy": 0.1520335778594017,
      "num_tokens": 32126536.0,
      "step": 14660
    },
    {
      "entropy": 5.634997606277466,
      "epoch": 1.4554386661373562,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047948510233227084,
      "loss": 5.4224,
      "mean_token_accuracy": 0.14677132964134215,
      "num_tokens": 32137059.0,
      "step": 14665
    },
    {
      "entropy": 5.568421459197998,
      "epoch": 1.4559348947995236,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004794703183924269,
      "loss": 5.3574,
      "mean_token_accuracy": 0.15463690757751464,
      "num_tokens": 32148186.0,
      "step": 14670
    },
    {
      "entropy": 5.620283460617065,
      "epoch": 1.4564311234616911,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004794555293821597,
      "loss": 5.4059,
      "mean_token_accuracy": 0.14564592242240906,
      "num_tokens": 32158244.0,
      "step": 14675
    },
    {
      "entropy": 5.590705919265747,
      "epoch": 1.4569273521238586,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000479440735301836,
      "loss": 5.2892,
      "mean_token_accuracy": 0.15670703798532487,
      "num_tokens": 32168404.0,
      "step": 14680
    },
    {
      "entropy": 5.528975248336792,
      "epoch": 1.4574235807860263,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004794259361518226,
      "loss": 5.2636,
      "mean_token_accuracy": 0.15128226280212403,
      "num_tokens": 32178744.0,
      "step": 14685
    },
    {
      "entropy": 5.551606273651123,
      "epoch": 1.4579198094481938,
      "grad_norm": 2.1875,
      "learning_rate": 0.00047941113193248643,
      "loss": 5.3909,
      "mean_token_accuracy": 0.14814578294754027,
      "num_tokens": 32189948.0,
      "step": 14690
    },
    {
      "entropy": 5.611333799362183,
      "epoch": 1.4584160381103612,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004793963226441947,
      "loss": 5.3764,
      "mean_token_accuracy": 0.149867282807827,
      "num_tokens": 32200979.0,
      "step": 14695
    },
    {
      "entropy": 5.556320333480835,
      "epoch": 1.4589122667725287,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047938150828731455,
      "loss": 5.3532,
      "mean_token_accuracy": 0.15096893012523652,
      "num_tokens": 32211569.0,
      "step": 14700
    },
    {
      "entropy": 5.577615356445312,
      "epoch": 1.4594084954346962,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004793666888622134,
      "loss": 5.3347,
      "mean_token_accuracy": 0.1519345074892044,
      "num_tokens": 32222429.0,
      "step": 14705
    },
    {
      "entropy": 5.623196363449097,
      "epoch": 1.459904724096864,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004793518643692587,
      "loss": 5.4358,
      "mean_token_accuracy": 0.14705090075731278,
      "num_tokens": 32234332.0,
      "step": 14710
    },
    {
      "entropy": 5.667341041564941,
      "epoch": 1.4604009527590314,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004793370348088182,
      "loss": 5.4306,
      "mean_token_accuracy": 0.15582174211740493,
      "num_tokens": 32244798.0,
      "step": 14715
    },
    {
      "entropy": 5.6184672832489015,
      "epoch": 1.4608971814211988,
      "grad_norm": 2.609375,
      "learning_rate": 0.0004793222001812595,
      "loss": 5.3472,
      "mean_token_accuracy": 0.15409290343523024,
      "num_tokens": 32257746.0,
      "step": 14720
    },
    {
      "entropy": 5.640293169021606,
      "epoch": 1.4613934100833665,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047930736048695045,
      "loss": 5.4445,
      "mean_token_accuracy": 0.14749422371387483,
      "num_tokens": 32270692.0,
      "step": 14725
    },
    {
      "entropy": 5.583708047866821,
      "epoch": 1.461889638745534,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004792925157262592,
      "loss": 5.33,
      "mean_token_accuracy": 0.15089503452181816,
      "num_tokens": 32281350.0,
      "step": 14730
    },
    {
      "entropy": 5.569955730438233,
      "epoch": 1.4623858674077015,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004792776658995537,
      "loss": 5.3601,
      "mean_token_accuracy": 0.1528308004140854,
      "num_tokens": 32293182.0,
      "step": 14735
    },
    {
      "entropy": 5.596712636947632,
      "epoch": 1.462882096069869,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004792628110072023,
      "loss": 5.3451,
      "mean_token_accuracy": 0.14926338493824004,
      "num_tokens": 32304903.0,
      "step": 14740
    },
    {
      "entropy": 5.679181671142578,
      "epoch": 1.4633783247320364,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004792479510495734,
      "loss": 5.5108,
      "mean_token_accuracy": 0.1481003351509571,
      "num_tokens": 32316074.0,
      "step": 14745
    },
    {
      "entropy": 5.564502811431884,
      "epoch": 1.463874553394204,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047923308602703535,
      "loss": 5.3276,
      "mean_token_accuracy": 0.16027457565069197,
      "num_tokens": 32327594.0,
      "step": 14750
    },
    {
      "entropy": 5.541444921493531,
      "epoch": 1.4643707820563716,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047921821593995684,
      "loss": 5.3612,
      "mean_token_accuracy": 0.15362461656332016,
      "num_tokens": 32339224.0,
      "step": 14755
    },
    {
      "entropy": 5.574941825866699,
      "epoch": 1.464867010718539,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004792033407887066,
      "loss": 5.3418,
      "mean_token_accuracy": 0.15104879140853883,
      "num_tokens": 32349451.0,
      "step": 14760
    },
    {
      "entropy": 5.6612025737762455,
      "epoch": 1.4653632393807066,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004791884605736536,
      "loss": 5.4268,
      "mean_token_accuracy": 0.14396267086267472,
      "num_tokens": 32361177.0,
      "step": 14765
    },
    {
      "entropy": 5.588083171844483,
      "epoch": 1.4658594680428743,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047917357529516676,
      "loss": 5.3339,
      "mean_token_accuracy": 0.14689060151576996,
      "num_tokens": 32372031.0,
      "step": 14770
    },
    {
      "entropy": 5.557452726364136,
      "epoch": 1.4663556967050417,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004791586849536152,
      "loss": 5.2605,
      "mean_token_accuracy": 0.15992897152900695,
      "num_tokens": 32383936.0,
      "step": 14775
    },
    {
      "entropy": 5.561427164077759,
      "epoch": 1.4668519253672092,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004791437895493682,
      "loss": 5.3922,
      "mean_token_accuracy": 0.1565810650587082,
      "num_tokens": 32394283.0,
      "step": 14780
    },
    {
      "entropy": 5.567660093307495,
      "epoch": 1.4673481540293767,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047912888908279507,
      "loss": 5.3843,
      "mean_token_accuracy": 0.15146982073783874,
      "num_tokens": 32405549.0,
      "step": 14785
    },
    {
      "entropy": 5.577366495132447,
      "epoch": 1.4678443826915442,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004791139835542653,
      "loss": 5.29,
      "mean_token_accuracy": 0.1522230625152588,
      "num_tokens": 32414896.0,
      "step": 14790
    },
    {
      "entropy": 5.57419719696045,
      "epoch": 1.4683406113537119,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00047909907296414864,
      "loss": 5.3932,
      "mean_token_accuracy": 0.15184588730335236,
      "num_tokens": 32426863.0,
      "step": 14795
    },
    {
      "entropy": 5.513531160354614,
      "epoch": 1.4688368400158793,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00047908415731281475,
      "loss": 5.3421,
      "mean_token_accuracy": 0.15287913531064987,
      "num_tokens": 32437703.0,
      "step": 14800
    },
    {
      "entropy": 5.604316139221192,
      "epoch": 1.4693330686780468,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047906923660063344,
      "loss": 5.4031,
      "mean_token_accuracy": 0.15393733829259873,
      "num_tokens": 32448517.0,
      "step": 14805
    },
    {
      "entropy": 5.587944507598877,
      "epoch": 1.4698292973402145,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000479054310827975,
      "loss": 5.3522,
      "mean_token_accuracy": 0.15401667281985282,
      "num_tokens": 32458782.0,
      "step": 14810
    },
    {
      "entropy": 5.5495370388031,
      "epoch": 1.470325526002382,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047903937999520913,
      "loss": 5.2396,
      "mean_token_accuracy": 0.15738397389650344,
      "num_tokens": 32470113.0,
      "step": 14815
    },
    {
      "entropy": 5.581517171859741,
      "epoch": 1.4708217546645495,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004790244441027064,
      "loss": 5.3371,
      "mean_token_accuracy": 0.15352957397699357,
      "num_tokens": 32482455.0,
      "step": 14820
    },
    {
      "entropy": 5.624271202087402,
      "epoch": 1.471317983326717,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004790095031508371,
      "loss": 5.3904,
      "mean_token_accuracy": 0.15690432041883468,
      "num_tokens": 32493536.0,
      "step": 14825
    },
    {
      "entropy": 5.526708793640137,
      "epoch": 1.4718142119888844,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047899455713997164,
      "loss": 5.281,
      "mean_token_accuracy": 0.15621823221445083,
      "num_tokens": 32503991.0,
      "step": 14830
    },
    {
      "entropy": 5.557015895843506,
      "epoch": 1.4723104406510519,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047897960607048084,
      "loss": 5.3797,
      "mean_token_accuracy": 0.15110863894224166,
      "num_tokens": 32515474.0,
      "step": 14835
    },
    {
      "entropy": 5.533372116088867,
      "epoch": 1.4728066693132196,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004789646499427353,
      "loss": 5.2681,
      "mean_token_accuracy": 0.15972076803445817,
      "num_tokens": 32526751.0,
      "step": 14840
    },
    {
      "entropy": 5.637294435501099,
      "epoch": 1.473302897975387,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004789496887571059,
      "loss": 5.4489,
      "mean_token_accuracy": 0.1464750736951828,
      "num_tokens": 32538191.0,
      "step": 14845
    },
    {
      "entropy": 5.602173662185669,
      "epoch": 1.4737991266375545,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047893472251396373,
      "loss": 5.3631,
      "mean_token_accuracy": 0.15140637308359145,
      "num_tokens": 32548841.0,
      "step": 14850
    },
    {
      "entropy": 5.551118278503418,
      "epoch": 1.4742953552997222,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047891975121368,
      "loss": 5.3046,
      "mean_token_accuracy": 0.15708068013191223,
      "num_tokens": 32558599.0,
      "step": 14855
    },
    {
      "entropy": 5.525722026824951,
      "epoch": 1.4747915839618897,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047890477485662563,
      "loss": 5.2653,
      "mean_token_accuracy": 0.16286808401346206,
      "num_tokens": 32568641.0,
      "step": 14860
    },
    {
      "entropy": 5.588715648651123,
      "epoch": 1.4752878126240572,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047888979344317236,
      "loss": 5.3594,
      "mean_token_accuracy": 0.1518983229994774,
      "num_tokens": 32578924.0,
      "step": 14865
    },
    {
      "entropy": 5.612540006637573,
      "epoch": 1.4757840412862246,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047887480697369155,
      "loss": 5.3365,
      "mean_token_accuracy": 0.15630028247833253,
      "num_tokens": 32590233.0,
      "step": 14870
    },
    {
      "entropy": 5.611448478698731,
      "epoch": 1.4762802699483921,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047885981544855475,
      "loss": 5.3805,
      "mean_token_accuracy": 0.15200439542531968,
      "num_tokens": 32600782.0,
      "step": 14875
    },
    {
      "entropy": 5.5485306739807125,
      "epoch": 1.4767764986105598,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004788448188681339,
      "loss": 5.3376,
      "mean_token_accuracy": 0.15879722237586974,
      "num_tokens": 32612611.0,
      "step": 14880
    },
    {
      "entropy": 5.56446647644043,
      "epoch": 1.4772727272727273,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004788298172328007,
      "loss": 5.4009,
      "mean_token_accuracy": 0.15555647611618043,
      "num_tokens": 32623436.0,
      "step": 14885
    },
    {
      "entropy": 5.543827056884766,
      "epoch": 1.4777689559348948,
      "grad_norm": 2.09375,
      "learning_rate": 0.00047881481054292736,
      "loss": 5.2543,
      "mean_token_accuracy": 0.15897083431482315,
      "num_tokens": 32633356.0,
      "step": 14890
    },
    {
      "entropy": 5.572874450683594,
      "epoch": 1.4782651845970625,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047879979879888584,
      "loss": 5.3321,
      "mean_token_accuracy": 0.16469699293375015,
      "num_tokens": 32643927.0,
      "step": 14895
    },
    {
      "entropy": 5.517231369018555,
      "epoch": 1.47876141325923,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00047878478200104845,
      "loss": 5.2933,
      "mean_token_accuracy": 0.15774290710687638,
      "num_tokens": 32655086.0,
      "step": 14900
    },
    {
      "entropy": 5.558468151092529,
      "epoch": 1.4792576419213974,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004787697601497876,
      "loss": 5.2493,
      "mean_token_accuracy": 0.15982231944799424,
      "num_tokens": 32665298.0,
      "step": 14905
    },
    {
      "entropy": 5.566588926315307,
      "epoch": 1.479753870583565,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004787547332454758,
      "loss": 5.3154,
      "mean_token_accuracy": 0.14966288954019547,
      "num_tokens": 32676717.0,
      "step": 14910
    },
    {
      "entropy": 5.622492074966431,
      "epoch": 1.4802500992457324,
      "grad_norm": 1.71875,
      "learning_rate": 0.00047873970128848567,
      "loss": 5.3704,
      "mean_token_accuracy": 0.14881788790225983,
      "num_tokens": 32686886.0,
      "step": 14915
    },
    {
      "entropy": 5.542652177810669,
      "epoch": 1.4807463279078998,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004787246642791899,
      "loss": 5.3148,
      "mean_token_accuracy": 0.15729339718818663,
      "num_tokens": 32698631.0,
      "step": 14920
    },
    {
      "entropy": 5.571841716766357,
      "epoch": 1.4812425565700675,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047870962221796147,
      "loss": 5.3273,
      "mean_token_accuracy": 0.15055980086326598,
      "num_tokens": 32710672.0,
      "step": 14925
    },
    {
      "entropy": 5.552709007263184,
      "epoch": 1.481738785232235,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004786945751051734,
      "loss": 5.3578,
      "mean_token_accuracy": 0.15674103647470475,
      "num_tokens": 32720858.0,
      "step": 14930
    },
    {
      "entropy": 5.517229795455933,
      "epoch": 1.4822350138944025,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004786795229411988,
      "loss": 5.2971,
      "mean_token_accuracy": 0.15582564920186998,
      "num_tokens": 32731592.0,
      "step": 14935
    },
    {
      "entropy": 5.624639844894409,
      "epoch": 1.4827312425565702,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047866446572641084,
      "loss": 5.3621,
      "mean_token_accuracy": 0.14827580004930496,
      "num_tokens": 32742777.0,
      "step": 14940
    },
    {
      "entropy": 5.608187675476074,
      "epoch": 1.4832274712187377,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004786494034611831,
      "loss": 5.2866,
      "mean_token_accuracy": 0.15557848066091537,
      "num_tokens": 32753035.0,
      "step": 14945
    },
    {
      "entropy": 5.54914927482605,
      "epoch": 1.4837236998809051,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000478634336145889,
      "loss": 5.3151,
      "mean_token_accuracy": 0.15387140214443207,
      "num_tokens": 32764318.0,
      "step": 14950
    },
    {
      "entropy": 5.582449340820313,
      "epoch": 1.4842199285430726,
      "grad_norm": 1.796875,
      "learning_rate": 0.00047861926378090197,
      "loss": 5.3319,
      "mean_token_accuracy": 0.14819187447428703,
      "num_tokens": 32775521.0,
      "step": 14955
    },
    {
      "entropy": 5.627120065689087,
      "epoch": 1.48471615720524,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004786041863665961,
      "loss": 5.3555,
      "mean_token_accuracy": 0.15588053166866303,
      "num_tokens": 32787101.0,
      "step": 14960
    },
    {
      "entropy": 5.58338041305542,
      "epoch": 1.4852123858674078,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004785891039033451,
      "loss": 5.35,
      "mean_token_accuracy": 0.15342047512531282,
      "num_tokens": 32798446.0,
      "step": 14965
    },
    {
      "entropy": 5.576963758468628,
      "epoch": 1.4857086145295753,
      "grad_norm": 1.421875,
      "learning_rate": 0.000478574016391523,
      "loss": 5.3662,
      "mean_token_accuracy": 0.1455719754099846,
      "num_tokens": 32809636.0,
      "step": 14970
    },
    {
      "entropy": 5.520108938217163,
      "epoch": 1.4862048431917427,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000478558923831504,
      "loss": 5.2511,
      "mean_token_accuracy": 0.15494602173566818,
      "num_tokens": 32820132.0,
      "step": 14975
    },
    {
      "entropy": 5.623526906967163,
      "epoch": 1.4867010718539102,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047854382622366235,
      "loss": 5.3542,
      "mean_token_accuracy": 0.15908877849578856,
      "num_tokens": 32830720.0,
      "step": 14980
    },
    {
      "entropy": 5.5590897560119625,
      "epoch": 1.487197300516078,
      "grad_norm": 1.828125,
      "learning_rate": 0.00047852872356837236,
      "loss": 5.3023,
      "mean_token_accuracy": 0.15576327592134476,
      "num_tokens": 32841621.0,
      "step": 14985
    },
    {
      "entropy": 5.62934365272522,
      "epoch": 1.4876935291782454,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004785136158660085,
      "loss": 5.444,
      "mean_token_accuracy": 0.1460772655904293,
      "num_tokens": 32852074.0,
      "step": 14990
    },
    {
      "entropy": 5.6316118240356445,
      "epoch": 1.4881897578404129,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047849850311694567,
      "loss": 5.3818,
      "mean_token_accuracy": 0.15325266420841216,
      "num_tokens": 32862556.0,
      "step": 14995
    },
    {
      "entropy": 5.5881103515625,
      "epoch": 1.4886859865025803,
      "grad_norm": 1.5,
      "learning_rate": 0.00047848338532155833,
      "loss": 5.3167,
      "mean_token_accuracy": 0.15975043624639512,
      "num_tokens": 32874407.0,
      "step": 15000
    },
    {
      "epoch": 1.4886859865025803,
      "eval_entropy": 5.454485900273485,
      "eval_loss": 5.4590840339660645,
      "eval_mean_token_accuracy": 0.15300249088317489,
      "eval_num_tokens": 32874407.0,
      "eval_runtime": 27.2062,
      "eval_samples_per_second": 1237.547,
      "eval_steps_per_second": 154.707,
      "step": 15000
    },
    {
      "entropy": 5.595745515823364,
      "epoch": 1.4891822151647478,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047846826248022155,
      "loss": 5.3495,
      "mean_token_accuracy": 0.14920330345630645,
      "num_tokens": 32884986.0,
      "step": 15005
    },
    {
      "entropy": 5.596615171432495,
      "epoch": 1.4896784438269155,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047845313459331034,
      "loss": 5.3921,
      "mean_token_accuracy": 0.15315815955400466,
      "num_tokens": 32896016.0,
      "step": 15010
    },
    {
      "entropy": 5.576980495452881,
      "epoch": 1.490174672489083,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047843800166119974,
      "loss": 5.2922,
      "mean_token_accuracy": 0.1561645671725273,
      "num_tokens": 32905648.0,
      "step": 15015
    },
    {
      "entropy": 5.6476099491119385,
      "epoch": 1.4906709011512504,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000478422863684265,
      "loss": 5.3595,
      "mean_token_accuracy": 0.15460566729307174,
      "num_tokens": 32917103.0,
      "step": 15020
    },
    {
      "entropy": 5.604568767547607,
      "epoch": 1.4911671298134181,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047840772066288167,
      "loss": 5.2978,
      "mean_token_accuracy": 0.15198874697089196,
      "num_tokens": 32928741.0,
      "step": 15025
    },
    {
      "entropy": 5.528544759750366,
      "epoch": 1.4916633584755856,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004783925725974252,
      "loss": 5.3662,
      "mean_token_accuracy": 0.1501271143555641,
      "num_tokens": 32939292.0,
      "step": 15030
    },
    {
      "entropy": 5.614155054092407,
      "epoch": 1.492159587137753,
      "grad_norm": 4.0,
      "learning_rate": 0.0004783774194882712,
      "loss": 5.3576,
      "mean_token_accuracy": 0.14903282076120378,
      "num_tokens": 32949940.0,
      "step": 15035
    },
    {
      "entropy": 5.596303224563599,
      "epoch": 1.4926558157999206,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004783622613357954,
      "loss": 5.331,
      "mean_token_accuracy": 0.15778712034225464,
      "num_tokens": 32961222.0,
      "step": 15040
    },
    {
      "entropy": 5.520463991165161,
      "epoch": 1.493152044462088,
      "grad_norm": 1.75,
      "learning_rate": 0.0004783470981403737,
      "loss": 5.3005,
      "mean_token_accuracy": 0.16081604808568956,
      "num_tokens": 32972095.0,
      "step": 15045
    },
    {
      "entropy": 5.580161714553833,
      "epoch": 1.4936482731242555,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004783319299023821,
      "loss": 5.3498,
      "mean_token_accuracy": 0.15642765164375305,
      "num_tokens": 32981834.0,
      "step": 15050
    },
    {
      "entropy": 5.6697088241577145,
      "epoch": 1.4941445017864232,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004783167566221969,
      "loss": 5.4458,
      "mean_token_accuracy": 0.1483989492058754,
      "num_tokens": 32994347.0,
      "step": 15055
    },
    {
      "entropy": 5.685481405258178,
      "epoch": 1.4946407304485907,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047830157830019423,
      "loss": 5.3996,
      "mean_token_accuracy": 0.14886631816625595,
      "num_tokens": 33004780.0,
      "step": 15060
    },
    {
      "entropy": 5.657601976394654,
      "epoch": 1.4951369591107582,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004782863949367505,
      "loss": 5.3609,
      "mean_token_accuracy": 0.14804850816726683,
      "num_tokens": 33015697.0,
      "step": 15065
    },
    {
      "entropy": 5.527649927139282,
      "epoch": 1.4956331877729259,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004782712065322421,
      "loss": 5.3298,
      "mean_token_accuracy": 0.14746382534503938,
      "num_tokens": 33026563.0,
      "step": 15070
    },
    {
      "entropy": 5.579450035095215,
      "epoch": 1.4961294164350933,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004782560130870459,
      "loss": 5.3449,
      "mean_token_accuracy": 0.1483881726861,
      "num_tokens": 33037290.0,
      "step": 15075
    },
    {
      "entropy": 5.6657496929168705,
      "epoch": 1.4966256450972608,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047824081460153846,
      "loss": 5.2788,
      "mean_token_accuracy": 0.15208484828472138,
      "num_tokens": 33047199.0,
      "step": 15080
    },
    {
      "entropy": 5.557948970794678,
      "epoch": 1.4971218737594283,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004782256110760969,
      "loss": 5.2972,
      "mean_token_accuracy": 0.15283251255750657,
      "num_tokens": 33057253.0,
      "step": 15085
    },
    {
      "entropy": 5.619180488586426,
      "epoch": 1.4976181024215958,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000478210402511098,
      "loss": 5.4462,
      "mean_token_accuracy": 0.14757110998034478,
      "num_tokens": 33068228.0,
      "step": 15090
    },
    {
      "entropy": 5.660345363616943,
      "epoch": 1.4981143310837635,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004781951889069189,
      "loss": 5.4014,
      "mean_token_accuracy": 0.1518438920378685,
      "num_tokens": 33079285.0,
      "step": 15095
    },
    {
      "entropy": 5.560277032852173,
      "epoch": 1.498610559745931,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047817997026393714,
      "loss": 5.2697,
      "mean_token_accuracy": 0.15824159234762192,
      "num_tokens": 33091253.0,
      "step": 15100
    },
    {
      "entropy": 5.531698274612427,
      "epoch": 1.4991067884080984,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047816474658252985,
      "loss": 5.293,
      "mean_token_accuracy": 0.1514116793870926,
      "num_tokens": 33102492.0,
      "step": 15105
    },
    {
      "entropy": 5.549141597747803,
      "epoch": 1.499603017070266,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047814951786307467,
      "loss": 5.3172,
      "mean_token_accuracy": 0.1559359535574913,
      "num_tokens": 33115038.0,
      "step": 15110
    },
    {
      "entropy": 5.563778352737427,
      "epoch": 1.5000992457324336,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004781342841059492,
      "loss": 5.2769,
      "mean_token_accuracy": 0.15677871704101562,
      "num_tokens": 33124470.0,
      "step": 15115
    },
    {
      "entropy": 5.575761413574218,
      "epoch": 1.500595474394601,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004781190453115311,
      "loss": 5.3304,
      "mean_token_accuracy": 0.1537366032600403,
      "num_tokens": 33135994.0,
      "step": 15120
    },
    {
      "entropy": 5.565511083602905,
      "epoch": 1.5010917030567685,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047810380148019847,
      "loss": 5.3262,
      "mean_token_accuracy": 0.1551915869116783,
      "num_tokens": 33145783.0,
      "step": 15125
    },
    {
      "entropy": 5.565045976638794,
      "epoch": 1.501587931718936,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047808855261232915,
      "loss": 5.4254,
      "mean_token_accuracy": 0.15126277804374694,
      "num_tokens": 33156404.0,
      "step": 15130
    },
    {
      "entropy": 5.576933097839356,
      "epoch": 1.5020841603811035,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047807329870830134,
      "loss": 5.4178,
      "mean_token_accuracy": 0.15279437154531478,
      "num_tokens": 33168226.0,
      "step": 15135
    },
    {
      "entropy": 5.576825714111328,
      "epoch": 1.5025803890432712,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004780580397684933,
      "loss": 5.3189,
      "mean_token_accuracy": 0.1499326631426811,
      "num_tokens": 33178265.0,
      "step": 15140
    },
    {
      "entropy": 5.55530104637146,
      "epoch": 1.5030766177054387,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004780427757932835,
      "loss": 5.2866,
      "mean_token_accuracy": 0.16153979748487474,
      "num_tokens": 33188992.0,
      "step": 15145
    },
    {
      "entropy": 5.543338584899902,
      "epoch": 1.5035728463676064,
      "grad_norm": 3.578125,
      "learning_rate": 0.0004780275067830503,
      "loss": 5.3529,
      "mean_token_accuracy": 0.1545162692666054,
      "num_tokens": 33199424.0,
      "step": 15150
    },
    {
      "entropy": 5.550428295135498,
      "epoch": 1.5040690750297738,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004780122327381725,
      "loss": 5.2877,
      "mean_token_accuracy": 0.15815422385931016,
      "num_tokens": 33210900.0,
      "step": 15155
    },
    {
      "entropy": 5.614867305755615,
      "epoch": 1.5045653036919413,
      "grad_norm": 2.6875,
      "learning_rate": 0.00047799695365902865,
      "loss": 5.3514,
      "mean_token_accuracy": 0.14761825352907182,
      "num_tokens": 33220796.0,
      "step": 15160
    },
    {
      "entropy": 5.580586290359497,
      "epoch": 1.5050615323541088,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004779816695459979,
      "loss": 5.3522,
      "mean_token_accuracy": 0.15039972364902496,
      "num_tokens": 33233454.0,
      "step": 15165
    },
    {
      "entropy": 5.573079776763916,
      "epoch": 1.5055577610162763,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004779663803994591,
      "loss": 5.2628,
      "mean_token_accuracy": 0.15735326409339906,
      "num_tokens": 33243574.0,
      "step": 15170
    },
    {
      "entropy": 5.564756202697754,
      "epoch": 1.5060539896784437,
      "grad_norm": 1.5,
      "learning_rate": 0.0004779510862197914,
      "loss": 5.3097,
      "mean_token_accuracy": 0.15298646986484526,
      "num_tokens": 33256298.0,
      "step": 15175
    },
    {
      "entropy": 5.582154846191406,
      "epoch": 1.5065502183406112,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004779357870073741,
      "loss": 5.4262,
      "mean_token_accuracy": 0.1486109673976898,
      "num_tokens": 33268089.0,
      "step": 15180
    },
    {
      "entropy": 5.569670629501343,
      "epoch": 1.507046447002779,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047792048276258653,
      "loss": 5.3198,
      "mean_token_accuracy": 0.14976810961961745,
      "num_tokens": 33279955.0,
      "step": 15185
    },
    {
      "entropy": 5.588629245758057,
      "epoch": 1.5075426756649464,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004779051734858083,
      "loss": 5.3176,
      "mean_token_accuracy": 0.15183554142713546,
      "num_tokens": 33291470.0,
      "step": 15190
    },
    {
      "entropy": 5.59222731590271,
      "epoch": 1.508038904327114,
      "grad_norm": 1.734375,
      "learning_rate": 0.00047788985917741894,
      "loss": 5.367,
      "mean_token_accuracy": 0.15891223549842834,
      "num_tokens": 33304169.0,
      "step": 15195
    },
    {
      "entropy": 5.479765701293945,
      "epoch": 1.5085351329892815,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047787453983779826,
      "loss": 5.1904,
      "mean_token_accuracy": 0.16050027012825013,
      "num_tokens": 33314436.0,
      "step": 15200
    },
    {
      "entropy": 5.558080387115479,
      "epoch": 1.509031361651449,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004778592154673262,
      "loss": 5.4075,
      "mean_token_accuracy": 0.14829804450273515,
      "num_tokens": 33325361.0,
      "step": 15205
    },
    {
      "entropy": 5.50170636177063,
      "epoch": 1.5095275903136165,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047784388606638263,
      "loss": 5.2707,
      "mean_token_accuracy": 0.15394257605075837,
      "num_tokens": 33336644.0,
      "step": 15210
    },
    {
      "entropy": 5.566920423507691,
      "epoch": 1.510023818975784,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004778285516353478,
      "loss": 5.3343,
      "mean_token_accuracy": 0.15421986430883408,
      "num_tokens": 33348418.0,
      "step": 15215
    },
    {
      "entropy": 5.546612787246704,
      "epoch": 1.5105200476379514,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047781321217460205,
      "loss": 5.2917,
      "mean_token_accuracy": 0.15833595693111419,
      "num_tokens": 33360615.0,
      "step": 15220
    },
    {
      "entropy": 5.545138359069824,
      "epoch": 1.5110162763001191,
      "grad_norm": 1.78125,
      "learning_rate": 0.00047779786768452554,
      "loss": 5.2796,
      "mean_token_accuracy": 0.15434435307979583,
      "num_tokens": 33372030.0,
      "step": 15225
    },
    {
      "entropy": 5.580491542816162,
      "epoch": 1.5115125049622866,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047778251816549895,
      "loss": 5.3225,
      "mean_token_accuracy": 0.1462794691324234,
      "num_tokens": 33382834.0,
      "step": 15230
    },
    {
      "entropy": 5.544415855407715,
      "epoch": 1.512008733624454,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047776716361790284,
      "loss": 5.309,
      "mean_token_accuracy": 0.15758906155824662,
      "num_tokens": 33393250.0,
      "step": 15235
    },
    {
      "entropy": 5.529370546340942,
      "epoch": 1.5125049622866218,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047775180404211804,
      "loss": 5.2687,
      "mean_token_accuracy": 0.1626941204071045,
      "num_tokens": 33402778.0,
      "step": 15240
    },
    {
      "entropy": 5.58170804977417,
      "epoch": 1.5130011909487893,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004777364394385254,
      "loss": 5.3578,
      "mean_token_accuracy": 0.1560498893260956,
      "num_tokens": 33413565.0,
      "step": 15245
    },
    {
      "entropy": 5.605157804489136,
      "epoch": 1.5134974196109567,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047772106980750593,
      "loss": 5.2853,
      "mean_token_accuracy": 0.1573310077190399,
      "num_tokens": 33424108.0,
      "step": 15250
    },
    {
      "entropy": 5.563763093948364,
      "epoch": 1.5139936482731242,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004777056951494407,
      "loss": 5.4154,
      "mean_token_accuracy": 0.1497437611222267,
      "num_tokens": 33435683.0,
      "step": 15255
    },
    {
      "entropy": 5.537083482742309,
      "epoch": 1.5144898769352917,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000477690315464711,
      "loss": 5.3494,
      "mean_token_accuracy": 0.15282171219587326,
      "num_tokens": 33446267.0,
      "step": 15260
    },
    {
      "entropy": 5.532860326766968,
      "epoch": 1.5149861055974592,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00047767493075369827,
      "loss": 5.2573,
      "mean_token_accuracy": 0.16792926639318467,
      "num_tokens": 33457901.0,
      "step": 15265
    },
    {
      "entropy": 5.593736457824707,
      "epoch": 1.5154823342596269,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000477659541016784,
      "loss": 5.4273,
      "mean_token_accuracy": 0.1498861566185951,
      "num_tokens": 33469215.0,
      "step": 15270
    },
    {
      "entropy": 5.588661861419678,
      "epoch": 1.5159785629217943,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047764414625434983,
      "loss": 5.2837,
      "mean_token_accuracy": 0.15507955998182296,
      "num_tokens": 33480384.0,
      "step": 15275
    },
    {
      "entropy": 5.573630619049072,
      "epoch": 1.516474791583962,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004776287464667774,
      "loss": 5.3104,
      "mean_token_accuracy": 0.15882946401834488,
      "num_tokens": 33491404.0,
      "step": 15280
    },
    {
      "entropy": 5.6242364883422855,
      "epoch": 1.5169710202461295,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047761334165444876,
      "loss": 5.4196,
      "mean_token_accuracy": 0.14349345713853837,
      "num_tokens": 33501888.0,
      "step": 15285
    },
    {
      "entropy": 5.64523344039917,
      "epoch": 1.517467248908297,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047759793181774586,
      "loss": 5.3674,
      "mean_token_accuracy": 0.15765542089939116,
      "num_tokens": 33513274.0,
      "step": 15290
    },
    {
      "entropy": 5.583764457702637,
      "epoch": 1.5179634775704645,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047758251695705073,
      "loss": 5.3416,
      "mean_token_accuracy": 0.15094257295131683,
      "num_tokens": 33524188.0,
      "step": 15295
    },
    {
      "entropy": 5.601757383346557,
      "epoch": 1.518459706232632,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047756709707274576,
      "loss": 5.4147,
      "mean_token_accuracy": 0.1519734099507332,
      "num_tokens": 33535207.0,
      "step": 15300
    },
    {
      "entropy": 5.645059394836426,
      "epoch": 1.5189559348947994,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004775516721652133,
      "loss": 5.3601,
      "mean_token_accuracy": 0.14853450804948806,
      "num_tokens": 33546154.0,
      "step": 15305
    },
    {
      "entropy": 5.537896060943604,
      "epoch": 1.519452163556967,
      "grad_norm": 1.375,
      "learning_rate": 0.0004775362422348358,
      "loss": 5.2604,
      "mean_token_accuracy": 0.16563362181186675,
      "num_tokens": 33555365.0,
      "step": 15310
    },
    {
      "entropy": 5.515996503829956,
      "epoch": 1.5199483922191346,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047752080728199586,
      "loss": 5.3294,
      "mean_token_accuracy": 0.15011829733848572,
      "num_tokens": 33566316.0,
      "step": 15315
    },
    {
      "entropy": 5.530770587921142,
      "epoch": 1.520444620881302,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004775053673070764,
      "loss": 5.2649,
      "mean_token_accuracy": 0.16024721711874007,
      "num_tokens": 33576549.0,
      "step": 15320
    },
    {
      "entropy": 5.542410945892334,
      "epoch": 1.5209408495434698,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004774899223104601,
      "loss": 5.3201,
      "mean_token_accuracy": 0.15816013514995575,
      "num_tokens": 33586681.0,
      "step": 15325
    },
    {
      "entropy": 5.6767477035522464,
      "epoch": 1.5214370782056372,
      "grad_norm": 1.625,
      "learning_rate": 0.0004774744722925301,
      "loss": 5.459,
      "mean_token_accuracy": 0.14515277072787286,
      "num_tokens": 33599201.0,
      "step": 15330
    },
    {
      "entropy": 5.608912801742553,
      "epoch": 1.5219333068678047,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004774590172536695,
      "loss": 5.3071,
      "mean_token_accuracy": 0.1541913777589798,
      "num_tokens": 33610217.0,
      "step": 15335
    },
    {
      "entropy": 5.564007902145386,
      "epoch": 1.5224295355299722,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047744355719426146,
      "loss": 5.3448,
      "mean_token_accuracy": 0.15395482778549194,
      "num_tokens": 33620586.0,
      "step": 15340
    },
    {
      "entropy": 5.583789539337158,
      "epoch": 1.5229257641921397,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047742809211468953,
      "loss": 5.3285,
      "mean_token_accuracy": 0.15753490328788758,
      "num_tokens": 33630712.0,
      "step": 15345
    },
    {
      "entropy": 5.517925024032593,
      "epoch": 1.5234219928543071,
      "grad_norm": 1.296875,
      "learning_rate": 0.000477412622015337,
      "loss": 5.2605,
      "mean_token_accuracy": 0.15581250786781312,
      "num_tokens": 33640367.0,
      "step": 15350
    },
    {
      "entropy": 5.61035647392273,
      "epoch": 1.5239182215164748,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004773971468965876,
      "loss": 5.3663,
      "mean_token_accuracy": 0.15176940560340882,
      "num_tokens": 33651816.0,
      "step": 15355
    },
    {
      "entropy": 5.598461437225342,
      "epoch": 1.5244144501786423,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047738166675882516,
      "loss": 5.2519,
      "mean_token_accuracy": 0.15924246460199357,
      "num_tokens": 33663661.0,
      "step": 15360
    },
    {
      "entropy": 5.515711164474487,
      "epoch": 1.52491067884081,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004773661816024335,
      "loss": 5.1484,
      "mean_token_accuracy": 0.16521305590867996,
      "num_tokens": 33673670.0,
      "step": 15365
    },
    {
      "entropy": 5.49716420173645,
      "epoch": 1.5254069075029775,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004773506914277965,
      "loss": 5.3986,
      "mean_token_accuracy": 0.14885758608579636,
      "num_tokens": 33685434.0,
      "step": 15370
    },
    {
      "entropy": 5.554122591018677,
      "epoch": 1.525903136165145,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047733519623529853,
      "loss": 5.3637,
      "mean_token_accuracy": 0.1568499341607094,
      "num_tokens": 33696751.0,
      "step": 15375
    },
    {
      "entropy": 5.601215600967407,
      "epoch": 1.5263993648273124,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004773196960253236,
      "loss": 5.3049,
      "mean_token_accuracy": 0.1556314781308174,
      "num_tokens": 33708141.0,
      "step": 15380
    },
    {
      "entropy": 5.480664443969727,
      "epoch": 1.52689559348948,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004773041907982561,
      "loss": 5.1752,
      "mean_token_accuracy": 0.1639094680547714,
      "num_tokens": 33719219.0,
      "step": 15385
    },
    {
      "entropy": 5.548329162597656,
      "epoch": 1.5273918221516474,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047728868055448064,
      "loss": 5.3097,
      "mean_token_accuracy": 0.15471617132425308,
      "num_tokens": 33730493.0,
      "step": 15390
    },
    {
      "entropy": 5.619878482818604,
      "epoch": 1.5278880508138148,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004772731652943818,
      "loss": 5.3697,
      "mean_token_accuracy": 0.14660182297229768,
      "num_tokens": 33740273.0,
      "step": 15395
    },
    {
      "entropy": 5.618013000488281,
      "epoch": 1.5283842794759825,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004772576450183444,
      "loss": 5.3267,
      "mean_token_accuracy": 0.15821680873632432,
      "num_tokens": 33752162.0,
      "step": 15400
    },
    {
      "entropy": 5.578912925720215,
      "epoch": 1.52888050813815,
      "grad_norm": 1.5,
      "learning_rate": 0.00047724211972675315,
      "loss": 5.294,
      "mean_token_accuracy": 0.15424714982509613,
      "num_tokens": 33763139.0,
      "step": 15405
    },
    {
      "entropy": 5.539679670333863,
      "epoch": 1.5293767368003177,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004772265894199931,
      "loss": 5.3392,
      "mean_token_accuracy": 0.15345700234174728,
      "num_tokens": 33774641.0,
      "step": 15410
    },
    {
      "entropy": 5.625232362747193,
      "epoch": 1.5298729654624852,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004772110540984494,
      "loss": 5.3393,
      "mean_token_accuracy": 0.15585496053099632,
      "num_tokens": 33785644.0,
      "step": 15415
    },
    {
      "entropy": 5.681424379348755,
      "epoch": 1.5303691941246527,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004771955137625073,
      "loss": 5.4563,
      "mean_token_accuracy": 0.14579481184482573,
      "num_tokens": 33797880.0,
      "step": 15420
    },
    {
      "entropy": 5.501276159286499,
      "epoch": 1.5308654227868201,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047717996841255224,
      "loss": 5.2755,
      "mean_token_accuracy": 0.15597875118255616,
      "num_tokens": 33808782.0,
      "step": 15425
    },
    {
      "entropy": 5.481735277175903,
      "epoch": 1.5313616514489876,
      "grad_norm": 2.359375,
      "learning_rate": 0.00047716441804896955,
      "loss": 5.2987,
      "mean_token_accuracy": 0.16102624237537383,
      "num_tokens": 33818483.0,
      "step": 15430
    },
    {
      "entropy": 5.546667146682739,
      "epoch": 1.531857880111155,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004771488626721449,
      "loss": 5.3032,
      "mean_token_accuracy": 0.15072073191404342,
      "num_tokens": 33828590.0,
      "step": 15435
    },
    {
      "entropy": 5.622566366195679,
      "epoch": 1.5323541087733228,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047713330228246406,
      "loss": 5.3957,
      "mean_token_accuracy": 0.1499764144420624,
      "num_tokens": 33838978.0,
      "step": 15440
    },
    {
      "entropy": 5.617811393737793,
      "epoch": 1.5328503374354903,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004771177368803129,
      "loss": 5.3338,
      "mean_token_accuracy": 0.15337470918893814,
      "num_tokens": 33851244.0,
      "step": 15445
    },
    {
      "entropy": 5.571711015701294,
      "epoch": 1.533346566097658,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047710216646607733,
      "loss": 5.2395,
      "mean_token_accuracy": 0.15889672040939332,
      "num_tokens": 33861666.0,
      "step": 15450
    },
    {
      "entropy": 5.530802965164185,
      "epoch": 1.5338427947598254,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047708659104014363,
      "loss": 5.2927,
      "mean_token_accuracy": 0.1546909213066101,
      "num_tokens": 33872416.0,
      "step": 15455
    },
    {
      "entropy": 5.486066627502441,
      "epoch": 1.534339023421993,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004770710106028978,
      "loss": 5.2817,
      "mean_token_accuracy": 0.15687309205532074,
      "num_tokens": 33881808.0,
      "step": 15460
    },
    {
      "entropy": 5.614900302886963,
      "epoch": 1.5348352520841604,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047705542515472647,
      "loss": 5.3786,
      "mean_token_accuracy": 0.15662708431482314,
      "num_tokens": 33892925.0,
      "step": 15465
    },
    {
      "entropy": 5.575819110870361,
      "epoch": 1.5353314807463279,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004770398346960159,
      "loss": 5.3214,
      "mean_token_accuracy": 0.14822439849376678,
      "num_tokens": 33903702.0,
      "step": 15470
    },
    {
      "entropy": 5.651213836669922,
      "epoch": 1.5358277094084953,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047702423922715284,
      "loss": 5.4176,
      "mean_token_accuracy": 0.15219579190015792,
      "num_tokens": 33917233.0,
      "step": 15475
    },
    {
      "entropy": 5.585955238342285,
      "epoch": 1.5363239380706628,
      "grad_norm": 8.3125,
      "learning_rate": 0.0004770086387485239,
      "loss": 5.3691,
      "mean_token_accuracy": 0.15480630993843078,
      "num_tokens": 33928362.0,
      "step": 15480
    },
    {
      "entropy": 5.6097804546356205,
      "epoch": 1.5368201667328305,
      "grad_norm": 1.296875,
      "learning_rate": 0.000476993033260516,
      "loss": 5.3916,
      "mean_token_accuracy": 0.15181665420532225,
      "num_tokens": 33938624.0,
      "step": 15485
    },
    {
      "entropy": 5.5448345184326175,
      "epoch": 1.537316395394998,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004769774227635162,
      "loss": 5.2764,
      "mean_token_accuracy": 0.1552410215139389,
      "num_tokens": 33950058.0,
      "step": 15490
    },
    {
      "entropy": 5.632225465774536,
      "epoch": 1.5378126240571657,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004769618072579115,
      "loss": 5.3818,
      "mean_token_accuracy": 0.14862419068813323,
      "num_tokens": 33962404.0,
      "step": 15495
    },
    {
      "entropy": 5.632731151580811,
      "epoch": 1.5383088527193332,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047694618674408917,
      "loss": 5.3282,
      "mean_token_accuracy": 0.1567629471421242,
      "num_tokens": 33973095.0,
      "step": 15500
    },
    {
      "entropy": 5.558283042907715,
      "epoch": 1.5388050813815006,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047693056122243655,
      "loss": 5.3546,
      "mean_token_accuracy": 0.15071016252040864,
      "num_tokens": 33984152.0,
      "step": 15505
    },
    {
      "entropy": 5.577968883514404,
      "epoch": 1.539301310043668,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047691493069334124,
      "loss": 5.3461,
      "mean_token_accuracy": 0.1511073336005211,
      "num_tokens": 33994854.0,
      "step": 15510
    },
    {
      "entropy": 5.611559200286865,
      "epoch": 1.5397975387058356,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047689929515719066,
      "loss": 5.3648,
      "mean_token_accuracy": 0.15496934354305267,
      "num_tokens": 34006573.0,
      "step": 15515
    },
    {
      "entropy": 5.564942312240601,
      "epoch": 1.540293767368003,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047688365461437263,
      "loss": 5.2727,
      "mean_token_accuracy": 0.15244106501340865,
      "num_tokens": 34016537.0,
      "step": 15520
    },
    {
      "entropy": 5.5523881912231445,
      "epoch": 1.5407899960301707,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047686800906527493,
      "loss": 5.3048,
      "mean_token_accuracy": 0.1652009978890419,
      "num_tokens": 34026825.0,
      "step": 15525
    },
    {
      "entropy": 5.605583953857422,
      "epoch": 1.5412862246923382,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004768523585102856,
      "loss": 5.325,
      "mean_token_accuracy": 0.1575642116367817,
      "num_tokens": 34036786.0,
      "step": 15530
    },
    {
      "entropy": 5.5970673084259035,
      "epoch": 1.541782453354506,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004768367029497928,
      "loss": 5.311,
      "mean_token_accuracy": 0.15605657249689103,
      "num_tokens": 34046784.0,
      "step": 15535
    },
    {
      "entropy": 5.602694988250732,
      "epoch": 1.5422786820166734,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047682104238418464,
      "loss": 5.3703,
      "mean_token_accuracy": 0.1536722093820572,
      "num_tokens": 34058250.0,
      "step": 15540
    },
    {
      "entropy": 5.529318952560425,
      "epoch": 1.5427749106788409,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047680537681384956,
      "loss": 5.3019,
      "mean_token_accuracy": 0.15570985227823259,
      "num_tokens": 34069969.0,
      "step": 15545
    },
    {
      "entropy": 5.596062850952149,
      "epoch": 1.5432711393410083,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004767897062391759,
      "loss": 5.4285,
      "mean_token_accuracy": 0.15010153353214264,
      "num_tokens": 34080817.0,
      "step": 15550
    },
    {
      "entropy": 5.547226905822754,
      "epoch": 1.5437673680031758,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004767740306605524,
      "loss": 5.3255,
      "mean_token_accuracy": 0.1500630721449852,
      "num_tokens": 34091138.0,
      "step": 15555
    },
    {
      "entropy": 5.563704299926758,
      "epoch": 1.5442635966653433,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004767583500783677,
      "loss": 5.2744,
      "mean_token_accuracy": 0.15848077312111855,
      "num_tokens": 34102035.0,
      "step": 15560
    },
    {
      "entropy": 5.574217367172241,
      "epoch": 1.5447598253275108,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047674266449301074,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1555880919098854,
      "num_tokens": 34112533.0,
      "step": 15565
    },
    {
      "entropy": 5.571885681152343,
      "epoch": 1.5452560539896785,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047672697390487034,
      "loss": 5.3278,
      "mean_token_accuracy": 0.16205270886421203,
      "num_tokens": 34124196.0,
      "step": 15570
    },
    {
      "entropy": 5.550203323364258,
      "epoch": 1.545752282651846,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004767112783143357,
      "loss": 5.3092,
      "mean_token_accuracy": 0.15690868943929673,
      "num_tokens": 34135635.0,
      "step": 15575
    },
    {
      "entropy": 5.535918283462524,
      "epoch": 1.5462485113140136,
      "grad_norm": 1.140625,
      "learning_rate": 0.000476695577721796,
      "loss": 5.245,
      "mean_token_accuracy": 0.15706741511821748,
      "num_tokens": 34146202.0,
      "step": 15580
    },
    {
      "entropy": 5.565359258651734,
      "epoch": 1.5467447399761811,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047667987212764065,
      "loss": 5.2941,
      "mean_token_accuracy": 0.1578651025891304,
      "num_tokens": 34157354.0,
      "step": 15585
    },
    {
      "entropy": 5.559172439575195,
      "epoch": 1.5472409686383486,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047666416153225894,
      "loss": 5.3416,
      "mean_token_accuracy": 0.14800440818071364,
      "num_tokens": 34168563.0,
      "step": 15590
    },
    {
      "entropy": 5.5476062297821045,
      "epoch": 1.547737197300516,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004766484459360406,
      "loss": 5.279,
      "mean_token_accuracy": 0.1565502643585205,
      "num_tokens": 34178604.0,
      "step": 15595
    },
    {
      "entropy": 5.574475193023682,
      "epoch": 1.5482334259626835,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047663272533937537,
      "loss": 5.3271,
      "mean_token_accuracy": 0.15411384254693986,
      "num_tokens": 34190321.0,
      "step": 15600
    },
    {
      "entropy": 5.528959894180298,
      "epoch": 1.548729654624851,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047661699974265294,
      "loss": 5.3285,
      "mean_token_accuracy": 0.15531688928604126,
      "num_tokens": 34201519.0,
      "step": 15605
    },
    {
      "entropy": 5.514243125915527,
      "epoch": 1.5492258832870187,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047660126914626344,
      "loss": 5.2786,
      "mean_token_accuracy": 0.15314695835113526,
      "num_tokens": 34212710.0,
      "step": 15610
    },
    {
      "entropy": 5.621606016159058,
      "epoch": 1.5497221119491862,
      "grad_norm": 1.5,
      "learning_rate": 0.0004765855335505969,
      "loss": 5.4111,
      "mean_token_accuracy": 0.15227942913770676,
      "num_tokens": 34223187.0,
      "step": 15615
    },
    {
      "entropy": 5.642950391769409,
      "epoch": 1.5502183406113537,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004765697929560434,
      "loss": 5.3526,
      "mean_token_accuracy": 0.1572491243481636,
      "num_tokens": 34234530.0,
      "step": 15620
    },
    {
      "entropy": 5.592506456375122,
      "epoch": 1.5507145692735214,
      "grad_norm": 1.5,
      "learning_rate": 0.00047655404736299354,
      "loss": 5.3496,
      "mean_token_accuracy": 0.15292095094919206,
      "num_tokens": 34245003.0,
      "step": 15625
    },
    {
      "entropy": 5.511945199966431,
      "epoch": 1.5512107979356888,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047653829677183745,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1600438952445984,
      "num_tokens": 34255630.0,
      "step": 15630
    },
    {
      "entropy": 5.564335823059082,
      "epoch": 1.5517070265978563,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047652254118296596,
      "loss": 5.32,
      "mean_token_accuracy": 0.16010782420635222,
      "num_tokens": 34266716.0,
      "step": 15635
    },
    {
      "entropy": 5.613068675994873,
      "epoch": 1.5522032552600238,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004765067805967697,
      "loss": 5.3406,
      "mean_token_accuracy": 0.15324259251356126,
      "num_tokens": 34278905.0,
      "step": 15640
    },
    {
      "entropy": 5.574022102355957,
      "epoch": 1.5526994839221913,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004764910150136395,
      "loss": 5.2902,
      "mean_token_accuracy": 0.15406236797571182,
      "num_tokens": 34289688.0,
      "step": 15645
    },
    {
      "entropy": 5.563397455215454,
      "epoch": 1.5531957125843587,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047647524443396625,
      "loss": 5.3232,
      "mean_token_accuracy": 0.15572215393185615,
      "num_tokens": 34300964.0,
      "step": 15650
    },
    {
      "entropy": 5.550938653945923,
      "epoch": 1.5536919412465264,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047645946885814105,
      "loss": 5.2996,
      "mean_token_accuracy": 0.1561855584383011,
      "num_tokens": 34311435.0,
      "step": 15655
    },
    {
      "entropy": 5.54950065612793,
      "epoch": 1.554188169908694,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004764436882865552,
      "loss": 5.3268,
      "mean_token_accuracy": 0.1505746752023697,
      "num_tokens": 34321742.0,
      "step": 15660
    },
    {
      "entropy": 5.618258857727051,
      "epoch": 1.5546843985708616,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004764279027195999,
      "loss": 5.3791,
      "mean_token_accuracy": 0.1524025857448578,
      "num_tokens": 34333796.0,
      "step": 15665
    },
    {
      "entropy": 5.7423210620880125,
      "epoch": 1.555180627233029,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047641211215766673,
      "loss": 5.4426,
      "mean_token_accuracy": 0.15378427356481553,
      "num_tokens": 34344521.0,
      "step": 15670
    },
    {
      "entropy": 5.547049951553345,
      "epoch": 1.5556768558951966,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004763963166011471,
      "loss": 5.3257,
      "mean_token_accuracy": 0.15916882306337357,
      "num_tokens": 34355629.0,
      "step": 15675
    },
    {
      "entropy": 5.536240720748902,
      "epoch": 1.556173084557364,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004763805160504328,
      "loss": 5.376,
      "mean_token_accuracy": 0.15737542361021042,
      "num_tokens": 34366935.0,
      "step": 15680
    },
    {
      "entropy": 5.543711185455322,
      "epoch": 1.5566693132195315,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004763647105059157,
      "loss": 5.2929,
      "mean_token_accuracy": 0.14850193858146668,
      "num_tokens": 34378355.0,
      "step": 15685
    },
    {
      "entropy": 5.565633392333984,
      "epoch": 1.557165541881699,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004763488999679877,
      "loss": 5.2397,
      "mean_token_accuracy": 0.15687926709651948,
      "num_tokens": 34388811.0,
      "step": 15690
    },
    {
      "entropy": 5.659170389175415,
      "epoch": 1.5576617705438665,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047633308443704076,
      "loss": 5.3575,
      "mean_token_accuracy": 0.1500616379082203,
      "num_tokens": 34399905.0,
      "step": 15695
    },
    {
      "entropy": 5.52790265083313,
      "epoch": 1.5581579992060341,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047631726391346717,
      "loss": 5.238,
      "mean_token_accuracy": 0.1550780475139618,
      "num_tokens": 34410338.0,
      "step": 15700
    },
    {
      "entropy": 5.570883750915527,
      "epoch": 1.5586542278682016,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004763014383976593,
      "loss": 5.4099,
      "mean_token_accuracy": 0.14863350614905357,
      "num_tokens": 34421204.0,
      "step": 15705
    },
    {
      "entropy": 5.5274858474731445,
      "epoch": 1.5591504565303693,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047628560789000945,
      "loss": 5.2247,
      "mean_token_accuracy": 0.16527197659015655,
      "num_tokens": 34431002.0,
      "step": 15710
    },
    {
      "entropy": 5.547233963012696,
      "epoch": 1.5596466851925368,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047626977239091027,
      "loss": 5.2691,
      "mean_token_accuracy": 0.15535202026367187,
      "num_tokens": 34442446.0,
      "step": 15715
    },
    {
      "entropy": 5.5566675662994385,
      "epoch": 1.5601429138547043,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004762539319007545,
      "loss": 5.2393,
      "mean_token_accuracy": 0.15874890089035035,
      "num_tokens": 34453590.0,
      "step": 15720
    },
    {
      "entropy": 5.614942312240601,
      "epoch": 1.5606391425168717,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004762380864199348,
      "loss": 5.3788,
      "mean_token_accuracy": 0.1544570729136467,
      "num_tokens": 34465614.0,
      "step": 15725
    },
    {
      "entropy": 5.626387023925782,
      "epoch": 1.5611353711790392,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004762222359488443,
      "loss": 5.4637,
      "mean_token_accuracy": 0.14764380753040313,
      "num_tokens": 34476616.0,
      "step": 15730
    },
    {
      "entropy": 5.5920453548431395,
      "epoch": 1.5616315998412067,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004762063804878758,
      "loss": 5.2594,
      "mean_token_accuracy": 0.15029204785823821,
      "num_tokens": 34486260.0,
      "step": 15735
    },
    {
      "entropy": 5.5983561992645265,
      "epoch": 1.5621278285033744,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047619052003742277,
      "loss": 5.3336,
      "mean_token_accuracy": 0.15426755249500274,
      "num_tokens": 34496392.0,
      "step": 15740
    },
    {
      "entropy": 5.574337196350098,
      "epoch": 1.5626240571655419,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004761746545978782,
      "loss": 5.3263,
      "mean_token_accuracy": 0.1618194982409477,
      "num_tokens": 34507218.0,
      "step": 15745
    },
    {
      "entropy": 5.621217060089111,
      "epoch": 1.5631202858277096,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004761587841696358,
      "loss": 5.398,
      "mean_token_accuracy": 0.14592298045754432,
      "num_tokens": 34518888.0,
      "step": 15750
    },
    {
      "entropy": 5.565575075149536,
      "epoch": 1.563616514489877,
      "grad_norm": 1.546875,
      "learning_rate": 0.000476142908753089,
      "loss": 5.3046,
      "mean_token_accuracy": 0.15362192913889886,
      "num_tokens": 34529865.0,
      "step": 15755
    },
    {
      "entropy": 5.587220525741577,
      "epoch": 1.5641127431520445,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004761270283486314,
      "loss": 5.3971,
      "mean_token_accuracy": 0.148468516767025,
      "num_tokens": 34542541.0,
      "step": 15760
    },
    {
      "entropy": 5.636463403701782,
      "epoch": 1.564608971814212,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047611114295665697,
      "loss": 5.3772,
      "mean_token_accuracy": 0.15202566832304001,
      "num_tokens": 34553056.0,
      "step": 15765
    },
    {
      "entropy": 5.6403051853179935,
      "epoch": 1.5651052004763795,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047609525257755967,
      "loss": 5.3544,
      "mean_token_accuracy": 0.15417518764734267,
      "num_tokens": 34563344.0,
      "step": 15770
    },
    {
      "entropy": 5.606562280654908,
      "epoch": 1.565601429138547,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004760793572117333,
      "loss": 5.2715,
      "mean_token_accuracy": 0.16136376112699508,
      "num_tokens": 34574320.0,
      "step": 15775
    },
    {
      "entropy": 5.535801839828491,
      "epoch": 1.5660976578007144,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004760634568595721,
      "loss": 5.3247,
      "mean_token_accuracy": 0.15110252499580384,
      "num_tokens": 34584728.0,
      "step": 15780
    },
    {
      "entropy": 5.62635440826416,
      "epoch": 1.5665938864628821,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004760475515214704,
      "loss": 5.4257,
      "mean_token_accuracy": 0.15070997625589372,
      "num_tokens": 34596395.0,
      "step": 15785
    },
    {
      "entropy": 5.558712482452393,
      "epoch": 1.5670901151250496,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004760316411978228,
      "loss": 5.3283,
      "mean_token_accuracy": 0.1510426864027977,
      "num_tokens": 34607950.0,
      "step": 15790
    },
    {
      "entropy": 5.558389472961426,
      "epoch": 1.5675863437872173,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047601572588902363,
      "loss": 5.3272,
      "mean_token_accuracy": 0.1585109606385231,
      "num_tokens": 34617640.0,
      "step": 15795
    },
    {
      "entropy": 5.555823802947998,
      "epoch": 1.5680825724493848,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047599980559546757,
      "loss": 5.3761,
      "mean_token_accuracy": 0.14542382061481476,
      "num_tokens": 34629475.0,
      "step": 15800
    },
    {
      "entropy": 5.568705034255982,
      "epoch": 1.5685788011115522,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004759838803175495,
      "loss": 5.3045,
      "mean_token_accuracy": 0.1501409277319908,
      "num_tokens": 34641423.0,
      "step": 15805
    },
    {
      "entropy": 5.661091423034668,
      "epoch": 1.5690750297737197,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047596795005566423,
      "loss": 5.3682,
      "mean_token_accuracy": 0.14950846806168555,
      "num_tokens": 34651894.0,
      "step": 15810
    },
    {
      "entropy": 5.628204584121704,
      "epoch": 1.5695712584358872,
      "grad_norm": 1.5,
      "learning_rate": 0.00047595201481020685,
      "loss": 5.3774,
      "mean_token_accuracy": 0.1514196902513504,
      "num_tokens": 34664177.0,
      "step": 15815
    },
    {
      "entropy": 5.571656322479248,
      "epoch": 1.5700674870980547,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047593607458157257,
      "loss": 5.332,
      "mean_token_accuracy": 0.15623770356178285,
      "num_tokens": 34674850.0,
      "step": 15820
    },
    {
      "entropy": 5.576591873168946,
      "epoch": 1.5705637157602224,
      "grad_norm": 1.78125,
      "learning_rate": 0.00047592012937015647,
      "loss": 5.3526,
      "mean_token_accuracy": 0.1536719247698784,
      "num_tokens": 34685719.0,
      "step": 15825
    },
    {
      "entropy": 5.617337656021118,
      "epoch": 1.5710599444223898,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047590417917635423,
      "loss": 5.3466,
      "mean_token_accuracy": 0.14578449502587318,
      "num_tokens": 34697261.0,
      "step": 15830
    },
    {
      "entropy": 5.589413070678711,
      "epoch": 1.5715561730845575,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004758882240005613,
      "loss": 5.3443,
      "mean_token_accuracy": 0.15478500500321388,
      "num_tokens": 34708196.0,
      "step": 15835
    },
    {
      "entropy": 5.593596363067627,
      "epoch": 1.572052401746725,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047587226384317323,
      "loss": 5.3844,
      "mean_token_accuracy": 0.1499848559498787,
      "num_tokens": 34718855.0,
      "step": 15840
    },
    {
      "entropy": 5.59573450088501,
      "epoch": 1.5725486304088925,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047585629870458584,
      "loss": 5.3026,
      "mean_token_accuracy": 0.15933698862791063,
      "num_tokens": 34731137.0,
      "step": 15845
    },
    {
      "entropy": 5.609903287887573,
      "epoch": 1.57304485907106,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004758403285851951,
      "loss": 5.2502,
      "mean_token_accuracy": 0.16093026399612426,
      "num_tokens": 34741628.0,
      "step": 15850
    },
    {
      "entropy": 5.5170488357543945,
      "epoch": 1.5735410877332274,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047582435348539695,
      "loss": 5.3411,
      "mean_token_accuracy": 0.1555830255150795,
      "num_tokens": 34752984.0,
      "step": 15855
    },
    {
      "entropy": 5.60999321937561,
      "epoch": 1.574037316395395,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004758083734055875,
      "loss": 5.3989,
      "mean_token_accuracy": 0.14826516062021255,
      "num_tokens": 34764363.0,
      "step": 15860
    },
    {
      "entropy": 5.632789850234985,
      "epoch": 1.5745335450575624,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004757923883461632,
      "loss": 5.3531,
      "mean_token_accuracy": 0.1497912272810936,
      "num_tokens": 34776289.0,
      "step": 15865
    },
    {
      "entropy": 5.605644607543946,
      "epoch": 1.57502977371973,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047577639830752034,
      "loss": 5.3368,
      "mean_token_accuracy": 0.1541541412472725,
      "num_tokens": 34786760.0,
      "step": 15870
    },
    {
      "entropy": 5.564853048324585,
      "epoch": 1.5755260023818976,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047576040329005536,
      "loss": 5.3081,
      "mean_token_accuracy": 0.15300208926200867,
      "num_tokens": 34797737.0,
      "step": 15875
    },
    {
      "entropy": 5.584435892105103,
      "epoch": 1.5760222310440652,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000475744403294165,
      "loss": 5.3352,
      "mean_token_accuracy": 0.15830534398555757,
      "num_tokens": 34809650.0,
      "step": 15880
    },
    {
      "entropy": 5.675848817825317,
      "epoch": 1.5765184597062327,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047572839832024606,
      "loss": 5.3794,
      "mean_token_accuracy": 0.14566903412342072,
      "num_tokens": 34820171.0,
      "step": 15885
    },
    {
      "entropy": 5.525732946395874,
      "epoch": 1.5770146883684002,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047571238836869537,
      "loss": 5.228,
      "mean_token_accuracy": 0.15823534578084947,
      "num_tokens": 34831286.0,
      "step": 15890
    },
    {
      "entropy": 5.469884061813355,
      "epoch": 1.5775109170305677,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004756963734399099,
      "loss": 5.2529,
      "mean_token_accuracy": 0.16067838966846465,
      "num_tokens": 34841767.0,
      "step": 15895
    },
    {
      "entropy": 5.579144239425659,
      "epoch": 1.5780071456927351,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047568035353428683,
      "loss": 5.3637,
      "mean_token_accuracy": 0.15196479558944703,
      "num_tokens": 34852644.0,
      "step": 15900
    },
    {
      "entropy": 5.622080278396607,
      "epoch": 1.5785033743549026,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047566432865222343,
      "loss": 5.3545,
      "mean_token_accuracy": 0.15344627648591996,
      "num_tokens": 34863605.0,
      "step": 15905
    },
    {
      "entropy": 5.486923742294311,
      "epoch": 1.5789996030170703,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047564829879411706,
      "loss": 5.3059,
      "mean_token_accuracy": 0.15795368403196336,
      "num_tokens": 34874594.0,
      "step": 15910
    },
    {
      "entropy": 5.579173564910889,
      "epoch": 1.5794958316792378,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004756322639603653,
      "loss": 5.3718,
      "mean_token_accuracy": 0.1493668183684349,
      "num_tokens": 34884627.0,
      "step": 15915
    },
    {
      "entropy": 5.668580627441406,
      "epoch": 1.5799920603414053,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047561622415136565,
      "loss": 5.3628,
      "mean_token_accuracy": 0.1543991059064865,
      "num_tokens": 34895679.0,
      "step": 15920
    },
    {
      "entropy": 5.519829034805298,
      "epoch": 1.580488289003573,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047560017936751597,
      "loss": 5.2924,
      "mean_token_accuracy": 0.15486378222703934,
      "num_tokens": 34906811.0,
      "step": 15925
    },
    {
      "entropy": 5.575310087203979,
      "epoch": 1.5809845176657404,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004755841296092141,
      "loss": 5.3395,
      "mean_token_accuracy": 0.15247920900583267,
      "num_tokens": 34918655.0,
      "step": 15930
    },
    {
      "entropy": 5.574757194519043,
      "epoch": 1.581480746327908,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047556807487685797,
      "loss": 5.3209,
      "mean_token_accuracy": 0.15134753733873368,
      "num_tokens": 34931034.0,
      "step": 15935
    },
    {
      "entropy": 5.579299068450927,
      "epoch": 1.5819769749900754,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004755520151708458,
      "loss": 5.2627,
      "mean_token_accuracy": 0.1562902271747589,
      "num_tokens": 34939707.0,
      "step": 15940
    },
    {
      "entropy": 5.49659743309021,
      "epoch": 1.5824732036522429,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004755359504915758,
      "loss": 5.3875,
      "mean_token_accuracy": 0.1469066321849823,
      "num_tokens": 34949237.0,
      "step": 15945
    },
    {
      "entropy": 5.605391359329223,
      "epoch": 1.5829694323144103,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004755198808394464,
      "loss": 5.4116,
      "mean_token_accuracy": 0.15757358819246292,
      "num_tokens": 34960380.0,
      "step": 15950
    },
    {
      "entropy": 5.66767258644104,
      "epoch": 1.583465660976578,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000475503806214856,
      "loss": 5.3717,
      "mean_token_accuracy": 0.1500846579670906,
      "num_tokens": 34971699.0,
      "step": 15955
    },
    {
      "entropy": 5.538005018234253,
      "epoch": 1.5839618896387455,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004754877266182033,
      "loss": 5.3185,
      "mean_token_accuracy": 0.15566515624523164,
      "num_tokens": 34982333.0,
      "step": 15960
    },
    {
      "entropy": 5.5492359638214115,
      "epoch": 1.5844581183009132,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000475471642049887,
      "loss": 5.357,
      "mean_token_accuracy": 0.1545894905924797,
      "num_tokens": 34993277.0,
      "step": 15965
    },
    {
      "entropy": 5.596325063705445,
      "epoch": 1.5849543469630807,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004754555525103058,
      "loss": 5.2899,
      "mean_token_accuracy": 0.15902315825223923,
      "num_tokens": 35004252.0,
      "step": 15970
    },
    {
      "entropy": 5.502462339401245,
      "epoch": 1.5854505756252482,
      "grad_norm": 1.5625,
      "learning_rate": 0.000475439457999859,
      "loss": 5.2934,
      "mean_token_accuracy": 0.1592414140701294,
      "num_tokens": 35016974.0,
      "step": 15975
    },
    {
      "entropy": 5.565835809707641,
      "epoch": 1.5859468042874156,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047542335851894553,
      "loss": 5.3876,
      "mean_token_accuracy": 0.15097211450338363,
      "num_tokens": 35029017.0,
      "step": 15980
    },
    {
      "entropy": 5.607384824752808,
      "epoch": 1.586443032949583,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047540725406796454,
      "loss": 5.4037,
      "mean_token_accuracy": 0.1489529624581337,
      "num_tokens": 35040239.0,
      "step": 15985
    },
    {
      "entropy": 5.629933166503906,
      "epoch": 1.5869392616117506,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004753911446473156,
      "loss": 5.3125,
      "mean_token_accuracy": 0.15295821130275727,
      "num_tokens": 35051426.0,
      "step": 15990
    },
    {
      "entropy": 5.487988233566284,
      "epoch": 1.587435490273918,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000475375030257398,
      "loss": 5.276,
      "mean_token_accuracy": 0.14909113943576813,
      "num_tokens": 35061777.0,
      "step": 15995
    },
    {
      "entropy": 5.549928474426269,
      "epoch": 1.5879317189360858,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004753589108986114,
      "loss": 5.2545,
      "mean_token_accuracy": 0.16583003401756286,
      "num_tokens": 35071822.0,
      "step": 16000
    },
    {
      "entropy": 5.573054790496826,
      "epoch": 1.5884279475982532,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047534278657135565,
      "loss": 5.325,
      "mean_token_accuracy": 0.1538211241364479,
      "num_tokens": 35081900.0,
      "step": 16005
    },
    {
      "entropy": 5.6045839309692385,
      "epoch": 1.588924176260421,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004753266572760304,
      "loss": 5.3769,
      "mean_token_accuracy": 0.1555659294128418,
      "num_tokens": 35092866.0,
      "step": 16010
    },
    {
      "entropy": 5.739051437377929,
      "epoch": 1.5894204049225884,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047531052301303577,
      "loss": 5.5089,
      "mean_token_accuracy": 0.15480010062456132,
      "num_tokens": 35103425.0,
      "step": 16015
    },
    {
      "entropy": 5.567584609985351,
      "epoch": 1.5899166335847559,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004752943837827718,
      "loss": 5.2544,
      "mean_token_accuracy": 0.15543368011713027,
      "num_tokens": 35114620.0,
      "step": 16020
    },
    {
      "entropy": 5.57315583229065,
      "epoch": 1.5904128622469234,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004752782395856387,
      "loss": 5.3655,
      "mean_token_accuracy": 0.14971593990921975,
      "num_tokens": 35125199.0,
      "step": 16025
    },
    {
      "entropy": 5.507650327682495,
      "epoch": 1.5909090909090908,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004752620904220368,
      "loss": 5.3378,
      "mean_token_accuracy": 0.15468747913837433,
      "num_tokens": 35135892.0,
      "step": 16030
    },
    {
      "entropy": 5.5833274841308596,
      "epoch": 1.5914053195712583,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004752459362923666,
      "loss": 5.3276,
      "mean_token_accuracy": 0.15307615101337432,
      "num_tokens": 35146186.0,
      "step": 16035
    },
    {
      "entropy": 5.588626432418823,
      "epoch": 1.591901548233426,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004752297771970287,
      "loss": 5.3775,
      "mean_token_accuracy": 0.15178096815943717,
      "num_tokens": 35157261.0,
      "step": 16040
    },
    {
      "entropy": 5.585022687911987,
      "epoch": 1.5923977768955935,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047521361313642373,
      "loss": 5.3147,
      "mean_token_accuracy": 0.15599383562803268,
      "num_tokens": 35168508.0,
      "step": 16045
    },
    {
      "entropy": 5.5035120964050295,
      "epoch": 1.5928940055577612,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004751974441109526,
      "loss": 5.2777,
      "mean_token_accuracy": 0.15483694672584533,
      "num_tokens": 35179444.0,
      "step": 16050
    },
    {
      "entropy": 5.534209251403809,
      "epoch": 1.5933902342199286,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047518127012101624,
      "loss": 5.2932,
      "mean_token_accuracy": 0.16178925186395646,
      "num_tokens": 35189517.0,
      "step": 16055
    },
    {
      "entropy": 5.56444296836853,
      "epoch": 1.5938864628820961,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004751650911670157,
      "loss": 5.3225,
      "mean_token_accuracy": 0.15087749809026718,
      "num_tokens": 35200191.0,
      "step": 16060
    },
    {
      "entropy": 5.609839963912964,
      "epoch": 1.5943826915442636,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047514890724935236,
      "loss": 5.342,
      "mean_token_accuracy": 0.14980731755495072,
      "num_tokens": 35211260.0,
      "step": 16065
    },
    {
      "entropy": 5.484301567077637,
      "epoch": 1.594878920206431,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004751327183684272,
      "loss": 5.3261,
      "mean_token_accuracy": 0.15323131382465363,
      "num_tokens": 35222408.0,
      "step": 16070
    },
    {
      "entropy": 5.604696369171142,
      "epoch": 1.5953751488685985,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047511652452464196,
      "loss": 5.3734,
      "mean_token_accuracy": 0.15365138426423072,
      "num_tokens": 35232760.0,
      "step": 16075
    },
    {
      "entropy": 5.632780408859253,
      "epoch": 1.595871377530766,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047510032571839816,
      "loss": 5.3338,
      "mean_token_accuracy": 0.15205200165510177,
      "num_tokens": 35243849.0,
      "step": 16080
    },
    {
      "entropy": 5.6109169006347654,
      "epoch": 1.5963676061929337,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004750841219500975,
      "loss": 5.3421,
      "mean_token_accuracy": 0.15368056893348694,
      "num_tokens": 35254962.0,
      "step": 16085
    },
    {
      "entropy": 5.567150211334228,
      "epoch": 1.5968638348551012,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004750679132201416,
      "loss": 5.4012,
      "mean_token_accuracy": 0.15037989169359206,
      "num_tokens": 35265895.0,
      "step": 16090
    },
    {
      "entropy": 5.660325956344605,
      "epoch": 1.597360063517269,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047505169952893276,
      "loss": 5.4151,
      "mean_token_accuracy": 0.1509908765554428,
      "num_tokens": 35277543.0,
      "step": 16095
    },
    {
      "entropy": 5.594936418533325,
      "epoch": 1.5978562921794364,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047503548087687267,
      "loss": 5.3243,
      "mean_token_accuracy": 0.15463293790817262,
      "num_tokens": 35289130.0,
      "step": 16100
    },
    {
      "entropy": 5.503684949874878,
      "epoch": 1.5983525208416038,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047501925726436376,
      "loss": 5.2629,
      "mean_token_accuracy": 0.15317713767290114,
      "num_tokens": 35298414.0,
      "step": 16105
    },
    {
      "entropy": 5.557555341720581,
      "epoch": 1.5988487495037713,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004750030286918082,
      "loss": 5.3457,
      "mean_token_accuracy": 0.1518327161669731,
      "num_tokens": 35310629.0,
      "step": 16110
    },
    {
      "entropy": 5.666381120681763,
      "epoch": 1.5993449781659388,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004749867951596086,
      "loss": 5.3677,
      "mean_token_accuracy": 0.1522211402654648,
      "num_tokens": 35321810.0,
      "step": 16115
    },
    {
      "entropy": 5.585275650024414,
      "epoch": 1.5998412068281063,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047497055666816736,
      "loss": 5.2703,
      "mean_token_accuracy": 0.15672306716442108,
      "num_tokens": 35333517.0,
      "step": 16120
    },
    {
      "entropy": 5.501574516296387,
      "epoch": 1.600337435490274,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004749543132178873,
      "loss": 5.2434,
      "mean_token_accuracy": 0.1599430963397026,
      "num_tokens": 35344170.0,
      "step": 16125
    },
    {
      "entropy": 5.57308611869812,
      "epoch": 1.6008336641524414,
      "grad_norm": 1.28125,
      "learning_rate": 0.000474938064809171,
      "loss": 5.4265,
      "mean_token_accuracy": 0.1503927528858185,
      "num_tokens": 35354963.0,
      "step": 16130
    },
    {
      "entropy": 5.5967014789581295,
      "epoch": 1.6013298928146091,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004749218114424216,
      "loss": 5.3762,
      "mean_token_accuracy": 0.15689027309417725,
      "num_tokens": 35365716.0,
      "step": 16135
    },
    {
      "entropy": 5.582081604003906,
      "epoch": 1.6018261214767766,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047490555311804206,
      "loss": 5.2923,
      "mean_token_accuracy": 0.15784256160259247,
      "num_tokens": 35376113.0,
      "step": 16140
    },
    {
      "entropy": 5.504783821105957,
      "epoch": 1.602322350138944,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047488928983643565,
      "loss": 5.2588,
      "mean_token_accuracy": 0.15081997513771056,
      "num_tokens": 35386178.0,
      "step": 16145
    },
    {
      "entropy": 5.487810754776001,
      "epoch": 1.6028185788011116,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047487302159800545,
      "loss": 5.2008,
      "mean_token_accuracy": 0.16245897859334946,
      "num_tokens": 35396874.0,
      "step": 16150
    },
    {
      "entropy": 5.54116792678833,
      "epoch": 1.603314807463279,
      "grad_norm": 1.453125,
      "learning_rate": 0.000474856748403155,
      "loss": 5.403,
      "mean_token_accuracy": 0.1522229328751564,
      "num_tokens": 35409405.0,
      "step": 16155
    },
    {
      "entropy": 5.621026420593262,
      "epoch": 1.6038110361254465,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004748404702522879,
      "loss": 5.331,
      "mean_token_accuracy": 0.15412441939115523,
      "num_tokens": 35421449.0,
      "step": 16160
    },
    {
      "entropy": 5.576800680160522,
      "epoch": 1.604307264787614,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047482418714580774,
      "loss": 5.3944,
      "mean_token_accuracy": 0.1493387848138809,
      "num_tokens": 35434612.0,
      "step": 16165
    },
    {
      "entropy": 5.554715585708618,
      "epoch": 1.6048034934497817,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004748078990841183,
      "loss": 5.2971,
      "mean_token_accuracy": 0.16132021993398665,
      "num_tokens": 35446016.0,
      "step": 16170
    },
    {
      "entropy": 5.64047532081604,
      "epoch": 1.6052997221119492,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004747916060676235,
      "loss": 5.378,
      "mean_token_accuracy": 0.15214113444089888,
      "num_tokens": 35457405.0,
      "step": 16175
    },
    {
      "entropy": 5.552808046340942,
      "epoch": 1.6057959507741169,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047477530809672747,
      "loss": 5.3867,
      "mean_token_accuracy": 0.1507679894566536,
      "num_tokens": 35468361.0,
      "step": 16180
    },
    {
      "entropy": 5.556333732604981,
      "epoch": 1.6062921794362843,
      "grad_norm": 1.25,
      "learning_rate": 0.0004747590051718342,
      "loss": 5.2776,
      "mean_token_accuracy": 0.15407151728868484,
      "num_tokens": 35480092.0,
      "step": 16185
    },
    {
      "entropy": 5.588628816604614,
      "epoch": 1.6067884080984518,
      "grad_norm": 1.25,
      "learning_rate": 0.00047474269729334806,
      "loss": 5.41,
      "mean_token_accuracy": 0.1463086262345314,
      "num_tokens": 35492131.0,
      "step": 16190
    },
    {
      "entropy": 5.634671878814697,
      "epoch": 1.6072846367606193,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047472638446167346,
      "loss": 5.3534,
      "mean_token_accuracy": 0.15103007853031158,
      "num_tokens": 35503600.0,
      "step": 16195
    },
    {
      "entropy": 5.654093647003174,
      "epoch": 1.6077808654227868,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047471006667721484,
      "loss": 5.4038,
      "mean_token_accuracy": 0.14761386588215827,
      "num_tokens": 35514716.0,
      "step": 16200
    },
    {
      "entropy": 5.541616153717041,
      "epoch": 1.6082770940849542,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000474693743940377,
      "loss": 5.3349,
      "mean_token_accuracy": 0.15194665938615798,
      "num_tokens": 35525688.0,
      "step": 16205
    },
    {
      "entropy": 5.571300506591797,
      "epoch": 1.608773322747122,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047467741625156453,
      "loss": 5.3268,
      "mean_token_accuracy": 0.16145787686109542,
      "num_tokens": 35536391.0,
      "step": 16210
    },
    {
      "entropy": 5.5164793014526365,
      "epoch": 1.6092695514092894,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047466108361118235,
      "loss": 5.2614,
      "mean_token_accuracy": 0.15969271659851075,
      "num_tokens": 35547016.0,
      "step": 16215
    },
    {
      "entropy": 5.540873241424561,
      "epoch": 1.6097657800714569,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047464474601963564,
      "loss": 5.332,
      "mean_token_accuracy": 0.15183869898319244,
      "num_tokens": 35558402.0,
      "step": 16220
    },
    {
      "entropy": 5.5944420337677006,
      "epoch": 1.6102620087336246,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004746284034773294,
      "loss": 5.3395,
      "mean_token_accuracy": 0.1556980550289154,
      "num_tokens": 35569779.0,
      "step": 16225
    },
    {
      "entropy": 5.636451625823975,
      "epoch": 1.610758237395792,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004746120559846689,
      "loss": 5.345,
      "mean_token_accuracy": 0.15460469275712968,
      "num_tokens": 35580044.0,
      "step": 16230
    },
    {
      "entropy": 5.639441013336182,
      "epoch": 1.6112544660579595,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047459570354205954,
      "loss": 5.4261,
      "mean_token_accuracy": 0.15355666726827621,
      "num_tokens": 35591216.0,
      "step": 16235
    },
    {
      "entropy": 5.515120649337769,
      "epoch": 1.611750694720127,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047457934614990684,
      "loss": 5.2443,
      "mean_token_accuracy": 0.1585569977760315,
      "num_tokens": 35602376.0,
      "step": 16240
    },
    {
      "entropy": 5.511484003067016,
      "epoch": 1.6122469233822945,
      "grad_norm": 1.25,
      "learning_rate": 0.0004745629838086164,
      "loss": 5.222,
      "mean_token_accuracy": 0.1580004259943962,
      "num_tokens": 35612184.0,
      "step": 16245
    },
    {
      "entropy": 5.578326272964477,
      "epoch": 1.612743152044462,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047454661651859394,
      "loss": 5.4002,
      "mean_token_accuracy": 0.15251383781433106,
      "num_tokens": 35624864.0,
      "step": 16250
    },
    {
      "entropy": 5.6110405921936035,
      "epoch": 1.6132393807066296,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004745302442802454,
      "loss": 5.333,
      "mean_token_accuracy": 0.15230897814035416,
      "num_tokens": 35636492.0,
      "step": 16255
    },
    {
      "entropy": 5.532247161865234,
      "epoch": 1.6137356093687971,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004745138670939768,
      "loss": 5.2664,
      "mean_token_accuracy": 0.1596395805478096,
      "num_tokens": 35646831.0,
      "step": 16260
    },
    {
      "entropy": 5.586107921600342,
      "epoch": 1.6142318380309648,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004744974849601941,
      "loss": 5.3658,
      "mean_token_accuracy": 0.15350607931613922,
      "num_tokens": 35657703.0,
      "step": 16265
    },
    {
      "entropy": 5.46837420463562,
      "epoch": 1.6147280666931323,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004744810978793038,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1651777908205986,
      "num_tokens": 35668334.0,
      "step": 16270
    },
    {
      "entropy": 5.509501171112061,
      "epoch": 1.6152242953552998,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047446470585171206,
      "loss": 5.3634,
      "mean_token_accuracy": 0.15173261761665344,
      "num_tokens": 35679796.0,
      "step": 16275
    },
    {
      "entropy": 5.539596509933472,
      "epoch": 1.6157205240174672,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047444830887782537,
      "loss": 5.241,
      "mean_token_accuracy": 0.155074842274189,
      "num_tokens": 35690825.0,
      "step": 16280
    },
    {
      "entropy": 5.643378973007202,
      "epoch": 1.6162167526796347,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004744319069580504,
      "loss": 5.3622,
      "mean_token_accuracy": 0.15133204981684684,
      "num_tokens": 35701824.0,
      "step": 16285
    },
    {
      "entropy": 5.610250616073609,
      "epoch": 1.6167129813418022,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047441550009279395,
      "loss": 5.2969,
      "mean_token_accuracy": 0.15842511579394342,
      "num_tokens": 35713250.0,
      "step": 16290
    },
    {
      "entropy": 5.562658739089966,
      "epoch": 1.6172092100039697,
      "grad_norm": 1.625,
      "learning_rate": 0.0004743990882824627,
      "loss": 5.3523,
      "mean_token_accuracy": 0.15480480194091797,
      "num_tokens": 35725049.0,
      "step": 16295
    },
    {
      "entropy": 5.517474222183227,
      "epoch": 1.6177054386661374,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047438267152746376,
      "loss": 5.2889,
      "mean_token_accuracy": 0.15361465960741044,
      "num_tokens": 35734676.0,
      "step": 16300
    },
    {
      "entropy": 5.614069890975952,
      "epoch": 1.6182016673283048,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047436624982820423,
      "loss": 5.3023,
      "mean_token_accuracy": 0.15760921090841293,
      "num_tokens": 35743953.0,
      "step": 16305
    },
    {
      "entropy": 5.512655973434448,
      "epoch": 1.6186978959904725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004743498231850913,
      "loss": 5.2221,
      "mean_token_accuracy": 0.16297280341386794,
      "num_tokens": 35753149.0,
      "step": 16310
    },
    {
      "entropy": 5.510140466690063,
      "epoch": 1.61919412465264,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004743333915985323,
      "loss": 5.3065,
      "mean_token_accuracy": 0.15266507267951965,
      "num_tokens": 35763783.0,
      "step": 16315
    },
    {
      "entropy": 5.519553899765015,
      "epoch": 1.6196903533148075,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047431695506893476,
      "loss": 5.2583,
      "mean_token_accuracy": 0.16117821633815765,
      "num_tokens": 35774663.0,
      "step": 16320
    },
    {
      "entropy": 5.552203512191772,
      "epoch": 1.620186581976975,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047430051359670616,
      "loss": 5.4042,
      "mean_token_accuracy": 0.1476173907518387,
      "num_tokens": 35786990.0,
      "step": 16325
    },
    {
      "entropy": 5.669633102416992,
      "epoch": 1.6206828106391424,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004742840671822542,
      "loss": 5.4267,
      "mean_token_accuracy": 0.15380464643239974,
      "num_tokens": 35798818.0,
      "step": 16330
    },
    {
      "entropy": 5.631000661849976,
      "epoch": 1.62117903930131,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000474267615825987,
      "loss": 5.3477,
      "mean_token_accuracy": 0.15315200686454772,
      "num_tokens": 35808813.0,
      "step": 16335
    },
    {
      "entropy": 5.574074792861938,
      "epoch": 1.6216752679634776,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047425115952831213,
      "loss": 5.3178,
      "mean_token_accuracy": 0.15582658499479293,
      "num_tokens": 35818664.0,
      "step": 16340
    },
    {
      "entropy": 5.490705966949463,
      "epoch": 1.622171496625645,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000474234698289638,
      "loss": 5.2506,
      "mean_token_accuracy": 0.16124146282672883,
      "num_tokens": 35828975.0,
      "step": 16345
    },
    {
      "entropy": 5.548087167739868,
      "epoch": 1.6226677252878128,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004742182321103725,
      "loss": 5.2832,
      "mean_token_accuracy": 0.15552945882081987,
      "num_tokens": 35839437.0,
      "step": 16350
    },
    {
      "entropy": 5.57461838722229,
      "epoch": 1.6231639539499803,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004742017609909242,
      "loss": 5.2511,
      "mean_token_accuracy": 0.16157738864421844,
      "num_tokens": 35850413.0,
      "step": 16355
    },
    {
      "entropy": 5.572533798217774,
      "epoch": 1.6236601826121477,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047418528493170157,
      "loss": 5.3417,
      "mean_token_accuracy": 0.15261703133583068,
      "num_tokens": 35862745.0,
      "step": 16360
    },
    {
      "entropy": 5.633335304260254,
      "epoch": 1.6241564112743152,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004741688039331129,
      "loss": 5.304,
      "mean_token_accuracy": 0.15144399404525757,
      "num_tokens": 35873934.0,
      "step": 16365
    },
    {
      "entropy": 5.522374439239502,
      "epoch": 1.6246526399364827,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047415231799556725,
      "loss": 5.2436,
      "mean_token_accuracy": 0.15993023961782454,
      "num_tokens": 35884279.0,
      "step": 16370
    },
    {
      "entropy": 5.49709644317627,
      "epoch": 1.6251488685986502,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047413582711947305,
      "loss": 5.2907,
      "mean_token_accuracy": 0.16134544909000398,
      "num_tokens": 35894144.0,
      "step": 16375
    },
    {
      "entropy": 5.527485513687134,
      "epoch": 1.6256450972608176,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047411933130523954,
      "loss": 5.2851,
      "mean_token_accuracy": 0.1510717526078224,
      "num_tokens": 35903641.0,
      "step": 16380
    },
    {
      "entropy": 5.511127901077271,
      "epoch": 1.6261413259229853,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004741028305532757,
      "loss": 5.2523,
      "mean_token_accuracy": 0.1546030528843403,
      "num_tokens": 35913634.0,
      "step": 16385
    },
    {
      "entropy": 5.548816537857055,
      "epoch": 1.6266375545851528,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004740863248639907,
      "loss": 5.3109,
      "mean_token_accuracy": 0.1562251552939415,
      "num_tokens": 35923106.0,
      "step": 16390
    },
    {
      "entropy": 5.596654605865479,
      "epoch": 1.6271337832473205,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047406981423779376,
      "loss": 5.2906,
      "mean_token_accuracy": 0.1561199188232422,
      "num_tokens": 35933709.0,
      "step": 16395
    },
    {
      "entropy": 5.575109767913818,
      "epoch": 1.627630011909488,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047405329867509447,
      "loss": 5.318,
      "mean_token_accuracy": 0.14654288738965987,
      "num_tokens": 35944734.0,
      "step": 16400
    },
    {
      "entropy": 5.550746250152588,
      "epoch": 1.6281262405716554,
      "grad_norm": 1.640625,
      "learning_rate": 0.00047403677817630226,
      "loss": 5.2729,
      "mean_token_accuracy": 0.1564690351486206,
      "num_tokens": 35956513.0,
      "step": 16405
    },
    {
      "entropy": 5.519120740890503,
      "epoch": 1.628622469233823,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004740202527418269,
      "loss": 5.304,
      "mean_token_accuracy": 0.15763405859470367,
      "num_tokens": 35967096.0,
      "step": 16410
    },
    {
      "entropy": 5.585912275314331,
      "epoch": 1.6291186978959904,
      "grad_norm": 1.484375,
      "learning_rate": 0.000474003722372078,
      "loss": 5.3626,
      "mean_token_accuracy": 0.15696593672037124,
      "num_tokens": 35979010.0,
      "step": 16415
    },
    {
      "entropy": 5.527207279205323,
      "epoch": 1.6296149265581579,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047398718706746556,
      "loss": 5.2408,
      "mean_token_accuracy": 0.15969198048114777,
      "num_tokens": 35989253.0,
      "step": 16420
    },
    {
      "entropy": 5.491158533096313,
      "epoch": 1.6301111552203256,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004739706468283998,
      "loss": 5.2326,
      "mean_token_accuracy": 0.16212774366140364,
      "num_tokens": 35999659.0,
      "step": 16425
    },
    {
      "entropy": 5.514856624603271,
      "epoch": 1.630607383882493,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047395410165529075,
      "loss": 5.3029,
      "mean_token_accuracy": 0.15482604503631592,
      "num_tokens": 36010295.0,
      "step": 16430
    },
    {
      "entropy": 5.571183919906616,
      "epoch": 1.6311036125446607,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047393755154854855,
      "loss": 5.3355,
      "mean_token_accuracy": 0.15046179443597793,
      "num_tokens": 36021575.0,
      "step": 16435
    },
    {
      "entropy": 5.583881998062134,
      "epoch": 1.6315998412068282,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004739209965085839,
      "loss": 5.3143,
      "mean_token_accuracy": 0.15252304673194886,
      "num_tokens": 36032764.0,
      "step": 16440
    },
    {
      "entropy": 5.568349027633667,
      "epoch": 1.6320960698689957,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000473904436535807,
      "loss": 5.2648,
      "mean_token_accuracy": 0.1637350156903267,
      "num_tokens": 36041864.0,
      "step": 16445
    },
    {
      "entropy": 5.487264251708984,
      "epoch": 1.6325922985311632,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004738878716306288,
      "loss": 5.245,
      "mean_token_accuracy": 0.1538069635629654,
      "num_tokens": 36051321.0,
      "step": 16450
    },
    {
      "entropy": 5.551449489593506,
      "epoch": 1.6330885271933306,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047387130179345996,
      "loss": 5.2446,
      "mean_token_accuracy": 0.15481723397970198,
      "num_tokens": 36061448.0,
      "step": 16455
    },
    {
      "entropy": 5.557281923294068,
      "epoch": 1.6335847558554981,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047385472702471133,
      "loss": 5.3723,
      "mean_token_accuracy": 0.15749264508485794,
      "num_tokens": 36071832.0,
      "step": 16460
    },
    {
      "entropy": 5.517821025848389,
      "epoch": 1.6340809845176656,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000473838147324794,
      "loss": 5.2804,
      "mean_token_accuracy": 0.1515105202794075,
      "num_tokens": 36082103.0,
      "step": 16465
    },
    {
      "entropy": 5.618150043487549,
      "epoch": 1.6345772131798333,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000473821562694119,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1554023876786232,
      "num_tokens": 36092063.0,
      "step": 16470
    },
    {
      "entropy": 5.583439064025879,
      "epoch": 1.6350734418420008,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004738049731330977,
      "loss": 5.314,
      "mean_token_accuracy": 0.1487298861145973,
      "num_tokens": 36102573.0,
      "step": 16475
    },
    {
      "entropy": 5.585085248947143,
      "epoch": 1.6355696705041685,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004737883786421414,
      "loss": 5.2827,
      "mean_token_accuracy": 0.15841763317584992,
      "num_tokens": 36112677.0,
      "step": 16480
    },
    {
      "entropy": 5.60122652053833,
      "epoch": 1.636065899166336,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047377177922166175,
      "loss": 5.353,
      "mean_token_accuracy": 0.15011394917964935,
      "num_tokens": 36123337.0,
      "step": 16485
    },
    {
      "entropy": 5.582880449295044,
      "epoch": 1.6365621278285034,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047375517487207024,
      "loss": 5.3443,
      "mean_token_accuracy": 0.15269685089588164,
      "num_tokens": 36134503.0,
      "step": 16490
    },
    {
      "entropy": 5.618263006210327,
      "epoch": 1.6370583564906709,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047373856559377876,
      "loss": 5.3752,
      "mean_token_accuracy": 0.1515375256538391,
      "num_tokens": 36145766.0,
      "step": 16495
    },
    {
      "entropy": 5.557055997848511,
      "epoch": 1.6375545851528384,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004737219513871989,
      "loss": 5.3369,
      "mean_token_accuracy": 0.15740207508206366,
      "num_tokens": 36156584.0,
      "step": 16500
    },
    {
      "entropy": 5.5577099323272705,
      "epoch": 1.6380508138150058,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047370533225274305,
      "loss": 5.3253,
      "mean_token_accuracy": 0.15923510044813155,
      "num_tokens": 36166999.0,
      "step": 16505
    },
    {
      "entropy": 5.590743112564087,
      "epoch": 1.6385470424771735,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004736887081908231,
      "loss": 5.3076,
      "mean_token_accuracy": 0.1567511409521103,
      "num_tokens": 36178724.0,
      "step": 16510
    },
    {
      "entropy": 5.647546005249024,
      "epoch": 1.639043271139341,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004736720792018513,
      "loss": 5.376,
      "mean_token_accuracy": 0.1528911255300045,
      "num_tokens": 36188912.0,
      "step": 16515
    },
    {
      "entropy": 5.528866100311279,
      "epoch": 1.6395394998015085,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004736554452862399,
      "loss": 5.2814,
      "mean_token_accuracy": 0.15536320507526397,
      "num_tokens": 36198525.0,
      "step": 16520
    },
    {
      "entropy": 5.50860562324524,
      "epoch": 1.6400357284636762,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047363880644440173,
      "loss": 5.2538,
      "mean_token_accuracy": 0.15529693365097047,
      "num_tokens": 36209384.0,
      "step": 16525
    },
    {
      "entropy": 5.6089410305023195,
      "epoch": 1.6405319571258437,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047362216267674906,
      "loss": 5.3487,
      "mean_token_accuracy": 0.14861616045236586,
      "num_tokens": 36220597.0,
      "step": 16530
    },
    {
      "entropy": 5.625215768814087,
      "epoch": 1.6410281857880111,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004736055139836948,
      "loss": 5.3814,
      "mean_token_accuracy": 0.15127637833356858,
      "num_tokens": 36230317.0,
      "step": 16535
    },
    {
      "entropy": 5.671610164642334,
      "epoch": 1.6415244144501786,
      "grad_norm": 1.828125,
      "learning_rate": 0.00047358886036565176,
      "loss": 5.4643,
      "mean_token_accuracy": 0.1518624559044838,
      "num_tokens": 36241645.0,
      "step": 16540
    },
    {
      "entropy": 5.621293020248413,
      "epoch": 1.642020643112346,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047357220182303285,
      "loss": 5.3937,
      "mean_token_accuracy": 0.1500053361058235,
      "num_tokens": 36253098.0,
      "step": 16545
    },
    {
      "entropy": 5.544845008850098,
      "epoch": 1.6425168717745136,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004735555383562513,
      "loss": 5.2945,
      "mean_token_accuracy": 0.15628247410058976,
      "num_tokens": 36264417.0,
      "step": 16550
    },
    {
      "entropy": 5.631740283966065,
      "epoch": 1.6430131004366813,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047353886996572014,
      "loss": 5.3845,
      "mean_token_accuracy": 0.15287363082170485,
      "num_tokens": 36276589.0,
      "step": 16555
    },
    {
      "entropy": 5.553200197219849,
      "epoch": 1.6435093290988487,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047352219665185293,
      "loss": 5.3078,
      "mean_token_accuracy": 0.14986671805381774,
      "num_tokens": 36287308.0,
      "step": 16560
    },
    {
      "entropy": 5.552547407150269,
      "epoch": 1.6440055577610164,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004735055184150629,
      "loss": 5.3105,
      "mean_token_accuracy": 0.15294841974973677,
      "num_tokens": 36298966.0,
      "step": 16565
    },
    {
      "entropy": 5.546921014785767,
      "epoch": 1.644501786423184,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047348883525576385,
      "loss": 5.3263,
      "mean_token_accuracy": 0.15294305086135865,
      "num_tokens": 36310498.0,
      "step": 16570
    },
    {
      "entropy": 5.599332571029663,
      "epoch": 1.6449980150853514,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004734721471743694,
      "loss": 5.3803,
      "mean_token_accuracy": 0.1561164602637291,
      "num_tokens": 36321589.0,
      "step": 16575
    },
    {
      "entropy": 5.572018718719482,
      "epoch": 1.6454942437475188,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004734554541712933,
      "loss": 5.3046,
      "mean_token_accuracy": 0.15115955621004104,
      "num_tokens": 36333041.0,
      "step": 16580
    },
    {
      "entropy": 5.608764696121216,
      "epoch": 1.6459904724096863,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004734387562469496,
      "loss": 5.36,
      "mean_token_accuracy": 0.14831943511962892,
      "num_tokens": 36344225.0,
      "step": 16585
    },
    {
      "entropy": 5.562346315383911,
      "epoch": 1.6464867010718538,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047342205340175235,
      "loss": 5.2456,
      "mean_token_accuracy": 0.1608218640089035,
      "num_tokens": 36355524.0,
      "step": 16590
    },
    {
      "entropy": 5.579918384552002,
      "epoch": 1.6469829297340213,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047340534563611574,
      "loss": 5.3248,
      "mean_token_accuracy": 0.1567131146788597,
      "num_tokens": 36366495.0,
      "step": 16595
    },
    {
      "entropy": 5.54401969909668,
      "epoch": 1.647479158396189,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047338863295045414,
      "loss": 5.3243,
      "mean_token_accuracy": 0.15729763209819794,
      "num_tokens": 36378316.0,
      "step": 16600
    },
    {
      "entropy": 5.586145210266113,
      "epoch": 1.6479753870583564,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047337191534518186,
      "loss": 5.4084,
      "mean_token_accuracy": 0.1492590382695198,
      "num_tokens": 36390500.0,
      "step": 16605
    },
    {
      "entropy": 5.587933778762817,
      "epoch": 1.6484716157205241,
      "grad_norm": 1.625,
      "learning_rate": 0.0004733551928207135,
      "loss": 5.3444,
      "mean_token_accuracy": 0.15185929089784622,
      "num_tokens": 36403831.0,
      "step": 16610
    },
    {
      "entropy": 5.575387239456177,
      "epoch": 1.6489678443826916,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004733384653774639,
      "loss": 5.2865,
      "mean_token_accuracy": 0.15543532073497773,
      "num_tokens": 36414896.0,
      "step": 16615
    },
    {
      "entropy": 5.5934216499328615,
      "epoch": 1.649464073044859,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047332173301584764,
      "loss": 5.2651,
      "mean_token_accuracy": 0.15499356687068938,
      "num_tokens": 36426260.0,
      "step": 16620
    },
    {
      "entropy": 5.512473678588867,
      "epoch": 1.6499603017070266,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047330499573627975,
      "loss": 5.2152,
      "mean_token_accuracy": 0.16574016958475113,
      "num_tokens": 36437125.0,
      "step": 16625
    },
    {
      "entropy": 5.500352001190185,
      "epoch": 1.650456530369194,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047328825353917526,
      "loss": 5.2613,
      "mean_token_accuracy": 0.1578000321984291,
      "num_tokens": 36447507.0,
      "step": 16630
    },
    {
      "entropy": 5.590629434585571,
      "epoch": 1.6509527590313615,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047327150642494944,
      "loss": 5.4283,
      "mean_token_accuracy": 0.15336142629384994,
      "num_tokens": 36459134.0,
      "step": 16635
    },
    {
      "entropy": 5.62870602607727,
      "epoch": 1.6514489876935292,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004732547543940174,
      "loss": 5.3436,
      "mean_token_accuracy": 0.15182484388351442,
      "num_tokens": 36469594.0,
      "step": 16640
    },
    {
      "entropy": 5.591393232345581,
      "epoch": 1.6519452163556967,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004732379974467947,
      "loss": 5.3325,
      "mean_token_accuracy": 0.15920318961143493,
      "num_tokens": 36480477.0,
      "step": 16645
    },
    {
      "entropy": 5.524727296829224,
      "epoch": 1.6524414450178644,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004732212355836968,
      "loss": 5.3148,
      "mean_token_accuracy": 0.16134185194969178,
      "num_tokens": 36491304.0,
      "step": 16650
    },
    {
      "entropy": 5.599442005157471,
      "epoch": 1.6529376736800319,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004732044688051395,
      "loss": 5.4051,
      "mean_token_accuracy": 0.14848437458276748,
      "num_tokens": 36502030.0,
      "step": 16655
    },
    {
      "entropy": 5.579805183410644,
      "epoch": 1.6534339023421993,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047318769711153833,
      "loss": 5.4297,
      "mean_token_accuracy": 0.1476582705974579,
      "num_tokens": 36514405.0,
      "step": 16660
    },
    {
      "entropy": 5.5849082469940186,
      "epoch": 1.6539301310043668,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004731709205033094,
      "loss": 5.2632,
      "mean_token_accuracy": 0.15359878540039062,
      "num_tokens": 36524626.0,
      "step": 16665
    },
    {
      "entropy": 5.577029514312744,
      "epoch": 1.6544263596665343,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047315413898086863,
      "loss": 5.2769,
      "mean_token_accuracy": 0.16225792616605758,
      "num_tokens": 36535340.0,
      "step": 16670
    },
    {
      "entropy": 5.589989233016968,
      "epoch": 1.6549225883287018,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004731373525446322,
      "loss": 5.2976,
      "mean_token_accuracy": 0.1548387736082077,
      "num_tokens": 36547189.0,
      "step": 16675
    },
    {
      "entropy": 5.560368394851684,
      "epoch": 1.6554188169908692,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004731205611950164,
      "loss": 5.3861,
      "mean_token_accuracy": 0.1514149934053421,
      "num_tokens": 36558177.0,
      "step": 16680
    },
    {
      "entropy": 5.5660406112670895,
      "epoch": 1.655915045653037,
      "grad_norm": 1.25,
      "learning_rate": 0.00047310376493243754,
      "loss": 5.199,
      "mean_token_accuracy": 0.16243930011987687,
      "num_tokens": 36569559.0,
      "step": 16685
    },
    {
      "entropy": 5.544949865341186,
      "epoch": 1.6564112743152044,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004730869637573122,
      "loss": 5.2647,
      "mean_token_accuracy": 0.15798574686050415,
      "num_tokens": 36579617.0,
      "step": 16690
    },
    {
      "entropy": 5.526273918151856,
      "epoch": 1.656907502977372,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047307015767005705,
      "loss": 5.2616,
      "mean_token_accuracy": 0.16097115427255632,
      "num_tokens": 36590427.0,
      "step": 16695
    },
    {
      "entropy": 5.5613655090332035,
      "epoch": 1.6574037316395396,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004730533466710888,
      "loss": 5.2625,
      "mean_token_accuracy": 0.15914074331521988,
      "num_tokens": 36602313.0,
      "step": 16700
    },
    {
      "entropy": 5.665493726730347,
      "epoch": 1.657899960301707,
      "grad_norm": 2.375,
      "learning_rate": 0.0004730365307608242,
      "loss": 5.3705,
      "mean_token_accuracy": 0.1539012610912323,
      "num_tokens": 36615191.0,
      "step": 16705
    },
    {
      "entropy": 5.540058565139771,
      "epoch": 1.6583961889638745,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047301970993968054,
      "loss": 5.2633,
      "mean_token_accuracy": 0.166679647564888,
      "num_tokens": 36624692.0,
      "step": 16710
    },
    {
      "entropy": 5.5788267135620115,
      "epoch": 1.658892417626042,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047300288420807463,
      "loss": 5.3385,
      "mean_token_accuracy": 0.15207265317440033,
      "num_tokens": 36635961.0,
      "step": 16715
    },
    {
      "entropy": 5.505636739730835,
      "epoch": 1.6593886462882095,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047298605356642387,
      "loss": 5.3092,
      "mean_token_accuracy": 0.15819020420312882,
      "num_tokens": 36646710.0,
      "step": 16720
    },
    {
      "entropy": 5.571049642562866,
      "epoch": 1.6598848749503772,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047296921801514565,
      "loss": 5.3986,
      "mean_token_accuracy": 0.15133154690265654,
      "num_tokens": 36657168.0,
      "step": 16725
    },
    {
      "entropy": 5.591034317016602,
      "epoch": 1.6603811036125447,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004729523775546574,
      "loss": 5.3563,
      "mean_token_accuracy": 0.1535052999854088,
      "num_tokens": 36668428.0,
      "step": 16730
    },
    {
      "entropy": 5.564067745208741,
      "epoch": 1.6608773322747123,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004729355321853767,
      "loss": 5.3411,
      "mean_token_accuracy": 0.1582286536693573,
      "num_tokens": 36679361.0,
      "step": 16735
    },
    {
      "entropy": 5.666923952102661,
      "epoch": 1.6613735609368798,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004729186819077214,
      "loss": 5.398,
      "mean_token_accuracy": 0.1550811618566513,
      "num_tokens": 36689480.0,
      "step": 16740
    },
    {
      "entropy": 5.680801439285278,
      "epoch": 1.6618697895990473,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004729018267221092,
      "loss": 5.438,
      "mean_token_accuracy": 0.15301423370838166,
      "num_tokens": 36700420.0,
      "step": 16745
    },
    {
      "entropy": 5.607660722732544,
      "epoch": 1.6623660182612148,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004728849666289581,
      "loss": 5.3797,
      "mean_token_accuracy": 0.15092367678880692,
      "num_tokens": 36710106.0,
      "step": 16750
    },
    {
      "entropy": 5.594700336456299,
      "epoch": 1.6628622469233822,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047286810162868625,
      "loss": 5.2297,
      "mean_token_accuracy": 0.16161254197359085,
      "num_tokens": 36719630.0,
      "step": 16755
    },
    {
      "entropy": 5.459393692016602,
      "epoch": 1.6633584755855497,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004728512317217119,
      "loss": 5.1973,
      "mean_token_accuracy": 0.1663515254855156,
      "num_tokens": 36729379.0,
      "step": 16760
    },
    {
      "entropy": 5.493234872817993,
      "epoch": 1.6638547042477172,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047283435690845336,
      "loss": 5.2755,
      "mean_token_accuracy": 0.16107940673828125,
      "num_tokens": 36740103.0,
      "step": 16765
    },
    {
      "entropy": 5.62037501335144,
      "epoch": 1.664350932909885,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047281747718932906,
      "loss": 5.3297,
      "mean_token_accuracy": 0.15738620311021806,
      "num_tokens": 36750928.0,
      "step": 16770
    },
    {
      "entropy": 5.584412956237793,
      "epoch": 1.6648471615720524,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004728005925647575,
      "loss": 5.3159,
      "mean_token_accuracy": 0.15442438200116157,
      "num_tokens": 36761415.0,
      "step": 16775
    },
    {
      "entropy": 5.523766422271729,
      "epoch": 1.66534339023422,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047278370303515744,
      "loss": 5.3351,
      "mean_token_accuracy": 0.15327319353818894,
      "num_tokens": 36772560.0,
      "step": 16780
    },
    {
      "entropy": 5.584780597686768,
      "epoch": 1.6658396188963875,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004727668086009478,
      "loss": 5.3596,
      "mean_token_accuracy": 0.15499525964260102,
      "num_tokens": 36783597.0,
      "step": 16785
    },
    {
      "entropy": 5.630609321594238,
      "epoch": 1.666335847558555,
      "grad_norm": 1.125,
      "learning_rate": 0.00047274990926254753,
      "loss": 5.4373,
      "mean_token_accuracy": 0.14935208708047867,
      "num_tokens": 36794219.0,
      "step": 16790
    },
    {
      "entropy": 5.5143654346466064,
      "epoch": 1.6668320762207225,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004727330050203755,
      "loss": 5.2421,
      "mean_token_accuracy": 0.15918112248182298,
      "num_tokens": 36804156.0,
      "step": 16795
    },
    {
      "entropy": 5.52657413482666,
      "epoch": 1.66732830488289,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004727160958748511,
      "loss": 5.2686,
      "mean_token_accuracy": 0.1578009694814682,
      "num_tokens": 36814414.0,
      "step": 16800
    },
    {
      "entropy": 5.532289695739746,
      "epoch": 1.6678245335450574,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047269918182639355,
      "loss": 5.239,
      "mean_token_accuracy": 0.156059867143631,
      "num_tokens": 36826173.0,
      "step": 16805
    },
    {
      "entropy": 5.604611206054687,
      "epoch": 1.6683207622072251,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004726822628754222,
      "loss": 5.361,
      "mean_token_accuracy": 0.15158332288265228,
      "num_tokens": 36837726.0,
      "step": 16810
    },
    {
      "entropy": 5.5995025634765625,
      "epoch": 1.6688169908693926,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047266533902235683,
      "loss": 5.3123,
      "mean_token_accuracy": 0.1491541087627411,
      "num_tokens": 36848605.0,
      "step": 16815
    },
    {
      "entropy": 5.579540872573853,
      "epoch": 1.66931321953156,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004726484102676169,
      "loss": 5.3143,
      "mean_token_accuracy": 0.1549732580780983,
      "num_tokens": 36858730.0,
      "step": 16820
    },
    {
      "entropy": 5.508907413482666,
      "epoch": 1.6698094481937278,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004726314766116223,
      "loss": 5.1763,
      "mean_token_accuracy": 0.15831146389245987,
      "num_tokens": 36868239.0,
      "step": 16825
    },
    {
      "entropy": 5.5061235427856445,
      "epoch": 1.6703056768558953,
      "grad_norm": 1.6875,
      "learning_rate": 0.000472614538054793,
      "loss": 5.2752,
      "mean_token_accuracy": 0.1591561570763588,
      "num_tokens": 36878077.0,
      "step": 16830
    },
    {
      "entropy": 5.558781671524048,
      "epoch": 1.6708019055180627,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047259759459754893,
      "loss": 5.3019,
      "mean_token_accuracy": 0.1484263464808464,
      "num_tokens": 36888874.0,
      "step": 16835
    },
    {
      "entropy": 5.57882604598999,
      "epoch": 1.6712981341802302,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004725806462403103,
      "loss": 5.3199,
      "mean_token_accuracy": 0.1567700795829296,
      "num_tokens": 36898584.0,
      "step": 16840
    },
    {
      "entropy": 5.565012836456299,
      "epoch": 1.6717943628423977,
      "grad_norm": 1.375,
      "learning_rate": 0.00047256369298349736,
      "loss": 5.3168,
      "mean_token_accuracy": 0.15655829459428788,
      "num_tokens": 36908662.0,
      "step": 16845
    },
    {
      "entropy": 5.532862758636474,
      "epoch": 1.6722905915045652,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047254673482753065,
      "loss": 5.2487,
      "mean_token_accuracy": 0.15651735812425613,
      "num_tokens": 36919482.0,
      "step": 16850
    },
    {
      "entropy": 5.570927858352661,
      "epoch": 1.6727868201667329,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047252977177283045,
      "loss": 5.2765,
      "mean_token_accuracy": 0.15110802203416823,
      "num_tokens": 36930470.0,
      "step": 16855
    },
    {
      "entropy": 5.554264736175537,
      "epoch": 1.6732830488289003,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047251280381981764,
      "loss": 5.2801,
      "mean_token_accuracy": 0.16025215089321138,
      "num_tokens": 36942837.0,
      "step": 16860
    },
    {
      "entropy": 5.588925695419311,
      "epoch": 1.673779277491068,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004724958309689129,
      "loss": 5.3557,
      "mean_token_accuracy": 0.15434966385364532,
      "num_tokens": 36952994.0,
      "step": 16865
    },
    {
      "entropy": 5.549069356918335,
      "epoch": 1.6742755061532355,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047247885322053705,
      "loss": 5.345,
      "mean_token_accuracy": 0.15192511528730393,
      "num_tokens": 36963973.0,
      "step": 16870
    },
    {
      "entropy": 5.484513950347901,
      "epoch": 1.674771734815403,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047246187057511127,
      "loss": 5.2711,
      "mean_token_accuracy": 0.15875779539346696,
      "num_tokens": 36975897.0,
      "step": 16875
    },
    {
      "entropy": 5.55510950088501,
      "epoch": 1.6752679634775705,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004724448830330565,
      "loss": 5.2956,
      "mean_token_accuracy": 0.15356604307889937,
      "num_tokens": 36986481.0,
      "step": 16880
    },
    {
      "entropy": 5.586087799072265,
      "epoch": 1.675764192139738,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047242789059479413,
      "loss": 5.3754,
      "mean_token_accuracy": 0.15363562554121019,
      "num_tokens": 36997863.0,
      "step": 16885
    },
    {
      "entropy": 5.611656713485718,
      "epoch": 1.6762604208019054,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047241089326074547,
      "loss": 5.337,
      "mean_token_accuracy": 0.15122737735509872,
      "num_tokens": 37009294.0,
      "step": 16890
    },
    {
      "entropy": 5.579777383804322,
      "epoch": 1.6767566494640729,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004723938910313321,
      "loss": 5.3022,
      "mean_token_accuracy": 0.16070909351110457,
      "num_tokens": 37019748.0,
      "step": 16895
    },
    {
      "entropy": 5.57478895187378,
      "epoch": 1.6772528781262406,
      "grad_norm": 1.625,
      "learning_rate": 0.00047237688390697555,
      "loss": 5.3721,
      "mean_token_accuracy": 0.15824934095144272,
      "num_tokens": 37030903.0,
      "step": 16900
    },
    {
      "entropy": 5.495475244522095,
      "epoch": 1.677749106788408,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004723598718880975,
      "loss": 5.2958,
      "mean_token_accuracy": 0.15939802080392837,
      "num_tokens": 37042551.0,
      "step": 16905
    },
    {
      "entropy": 5.630831623077393,
      "epoch": 1.6782453354505757,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047234285497512005,
      "loss": 5.3549,
      "mean_token_accuracy": 0.1499130293726921,
      "num_tokens": 37053151.0,
      "step": 16910
    },
    {
      "entropy": 5.638395881652832,
      "epoch": 1.6787415641127432,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047232583316846494,
      "loss": 5.3759,
      "mean_token_accuracy": 0.14672162905335426,
      "num_tokens": 37064421.0,
      "step": 16915
    },
    {
      "entropy": 5.56446681022644,
      "epoch": 1.6792377927749107,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004723088064685544,
      "loss": 5.3086,
      "mean_token_accuracy": 0.16018568575382233,
      "num_tokens": 37074758.0,
      "step": 16920
    },
    {
      "entropy": 5.584210538864136,
      "epoch": 1.6797340214370782,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047229177487581055,
      "loss": 5.2529,
      "mean_token_accuracy": 0.1550099015235901,
      "num_tokens": 37084898.0,
      "step": 16925
    },
    {
      "entropy": 5.5861358642578125,
      "epoch": 1.6802302500992456,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047227473839065587,
      "loss": 5.3437,
      "mean_token_accuracy": 0.15191148668527604,
      "num_tokens": 37094366.0,
      "step": 16930
    },
    {
      "entropy": 5.671796894073486,
      "epoch": 1.6807264787614131,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047225769701351284,
      "loss": 5.3994,
      "mean_token_accuracy": 0.1574471414089203,
      "num_tokens": 37104965.0,
      "step": 16935
    },
    {
      "entropy": 5.521795988082886,
      "epoch": 1.6812227074235808,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004722406507448038,
      "loss": 5.2854,
      "mean_token_accuracy": 0.1530693456530571,
      "num_tokens": 37115622.0,
      "step": 16940
    },
    {
      "entropy": 5.603211402893066,
      "epoch": 1.6817189360857483,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047222359958495175,
      "loss": 5.3557,
      "mean_token_accuracy": 0.1503603234887123,
      "num_tokens": 37127234.0,
      "step": 16945
    },
    {
      "entropy": 5.576606369018554,
      "epoch": 1.682215164747916,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004722065435343794,
      "loss": 5.2248,
      "mean_token_accuracy": 0.16062504500150682,
      "num_tokens": 37136533.0,
      "step": 16950
    },
    {
      "entropy": 5.5512103080749515,
      "epoch": 1.6827113934100835,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004721894825935097,
      "loss": 5.337,
      "mean_token_accuracy": 0.1524190902709961,
      "num_tokens": 37147821.0,
      "step": 16955
    },
    {
      "entropy": 5.521806383132935,
      "epoch": 1.683207622072251,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004721724167627657,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1565449357032776,
      "num_tokens": 37158929.0,
      "step": 16960
    },
    {
      "entropy": 5.6376128673553465,
      "epoch": 1.6837038507344184,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004721553460425708,
      "loss": 5.3517,
      "mean_token_accuracy": 0.16015688627958297,
      "num_tokens": 37169817.0,
      "step": 16965
    },
    {
      "entropy": 5.62782473564148,
      "epoch": 1.684200079396586,
      "grad_norm": 2.34375,
      "learning_rate": 0.000472138270433348,
      "loss": 5.4626,
      "mean_token_accuracy": 0.14678781107068062,
      "num_tokens": 37181009.0,
      "step": 16970
    },
    {
      "entropy": 5.551593208312989,
      "epoch": 1.6846963080587534,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047212118993552097,
      "loss": 5.2817,
      "mean_token_accuracy": 0.15696789175271988,
      "num_tokens": 37192110.0,
      "step": 16975
    },
    {
      "entropy": 5.572682809829712,
      "epoch": 1.6851925367209208,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004721041045495131,
      "loss": 5.3074,
      "mean_token_accuracy": 0.15926933139562607,
      "num_tokens": 37203426.0,
      "step": 16980
    },
    {
      "entropy": 5.64717059135437,
      "epoch": 1.6856887653830885,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047208701427574817,
      "loss": 5.4222,
      "mean_token_accuracy": 0.14812846779823302,
      "num_tokens": 37215025.0,
      "step": 16985
    },
    {
      "entropy": 5.660180330276489,
      "epoch": 1.686184994045256,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047206991911465,
      "loss": 5.3308,
      "mean_token_accuracy": 0.15768174976110458,
      "num_tokens": 37225653.0,
      "step": 16990
    },
    {
      "entropy": 5.563625383377075,
      "epoch": 1.6866812227074237,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004720528190666425,
      "loss": 5.3133,
      "mean_token_accuracy": 0.1534594789147377,
      "num_tokens": 37237282.0,
      "step": 16995
    },
    {
      "entropy": 5.489155864715576,
      "epoch": 1.6871774513695912,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047203571413214967,
      "loss": 5.2656,
      "mean_token_accuracy": 0.1669609799981117,
      "num_tokens": 37248476.0,
      "step": 17000
    },
    {
      "entropy": 5.593642997741699,
      "epoch": 1.6876736800317587,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004720186043115957,
      "loss": 5.2581,
      "mean_token_accuracy": 0.15342743843793868,
      "num_tokens": 37259354.0,
      "step": 17005
    },
    {
      "entropy": 5.58482027053833,
      "epoch": 1.6881699086939261,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004720014896054048,
      "loss": 5.3316,
      "mean_token_accuracy": 0.1536075085401535,
      "num_tokens": 37269895.0,
      "step": 17010
    },
    {
      "entropy": 5.540448474884033,
      "epoch": 1.6886661373560936,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004719843700140017,
      "loss": 5.3031,
      "mean_token_accuracy": 0.16014419198036195,
      "num_tokens": 37281201.0,
      "step": 17015
    },
    {
      "entropy": 5.527393007278443,
      "epoch": 1.689162366018261,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004719672455378105,
      "loss": 5.3246,
      "mean_token_accuracy": 0.1528385803103447,
      "num_tokens": 37293362.0,
      "step": 17020
    },
    {
      "entropy": 5.6267742156982425,
      "epoch": 1.6896585946804288,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000471950116177256,
      "loss": 5.3451,
      "mean_token_accuracy": 0.15028529167175292,
      "num_tokens": 37304291.0,
      "step": 17025
    },
    {
      "entropy": 5.621040725708008,
      "epoch": 1.6901548233425963,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004719329819327631,
      "loss": 5.3173,
      "mean_token_accuracy": 0.15144096463918685,
      "num_tokens": 37315608.0,
      "step": 17030
    },
    {
      "entropy": 5.561187791824341,
      "epoch": 1.690651052004764,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004719158428047566,
      "loss": 5.2697,
      "mean_token_accuracy": 0.1527626022696495,
      "num_tokens": 37327559.0,
      "step": 17035
    },
    {
      "entropy": 5.522388315200805,
      "epoch": 1.6911472806669314,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004718986987936615,
      "loss": 5.3101,
      "mean_token_accuracy": 0.15939072519540787,
      "num_tokens": 37338101.0,
      "step": 17040
    },
    {
      "entropy": 5.570981359481811,
      "epoch": 1.691643509329099,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047188154989990296,
      "loss": 5.3973,
      "mean_token_accuracy": 0.15344681069254876,
      "num_tokens": 37349124.0,
      "step": 17045
    },
    {
      "entropy": 5.560146331787109,
      "epoch": 1.6921397379912664,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004718643961239062,
      "loss": 5.257,
      "mean_token_accuracy": 0.16170873939990998,
      "num_tokens": 37360061.0,
      "step": 17050
    },
    {
      "entropy": 5.579184865951538,
      "epoch": 1.6926359666534339,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047184723746609666,
      "loss": 5.3326,
      "mean_token_accuracy": 0.1584179386496544,
      "num_tokens": 37371210.0,
      "step": 17055
    },
    {
      "entropy": 5.539211654663086,
      "epoch": 1.6931321953156013,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047183007392689977,
      "loss": 5.2673,
      "mean_token_accuracy": 0.15691950917243958,
      "num_tokens": 37382547.0,
      "step": 17060
    },
    {
      "entropy": 5.56604700088501,
      "epoch": 1.6936284239777688,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047181290550674115,
      "loss": 5.3245,
      "mean_token_accuracy": 0.16095831543207167,
      "num_tokens": 37392311.0,
      "step": 17065
    },
    {
      "entropy": 5.592380952835083,
      "epoch": 1.6941246526399365,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004717957322060466,
      "loss": 5.2785,
      "mean_token_accuracy": 0.1578421860933304,
      "num_tokens": 37402626.0,
      "step": 17070
    },
    {
      "entropy": 5.574470567703247,
      "epoch": 1.694620881302104,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004717785540252419,
      "loss": 5.2798,
      "mean_token_accuracy": 0.15093974173069,
      "num_tokens": 37412100.0,
      "step": 17075
    },
    {
      "entropy": 5.470191574096679,
      "epoch": 1.6951171099642717,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004717613709647532,
      "loss": 5.177,
      "mean_token_accuracy": 0.160963237285614,
      "num_tokens": 37421735.0,
      "step": 17080
    },
    {
      "entropy": 5.607144689559936,
      "epoch": 1.6956133386264391,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047174418302500635,
      "loss": 5.3989,
      "mean_token_accuracy": 0.14853432029485703,
      "num_tokens": 37432179.0,
      "step": 17085
    },
    {
      "entropy": 5.550100421905517,
      "epoch": 1.6961095672886066,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004717269902064278,
      "loss": 5.2359,
      "mean_token_accuracy": 0.16764892935752868,
      "num_tokens": 37443276.0,
      "step": 17090
    },
    {
      "entropy": 5.5500537872314455,
      "epoch": 1.696605795950774,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004717097925094438,
      "loss": 5.3655,
      "mean_token_accuracy": 0.15181733369827272,
      "num_tokens": 37454537.0,
      "step": 17095
    },
    {
      "entropy": 5.558017683029175,
      "epoch": 1.6971020246129416,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004716925899344808,
      "loss": 5.2774,
      "mean_token_accuracy": 0.15302327424287795,
      "num_tokens": 37465493.0,
      "step": 17100
    },
    {
      "entropy": 5.515025949478149,
      "epoch": 1.697598253275109,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004716753824819654,
      "loss": 5.2644,
      "mean_token_accuracy": 0.15756237804889678,
      "num_tokens": 37475511.0,
      "step": 17105
    },
    {
      "entropy": 5.50667405128479,
      "epoch": 1.6980944819372767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004716581701523243,
      "loss": 5.1881,
      "mean_token_accuracy": 0.1556583359837532,
      "num_tokens": 37487046.0,
      "step": 17110
    },
    {
      "entropy": 5.567019939422607,
      "epoch": 1.6985907105994442,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047164095294598433,
      "loss": 5.3495,
      "mean_token_accuracy": 0.15532005131244658,
      "num_tokens": 37498398.0,
      "step": 17115
    },
    {
      "entropy": 5.573226594924927,
      "epoch": 1.699086939261612,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047162373086337243,
      "loss": 5.3411,
      "mean_token_accuracy": 0.1551344722509384,
      "num_tokens": 37509397.0,
      "step": 17120
    },
    {
      "entropy": 5.543571949005127,
      "epoch": 1.6995831679237794,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047160650390491574,
      "loss": 5.2299,
      "mean_token_accuracy": 0.16128843873739243,
      "num_tokens": 37519284.0,
      "step": 17125
    },
    {
      "entropy": 5.576875972747803,
      "epoch": 1.7000793965859469,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004715892720710414,
      "loss": 5.3305,
      "mean_token_accuracy": 0.15937667191028596,
      "num_tokens": 37529699.0,
      "step": 17130
    },
    {
      "entropy": 5.63898811340332,
      "epoch": 1.7005756252481143,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004715720353621767,
      "loss": 5.3838,
      "mean_token_accuracy": 0.14993614703416824,
      "num_tokens": 37541756.0,
      "step": 17135
    },
    {
      "entropy": 5.5366870880126955,
      "epoch": 1.7010718539102818,
      "grad_norm": 1.25,
      "learning_rate": 0.00047155479377874917,
      "loss": 5.2552,
      "mean_token_accuracy": 0.16365874111652373,
      "num_tokens": 37552693.0,
      "step": 17140
    },
    {
      "entropy": 5.562690687179566,
      "epoch": 1.7015680825724493,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047153754732118616,
      "loss": 5.248,
      "mean_token_accuracy": 0.1595419839024544,
      "num_tokens": 37563826.0,
      "step": 17145
    },
    {
      "entropy": 5.571802473068237,
      "epoch": 1.7020643112346168,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004715202959899156,
      "loss": 5.3036,
      "mean_token_accuracy": 0.1549629360437393,
      "num_tokens": 37576645.0,
      "step": 17150
    },
    {
      "entropy": 5.463922739028931,
      "epoch": 1.7025605398967845,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004715030397853651,
      "loss": 5.1752,
      "mean_token_accuracy": 0.16198420748114586,
      "num_tokens": 37587467.0,
      "step": 17155
    },
    {
      "entropy": 5.542828941345215,
      "epoch": 1.703056768558952,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047148577870796264,
      "loss": 5.3419,
      "mean_token_accuracy": 0.15009882897138596,
      "num_tokens": 37598475.0,
      "step": 17160
    },
    {
      "entropy": 5.662155437469482,
      "epoch": 1.7035529972211196,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004714685127581363,
      "loss": 5.4323,
      "mean_token_accuracy": 0.1517896145582199,
      "num_tokens": 37610260.0,
      "step": 17165
    },
    {
      "entropy": 5.616348600387573,
      "epoch": 1.704049225883287,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047145124193631415,
      "loss": 5.3585,
      "mean_token_accuracy": 0.15538112372159957,
      "num_tokens": 37621129.0,
      "step": 17170
    },
    {
      "entropy": 5.547582912445068,
      "epoch": 1.7045454545454546,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047143396624292454,
      "loss": 5.2678,
      "mean_token_accuracy": 0.1562929093837738,
      "num_tokens": 37633130.0,
      "step": 17175
    },
    {
      "entropy": 5.606756973266601,
      "epoch": 1.705041683207622,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00047141668567839584,
      "loss": 5.3917,
      "mean_token_accuracy": 0.1532363310456276,
      "num_tokens": 37644247.0,
      "step": 17180
    },
    {
      "entropy": 5.612619686126709,
      "epoch": 1.7055379118697895,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004713994002431566,
      "loss": 5.3238,
      "mean_token_accuracy": 0.15229059606790543,
      "num_tokens": 37655602.0,
      "step": 17185
    },
    {
      "entropy": 5.512532186508179,
      "epoch": 1.706034140531957,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047138210993763534,
      "loss": 5.2392,
      "mean_token_accuracy": 0.15669836699962617,
      "num_tokens": 37667150.0,
      "step": 17190
    },
    {
      "entropy": 5.507809162139893,
      "epoch": 1.7065303691941247,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000471364814762261,
      "loss": 5.2436,
      "mean_token_accuracy": 0.1620771273970604,
      "num_tokens": 37677626.0,
      "step": 17195
    },
    {
      "entropy": 5.633812999725341,
      "epoch": 1.7070265978562922,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004713475147174624,
      "loss": 5.3828,
      "mean_token_accuracy": 0.1537002980709076,
      "num_tokens": 37689050.0,
      "step": 17200
    },
    {
      "entropy": 5.664777755737305,
      "epoch": 1.7075228265184597,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004713302098036685,
      "loss": 5.2681,
      "mean_token_accuracy": 0.16411532014608382,
      "num_tokens": 37699190.0,
      "step": 17205
    },
    {
      "entropy": 5.4560198307037355,
      "epoch": 1.7080190551806274,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004713129000213085,
      "loss": 5.2626,
      "mean_token_accuracy": 0.1605716422200203,
      "num_tokens": 37710465.0,
      "step": 17210
    },
    {
      "entropy": 5.522910213470459,
      "epoch": 1.7085152838427948,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004712955853708115,
      "loss": 5.3688,
      "mean_token_accuracy": 0.15560415685176848,
      "num_tokens": 37722426.0,
      "step": 17215
    },
    {
      "entropy": 5.557200336456299,
      "epoch": 1.7090115125049623,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000471278265852607,
      "loss": 5.2665,
      "mean_token_accuracy": 0.15996670722961426,
      "num_tokens": 37733104.0,
      "step": 17220
    },
    {
      "entropy": 5.526095724105835,
      "epoch": 1.7095077411671298,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004712609414671245,
      "loss": 5.2021,
      "mean_token_accuracy": 0.16054170429706574,
      "num_tokens": 37742747.0,
      "step": 17225
    },
    {
      "entropy": 5.519110012054443,
      "epoch": 1.7100039698292973,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004712436122147935,
      "loss": 5.2548,
      "mean_token_accuracy": 0.15934465378522872,
      "num_tokens": 37753338.0,
      "step": 17230
    },
    {
      "entropy": 5.5226866722106935,
      "epoch": 1.7105001984914647,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004712262780960438,
      "loss": 5.2313,
      "mean_token_accuracy": 0.1620363861322403,
      "num_tokens": 37763122.0,
      "step": 17235
    },
    {
      "entropy": 5.5014573097229,
      "epoch": 1.7109964271536324,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047120893911130525,
      "loss": 5.2414,
      "mean_token_accuracy": 0.15666421204805375,
      "num_tokens": 37774677.0,
      "step": 17240
    },
    {
      "entropy": 5.519226121902466,
      "epoch": 1.7114926558158,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047119159526100773,
      "loss": 5.2522,
      "mean_token_accuracy": 0.14980335980653764,
      "num_tokens": 37786186.0,
      "step": 17245
    },
    {
      "entropy": 5.517546653747559,
      "epoch": 1.7119888844779676,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047117424654558146,
      "loss": 5.2891,
      "mean_token_accuracy": 0.16381997764110565,
      "num_tokens": 37796837.0,
      "step": 17250
    },
    {
      "entropy": 5.585890579223633,
      "epoch": 1.712485113140135,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004711568929654566,
      "loss": 5.3462,
      "mean_token_accuracy": 0.15599875301122665,
      "num_tokens": 37808813.0,
      "step": 17255
    },
    {
      "entropy": 5.499355745315552,
      "epoch": 1.7129813418023025,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004711395345210635,
      "loss": 5.2124,
      "mean_token_accuracy": 0.16025424599647523,
      "num_tokens": 37819770.0,
      "step": 17260
    },
    {
      "entropy": 5.541247844696045,
      "epoch": 1.71347757046447,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004711221712128325,
      "loss": 5.2973,
      "mean_token_accuracy": 0.15971039682626725,
      "num_tokens": 37831831.0,
      "step": 17265
    },
    {
      "entropy": 5.545330762863159,
      "epoch": 1.7139737991266375,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004711048030411943,
      "loss": 5.3331,
      "mean_token_accuracy": 0.1545376867055893,
      "num_tokens": 37843092.0,
      "step": 17270
    },
    {
      "entropy": 5.533629083633423,
      "epoch": 1.714470027788805,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047108743000657963,
      "loss": 5.2843,
      "mean_token_accuracy": 0.1695124998688698,
      "num_tokens": 37853192.0,
      "step": 17275
    },
    {
      "entropy": 5.661731433868408,
      "epoch": 1.7149662564509724,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004710700521094192,
      "loss": 5.3211,
      "mean_token_accuracy": 0.1533560946583748,
      "num_tokens": 37863510.0,
      "step": 17280
    },
    {
      "entropy": 5.54509859085083,
      "epoch": 1.7154624851131401,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047105266935014385,
      "loss": 5.2534,
      "mean_token_accuracy": 0.16175752282142639,
      "num_tokens": 37876156.0,
      "step": 17285
    },
    {
      "entropy": 5.49203634262085,
      "epoch": 1.7159587137753076,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047103528172918487,
      "loss": 5.2851,
      "mean_token_accuracy": 0.15938131213188172,
      "num_tokens": 37887153.0,
      "step": 17290
    },
    {
      "entropy": 5.589544200897217,
      "epoch": 1.7164549424374753,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004710178892469733,
      "loss": 5.2836,
      "mean_token_accuracy": 0.15960506796836854,
      "num_tokens": 37898820.0,
      "step": 17295
    },
    {
      "entropy": 5.577250623703003,
      "epoch": 1.7169511710996428,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047100049190394044,
      "loss": 5.2572,
      "mean_token_accuracy": 0.15494226217269896,
      "num_tokens": 37910257.0,
      "step": 17300
    },
    {
      "entropy": 5.516682863235474,
      "epoch": 1.7174473997618103,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004709830897005178,
      "loss": 5.3641,
      "mean_token_accuracy": 0.15409231930971146,
      "num_tokens": 37921574.0,
      "step": 17305
    },
    {
      "entropy": 5.588931131362915,
      "epoch": 1.7179436284239777,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004709656826371367,
      "loss": 5.3358,
      "mean_token_accuracy": 0.15544934421777726,
      "num_tokens": 37932741.0,
      "step": 17310
    },
    {
      "entropy": 5.585829830169677,
      "epoch": 1.7184398570861452,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004709482707142291,
      "loss": 5.3051,
      "mean_token_accuracy": 0.15546197444200516,
      "num_tokens": 37944592.0,
      "step": 17315
    },
    {
      "entropy": 5.552493953704834,
      "epoch": 1.7189360857483127,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047093085393222646,
      "loss": 5.2533,
      "mean_token_accuracy": 0.16265860348939895,
      "num_tokens": 37954983.0,
      "step": 17320
    },
    {
      "entropy": 5.5698655128479,
      "epoch": 1.7194323144104804,
      "grad_norm": 1.828125,
      "learning_rate": 0.00047091343229156094,
      "loss": 5.314,
      "mean_token_accuracy": 0.1587136685848236,
      "num_tokens": 37964491.0,
      "step": 17325
    },
    {
      "entropy": 5.542346048355102,
      "epoch": 1.7199285430726479,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004708960057926644,
      "loss": 5.2073,
      "mean_token_accuracy": 0.16317139565944672,
      "num_tokens": 37976140.0,
      "step": 17330
    },
    {
      "entropy": 5.5543430805206295,
      "epoch": 1.7204247717348156,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047087857443596905,
      "loss": 5.2793,
      "mean_token_accuracy": 0.1609412834048271,
      "num_tokens": 37987776.0,
      "step": 17335
    },
    {
      "entropy": 5.537221765518188,
      "epoch": 1.720921000396983,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047086113822190716,
      "loss": 5.3304,
      "mean_token_accuracy": 0.15319294035434722,
      "num_tokens": 37998088.0,
      "step": 17340
    },
    {
      "entropy": 5.48595404624939,
      "epoch": 1.7214172290591505,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047084369715091103,
      "loss": 5.3274,
      "mean_token_accuracy": 0.14934306740760803,
      "num_tokens": 38009802.0,
      "step": 17345
    },
    {
      "entropy": 5.595892810821534,
      "epoch": 1.721913457721318,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004708262512234133,
      "loss": 5.3657,
      "mean_token_accuracy": 0.16491689532995224,
      "num_tokens": 38020797.0,
      "step": 17350
    },
    {
      "entropy": 5.6584601402282715,
      "epoch": 1.7224096863834855,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004708088004398464,
      "loss": 5.3632,
      "mean_token_accuracy": 0.15144674032926558,
      "num_tokens": 38031816.0,
      "step": 17355
    },
    {
      "entropy": 5.59825701713562,
      "epoch": 1.722905915045653,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004707913448006433,
      "loss": 5.3412,
      "mean_token_accuracy": 0.1491363100707531,
      "num_tokens": 38043953.0,
      "step": 17360
    },
    {
      "entropy": 5.495429515838623,
      "epoch": 1.7234021437078204,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004707738843062367,
      "loss": 5.2101,
      "mean_token_accuracy": 0.167082978785038,
      "num_tokens": 38053929.0,
      "step": 17365
    },
    {
      "entropy": 5.636122131347657,
      "epoch": 1.723898372369988,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004707564189570596,
      "loss": 5.4539,
      "mean_token_accuracy": 0.15333012044429778,
      "num_tokens": 38067091.0,
      "step": 17370
    },
    {
      "entropy": 5.657432413101196,
      "epoch": 1.7243946010321556,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004707389487535451,
      "loss": 5.2955,
      "mean_token_accuracy": 0.15900218933820726,
      "num_tokens": 38079002.0,
      "step": 17375
    },
    {
      "entropy": 5.591316556930542,
      "epoch": 1.7248908296943233,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004707214736961265,
      "loss": 5.2987,
      "mean_token_accuracy": 0.15801535099744796,
      "num_tokens": 38089526.0,
      "step": 17380
    },
    {
      "entropy": 5.539578104019165,
      "epoch": 1.7253870583564908,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004707039937852371,
      "loss": 5.2771,
      "mean_token_accuracy": 0.15809779912233352,
      "num_tokens": 38099902.0,
      "step": 17385
    },
    {
      "entropy": 5.507629823684693,
      "epoch": 1.7258832870186582,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047068650902131034,
      "loss": 5.253,
      "mean_token_accuracy": 0.16492314785718917,
      "num_tokens": 38111725.0,
      "step": 17390
    },
    {
      "entropy": 5.575964021682739,
      "epoch": 1.7263795156808257,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004706690194047798,
      "loss": 5.2251,
      "mean_token_accuracy": 0.1590568721294403,
      "num_tokens": 38123522.0,
      "step": 17395
    },
    {
      "entropy": 5.612242364883423,
      "epoch": 1.7268757443429932,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004706515249360792,
      "loss": 5.3447,
      "mean_token_accuracy": 0.15331568270921708,
      "num_tokens": 38133968.0,
      "step": 17400
    },
    {
      "entropy": 5.5302082061767575,
      "epoch": 1.7273719730051607,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004706340256156425,
      "loss": 5.3979,
      "mean_token_accuracy": 0.14713698625564575,
      "num_tokens": 38145288.0,
      "step": 17405
    },
    {
      "entropy": 5.58294997215271,
      "epoch": 1.7278682016673284,
      "grad_norm": 2.75,
      "learning_rate": 0.0004706165214439034,
      "loss": 5.2986,
      "mean_token_accuracy": 0.15334606319665908,
      "num_tokens": 38156141.0,
      "step": 17410
    },
    {
      "entropy": 5.638806390762329,
      "epoch": 1.7283644303294958,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004705990124212961,
      "loss": 5.2736,
      "mean_token_accuracy": 0.16556357145309447,
      "num_tokens": 38167211.0,
      "step": 17415
    },
    {
      "entropy": 5.533245277404785,
      "epoch": 1.7288606589916635,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004705814985482548,
      "loss": 5.3117,
      "mean_token_accuracy": 0.15492815524339676,
      "num_tokens": 38177360.0,
      "step": 17420
    },
    {
      "entropy": 5.495733690261841,
      "epoch": 1.729356887653831,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00047056397982521366,
      "loss": 5.221,
      "mean_token_accuracy": 0.16426936089992522,
      "num_tokens": 38187857.0,
      "step": 17425
    },
    {
      "entropy": 5.596555709838867,
      "epoch": 1.7298531163159985,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004705464562526074,
      "loss": 5.3314,
      "mean_token_accuracy": 0.14911005049943923,
      "num_tokens": 38198161.0,
      "step": 17430
    },
    {
      "entropy": 5.570596551895141,
      "epoch": 1.730349344978166,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004705289278308703,
      "loss": 5.2822,
      "mean_token_accuracy": 0.16063248068094255,
      "num_tokens": 38208612.0,
      "step": 17435
    },
    {
      "entropy": 5.570613050460816,
      "epoch": 1.7308455736403334,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047051139456043707,
      "loss": 5.3241,
      "mean_token_accuracy": 0.15265615433454513,
      "num_tokens": 38220153.0,
      "step": 17440
    },
    {
      "entropy": 5.539574527740479,
      "epoch": 1.731341802302501,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004704938564417427,
      "loss": 5.2779,
      "mean_token_accuracy": 0.16104727387428283,
      "num_tokens": 38230790.0,
      "step": 17445
    },
    {
      "entropy": 5.599995565414429,
      "epoch": 1.7318380309646684,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047047631347522175,
      "loss": 5.2449,
      "mean_token_accuracy": 0.1610218405723572,
      "num_tokens": 38242585.0,
      "step": 17450
    },
    {
      "entropy": 5.5565718650817875,
      "epoch": 1.732334259626836,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047045876566130957,
      "loss": 5.2866,
      "mean_token_accuracy": 0.15610651522874833,
      "num_tokens": 38253316.0,
      "step": 17455
    },
    {
      "entropy": 5.5149177551269535,
      "epoch": 1.7328304882890035,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004704412130004411,
      "loss": 5.2551,
      "mean_token_accuracy": 0.16203157901763915,
      "num_tokens": 38263230.0,
      "step": 17460
    },
    {
      "entropy": 5.617604398727417,
      "epoch": 1.7333267169511712,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00047042365549305174,
      "loss": 5.3354,
      "mean_token_accuracy": 0.153751440346241,
      "num_tokens": 38273416.0,
      "step": 17465
    },
    {
      "entropy": 5.543125057220459,
      "epoch": 1.7338229456133387,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004704060931395768,
      "loss": 5.2669,
      "mean_token_accuracy": 0.15701694041490555,
      "num_tokens": 38285764.0,
      "step": 17470
    },
    {
      "entropy": 5.585240077972412,
      "epoch": 1.7343191742755062,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047038852594045186,
      "loss": 5.3636,
      "mean_token_accuracy": 0.1569289669394493,
      "num_tokens": 38296899.0,
      "step": 17475
    },
    {
      "entropy": 5.639583969116211,
      "epoch": 1.7348154029376737,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047037095389611245,
      "loss": 5.3598,
      "mean_token_accuracy": 0.15980358570814132,
      "num_tokens": 38309050.0,
      "step": 17480
    },
    {
      "entropy": 5.549654245376587,
      "epoch": 1.7353116315998411,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004703533770069944,
      "loss": 5.3154,
      "mean_token_accuracy": 0.1576733723282814,
      "num_tokens": 38320831.0,
      "step": 17485
    },
    {
      "entropy": 5.550827264785767,
      "epoch": 1.7358078602620086,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047033579527353355,
      "loss": 5.3188,
      "mean_token_accuracy": 0.1538187585771084,
      "num_tokens": 38332096.0,
      "step": 17490
    },
    {
      "entropy": 5.6151097297668455,
      "epoch": 1.7363040889241763,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004703182086961659,
      "loss": 5.3419,
      "mean_token_accuracy": 0.15780805796384811,
      "num_tokens": 38343110.0,
      "step": 17495
    },
    {
      "entropy": 5.614585018157959,
      "epoch": 1.7368003175863438,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004703006172753274,
      "loss": 5.3329,
      "mean_token_accuracy": 0.16204356253147126,
      "num_tokens": 38353971.0,
      "step": 17500
    },
    {
      "entropy": 5.513349342346191,
      "epoch": 1.7372965462485113,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047028302101145467,
      "loss": 5.2598,
      "mean_token_accuracy": 0.1626959577202797,
      "num_tokens": 38365114.0,
      "step": 17505
    },
    {
      "entropy": 5.63951301574707,
      "epoch": 1.737792774910679,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004702654199049836,
      "loss": 5.2759,
      "mean_token_accuracy": 0.15487973839044572,
      "num_tokens": 38375784.0,
      "step": 17510
    },
    {
      "entropy": 5.545527791976928,
      "epoch": 1.7382890035728464,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000470247813956351,
      "loss": 5.2643,
      "mean_token_accuracy": 0.15653037428855895,
      "num_tokens": 38386053.0,
      "step": 17515
    },
    {
      "entropy": 5.450185871124267,
      "epoch": 1.738785232235014,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047023020316599334,
      "loss": 5.253,
      "mean_token_accuracy": 0.15928541570901872,
      "num_tokens": 38396545.0,
      "step": 17520
    },
    {
      "entropy": 5.538559150695801,
      "epoch": 1.7392814608971814,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047021258753434725,
      "loss": 5.2275,
      "mean_token_accuracy": 0.15978810042142869,
      "num_tokens": 38406344.0,
      "step": 17525
    },
    {
      "entropy": 5.559768581390381,
      "epoch": 1.7397776895593489,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047019496706184966,
      "loss": 5.2154,
      "mean_token_accuracy": 0.16572791039943696,
      "num_tokens": 38416388.0,
      "step": 17530
    },
    {
      "entropy": 5.60798830986023,
      "epoch": 1.7402739182215163,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004701773417489375,
      "loss": 5.3721,
      "mean_token_accuracy": 0.14739905446767806,
      "num_tokens": 38427608.0,
      "step": 17535
    },
    {
      "entropy": 5.596655035018921,
      "epoch": 1.740770146883684,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047015971159604787,
      "loss": 5.2669,
      "mean_token_accuracy": 0.15767776519060134,
      "num_tokens": 38437921.0,
      "step": 17540
    },
    {
      "entropy": 5.566097831726074,
      "epoch": 1.7412663755458515,
      "grad_norm": 1.640625,
      "learning_rate": 0.000470142076603618,
      "loss": 5.3366,
      "mean_token_accuracy": 0.15474507808685303,
      "num_tokens": 38448767.0,
      "step": 17545
    },
    {
      "entropy": 5.567763996124268,
      "epoch": 1.7417626042080192,
      "grad_norm": 1.640625,
      "learning_rate": 0.00047012443677208496,
      "loss": 5.4064,
      "mean_token_accuracy": 0.1521571010351181,
      "num_tokens": 38460052.0,
      "step": 17550
    },
    {
      "entropy": 5.537759494781494,
      "epoch": 1.7422588328701867,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004701067921018864,
      "loss": 5.2476,
      "mean_token_accuracy": 0.1598023623228073,
      "num_tokens": 38470177.0,
      "step": 17555
    },
    {
      "entropy": 5.597159385681152,
      "epoch": 1.7427550615323542,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047008914259345986,
      "loss": 5.2746,
      "mean_token_accuracy": 0.1537667542695999,
      "num_tokens": 38481608.0,
      "step": 17560
    },
    {
      "entropy": 5.578896141052246,
      "epoch": 1.7432512901945216,
      "grad_norm": 1.5,
      "learning_rate": 0.00047007148824724295,
      "loss": 5.338,
      "mean_token_accuracy": 0.15344153940677643,
      "num_tokens": 38492534.0,
      "step": 17565
    },
    {
      "entropy": 5.5659726619720455,
      "epoch": 1.743747518856689,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004700538290636734,
      "loss": 5.3331,
      "mean_token_accuracy": 0.15721188485622406,
      "num_tokens": 38503629.0,
      "step": 17570
    },
    {
      "entropy": 5.535835695266724,
      "epoch": 1.7442437475188566,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047003616504318925,
      "loss": 5.2319,
      "mean_token_accuracy": 0.15641989707946777,
      "num_tokens": 38514428.0,
      "step": 17575
    },
    {
      "entropy": 5.601305866241455,
      "epoch": 1.744739976181024,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047001849618622844,
      "loss": 5.3004,
      "mean_token_accuracy": 0.1533289298415184,
      "num_tokens": 38525468.0,
      "step": 17580
    },
    {
      "entropy": 5.545747089385986,
      "epoch": 1.7452362048431918,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047000082249322916,
      "loss": 5.2683,
      "mean_token_accuracy": 0.1576245367527008,
      "num_tokens": 38535095.0,
      "step": 17585
    },
    {
      "entropy": 5.567392301559448,
      "epoch": 1.7457324335053592,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004699831439646297,
      "loss": 5.2769,
      "mean_token_accuracy": 0.15296800881624223,
      "num_tokens": 38546867.0,
      "step": 17590
    },
    {
      "entropy": 5.4986546516418455,
      "epoch": 1.746228662167527,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004699654606008684,
      "loss": 5.1596,
      "mean_token_accuracy": 0.16639846861362456,
      "num_tokens": 38557231.0,
      "step": 17595
    },
    {
      "entropy": 5.588666725158691,
      "epoch": 1.7467248908296944,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004699477724023838,
      "loss": 5.3221,
      "mean_token_accuracy": 0.1553348869085312,
      "num_tokens": 38568857.0,
      "step": 17600
    },
    {
      "entropy": 5.594071340560913,
      "epoch": 1.7472211194918619,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004699300793696145,
      "loss": 5.2905,
      "mean_token_accuracy": 0.15402051508426667,
      "num_tokens": 38579177.0,
      "step": 17605
    },
    {
      "entropy": 5.5358555793762205,
      "epoch": 1.7477173481540293,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004699123815029993,
      "loss": 5.1563,
      "mean_token_accuracy": 0.165372034907341,
      "num_tokens": 38588683.0,
      "step": 17610
    },
    {
      "entropy": 5.466896820068359,
      "epoch": 1.7482135768161968,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046989467880297704,
      "loss": 5.2652,
      "mean_token_accuracy": 0.16071700155735016,
      "num_tokens": 38599636.0,
      "step": 17615
    },
    {
      "entropy": 5.626954841613769,
      "epoch": 1.7487098054783643,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046987697126998673,
      "loss": 5.3604,
      "mean_token_accuracy": 0.15165707767009734,
      "num_tokens": 38609360.0,
      "step": 17620
    },
    {
      "entropy": 5.622626924514771,
      "epoch": 1.749206034140532,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004698592589044674,
      "loss": 5.3029,
      "mean_token_accuracy": 0.15392565354704857,
      "num_tokens": 38620996.0,
      "step": 17625
    },
    {
      "entropy": 5.605477571487427,
      "epoch": 1.7497022628026995,
      "grad_norm": 1.703125,
      "learning_rate": 0.00046984154170685833,
      "loss": 5.4224,
      "mean_token_accuracy": 0.15528785735368728,
      "num_tokens": 38632935.0,
      "step": 17630
    },
    {
      "entropy": 5.615880155563355,
      "epoch": 1.7501984914648672,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004698238196775989,
      "loss": 5.3884,
      "mean_token_accuracy": 0.1533407375216484,
      "num_tokens": 38644299.0,
      "step": 17635
    },
    {
      "entropy": 5.56966757774353,
      "epoch": 1.7506947201270346,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046980609281712857,
      "loss": 5.3342,
      "mean_token_accuracy": 0.15662722885608674,
      "num_tokens": 38654970.0,
      "step": 17640
    },
    {
      "entropy": 5.581515073776245,
      "epoch": 1.7511909487892021,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004697883611258868,
      "loss": 5.3955,
      "mean_token_accuracy": 0.14752071127295494,
      "num_tokens": 38666302.0,
      "step": 17645
    },
    {
      "entropy": 5.552978563308716,
      "epoch": 1.7516871774513696,
      "grad_norm": 2.125,
      "learning_rate": 0.0004697706246043136,
      "loss": 5.2852,
      "mean_token_accuracy": 0.1581949084997177,
      "num_tokens": 38678346.0,
      "step": 17650
    },
    {
      "entropy": 5.547442197799683,
      "epoch": 1.752183406113537,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046975288325284845,
      "loss": 5.2154,
      "mean_token_accuracy": 0.15842795372009277,
      "num_tokens": 38688242.0,
      "step": 17655
    },
    {
      "entropy": 5.654832410812378,
      "epoch": 1.7526796347757045,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004697351370719315,
      "loss": 5.3951,
      "mean_token_accuracy": 0.1533050313591957,
      "num_tokens": 38698756.0,
      "step": 17660
    },
    {
      "entropy": 5.578062057495117,
      "epoch": 1.753175863437872,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004697173860620027,
      "loss": 5.3679,
      "mean_token_accuracy": 0.1517864301800728,
      "num_tokens": 38710410.0,
      "step": 17665
    },
    {
      "entropy": 5.627927017211914,
      "epoch": 1.7536720921000397,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004696996302235024,
      "loss": 5.3182,
      "mean_token_accuracy": 0.15006345510482788,
      "num_tokens": 38721917.0,
      "step": 17670
    },
    {
      "entropy": 5.593115568161011,
      "epoch": 1.7541683207622072,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046968186955687074,
      "loss": 5.33,
      "mean_token_accuracy": 0.16267426162958146,
      "num_tokens": 38733085.0,
      "step": 17675
    },
    {
      "entropy": 5.54930567741394,
      "epoch": 1.7546645494243749,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004696641040625483,
      "loss": 5.353,
      "mean_token_accuracy": 0.1543727770447731,
      "num_tokens": 38744785.0,
      "step": 17680
    },
    {
      "entropy": 5.553816413879394,
      "epoch": 1.7551607780865424,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046964633374097544,
      "loss": 5.3448,
      "mean_token_accuracy": 0.1635530784726143,
      "num_tokens": 38755626.0,
      "step": 17685
    },
    {
      "entropy": 5.607366418838501,
      "epoch": 1.7556570067487098,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046962855859259306,
      "loss": 5.22,
      "mean_token_accuracy": 0.15754439681768417,
      "num_tokens": 38766360.0,
      "step": 17690
    },
    {
      "entropy": 5.543725681304932,
      "epoch": 1.7561532354108773,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046961077861784167,
      "loss": 5.2884,
      "mean_token_accuracy": 0.1567128911614418,
      "num_tokens": 38776763.0,
      "step": 17695
    },
    {
      "entropy": 5.496878814697266,
      "epoch": 1.7566494640730448,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004695929938171624,
      "loss": 5.357,
      "mean_token_accuracy": 0.159503473341465,
      "num_tokens": 38787664.0,
      "step": 17700
    },
    {
      "entropy": 5.5409252643585205,
      "epoch": 1.7571456927352123,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004695752041909962,
      "loss": 5.2871,
      "mean_token_accuracy": 0.15526586323976516,
      "num_tokens": 38798018.0,
      "step": 17705
    },
    {
      "entropy": 5.583305788040161,
      "epoch": 1.75764192139738,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004695574097397842,
      "loss": 5.2406,
      "mean_token_accuracy": 0.15646658688783646,
      "num_tokens": 38810381.0,
      "step": 17710
    },
    {
      "entropy": 5.530179643630982,
      "epoch": 1.7581381500595474,
      "grad_norm": 1.75,
      "learning_rate": 0.00046953961046396767,
      "loss": 5.2347,
      "mean_token_accuracy": 0.1654270589351654,
      "num_tokens": 38821042.0,
      "step": 17715
    },
    {
      "entropy": 5.498606967926025,
      "epoch": 1.7586343787217151,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046952180636398793,
      "loss": 5.181,
      "mean_token_accuracy": 0.16199980527162552,
      "num_tokens": 38830890.0,
      "step": 17720
    },
    {
      "entropy": 5.523797559738159,
      "epoch": 1.7591306073838826,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004695039974402866,
      "loss": 5.2544,
      "mean_token_accuracy": 0.16240211427211762,
      "num_tokens": 38842648.0,
      "step": 17725
    },
    {
      "entropy": 5.5692214012146,
      "epoch": 1.75962683604605,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046948618369330533,
      "loss": 5.1986,
      "mean_token_accuracy": 0.16712820529937744,
      "num_tokens": 38852740.0,
      "step": 17730
    },
    {
      "entropy": 5.524205732345581,
      "epoch": 1.7601230647082176,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046946836512348566,
      "loss": 5.338,
      "mean_token_accuracy": 0.14946184903383256,
      "num_tokens": 38863371.0,
      "step": 17735
    },
    {
      "entropy": 5.578400182723999,
      "epoch": 1.760619293370385,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046945054173126965,
      "loss": 5.2997,
      "mean_token_accuracy": 0.1581171676516533,
      "num_tokens": 38873893.0,
      "step": 17740
    },
    {
      "entropy": 5.587970638275147,
      "epoch": 1.7611155220325525,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046943271351709923,
      "loss": 5.3651,
      "mean_token_accuracy": 0.15448251962661744,
      "num_tokens": 38884778.0,
      "step": 17745
    },
    {
      "entropy": 5.524837684631348,
      "epoch": 1.76161175069472,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004694148804814164,
      "loss": 5.3023,
      "mean_token_accuracy": 0.1565079554915428,
      "num_tokens": 38896011.0,
      "step": 17750
    },
    {
      "entropy": 5.58945484161377,
      "epoch": 1.7621079793568877,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004693970426246635,
      "loss": 5.3404,
      "mean_token_accuracy": 0.1499084785580635,
      "num_tokens": 38907686.0,
      "step": 17755
    },
    {
      "entropy": 5.511695623397827,
      "epoch": 1.7626042080190552,
      "grad_norm": 1.65625,
      "learning_rate": 0.00046937919994728287,
      "loss": 5.2995,
      "mean_token_accuracy": 0.15649536103010178,
      "num_tokens": 38919180.0,
      "step": 17760
    },
    {
      "entropy": 5.55660719871521,
      "epoch": 1.7631004366812228,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004693613524497168,
      "loss": 5.2615,
      "mean_token_accuracy": 0.15980495065450667,
      "num_tokens": 38929600.0,
      "step": 17765
    },
    {
      "entropy": 5.599948120117188,
      "epoch": 1.7635966653433903,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046934350013240814,
      "loss": 5.2173,
      "mean_token_accuracy": 0.1643930569291115,
      "num_tokens": 38940404.0,
      "step": 17770
    },
    {
      "entropy": 5.54900803565979,
      "epoch": 1.7640928940055578,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004693256429957994,
      "loss": 5.2655,
      "mean_token_accuracy": 0.1627445250749588,
      "num_tokens": 38951929.0,
      "step": 17775
    },
    {
      "entropy": 5.5069153785705565,
      "epoch": 1.7645891226677253,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004693077810403335,
      "loss": 5.2619,
      "mean_token_accuracy": 0.1603559970855713,
      "num_tokens": 38961954.0,
      "step": 17780
    },
    {
      "entropy": 5.527059745788574,
      "epoch": 1.7650853513298927,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046928991426645325,
      "loss": 5.2301,
      "mean_token_accuracy": 0.16304759681224823,
      "num_tokens": 38970974.0,
      "step": 17785
    },
    {
      "entropy": 5.57023983001709,
      "epoch": 1.7655815799920602,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046927204267460177,
      "loss": 5.3242,
      "mean_token_accuracy": 0.15539026409387588,
      "num_tokens": 38982481.0,
      "step": 17790
    },
    {
      "entropy": 5.609460020065308,
      "epoch": 1.766077808654228,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046925416626522235,
      "loss": 5.2894,
      "mean_token_accuracy": 0.1610730230808258,
      "num_tokens": 38993279.0,
      "step": 17795
    },
    {
      "entropy": 5.516831350326538,
      "epoch": 1.7665740373163954,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004692362850387581,
      "loss": 5.2596,
      "mean_token_accuracy": 0.16179780215024947,
      "num_tokens": 39003748.0,
      "step": 17800
    },
    {
      "entropy": 5.493751859664917,
      "epoch": 1.7670702659785629,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00046921839899565263,
      "loss": 5.2757,
      "mean_token_accuracy": 0.15366032123565673,
      "num_tokens": 39014902.0,
      "step": 17805
    },
    {
      "entropy": 5.569702577590943,
      "epoch": 1.7675664946407306,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004692005081363493,
      "loss": 5.3485,
      "mean_token_accuracy": 0.15385346710681916,
      "num_tokens": 39026484.0,
      "step": 17810
    },
    {
      "entropy": 5.572071313858032,
      "epoch": 1.768062723302898,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046918261246129176,
      "loss": 5.3231,
      "mean_token_accuracy": 0.1550864115357399,
      "num_tokens": 39036268.0,
      "step": 17815
    },
    {
      "entropy": 5.6027590274810795,
      "epoch": 1.7685589519650655,
      "grad_norm": 1.359375,
      "learning_rate": 0.000469164711970924,
      "loss": 5.3852,
      "mean_token_accuracy": 0.15538707077503205,
      "num_tokens": 39049092.0,
      "step": 17820
    },
    {
      "entropy": 5.604225301742554,
      "epoch": 1.769055180627233,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004691468066656896,
      "loss": 5.3117,
      "mean_token_accuracy": 0.15177865773439408,
      "num_tokens": 39059255.0,
      "step": 17825
    },
    {
      "entropy": 5.5226318359375,
      "epoch": 1.7695514092894005,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046912889654603286,
      "loss": 5.2319,
      "mean_token_accuracy": 0.16025121808052062,
      "num_tokens": 39069913.0,
      "step": 17830
    },
    {
      "entropy": 5.561707019805908,
      "epoch": 1.770047637951568,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004691109816123977,
      "loss": 5.258,
      "mean_token_accuracy": 0.162455515563488,
      "num_tokens": 39080666.0,
      "step": 17835
    },
    {
      "entropy": 5.584376001358033,
      "epoch": 1.7705438666137356,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004690930618652286,
      "loss": 5.3156,
      "mean_token_accuracy": 0.15546325147151946,
      "num_tokens": 39091562.0,
      "step": 17840
    },
    {
      "entropy": 5.613411378860474,
      "epoch": 1.7710400952759031,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046907513730496975,
      "loss": 5.3883,
      "mean_token_accuracy": 0.14589001312851907,
      "num_tokens": 39102567.0,
      "step": 17845
    },
    {
      "entropy": 5.66389832496643,
      "epoch": 1.7715363239380708,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004690572079320657,
      "loss": 5.2794,
      "mean_token_accuracy": 0.15894893258810044,
      "num_tokens": 39113260.0,
      "step": 17850
    },
    {
      "entropy": 5.547697591781616,
      "epoch": 1.7720325526002383,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00046903927374696106,
      "loss": 5.3274,
      "mean_token_accuracy": 0.15917097330093383,
      "num_tokens": 39125790.0,
      "step": 17855
    },
    {
      "entropy": 5.557540321350098,
      "epoch": 1.7725287812624058,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004690213347501005,
      "loss": 5.3732,
      "mean_token_accuracy": 0.15195812582969664,
      "num_tokens": 39137220.0,
      "step": 17860
    },
    {
      "entropy": 5.614561986923218,
      "epoch": 1.7730250099245732,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004690033909419289,
      "loss": 5.2342,
      "mean_token_accuracy": 0.16195437014102937,
      "num_tokens": 39146362.0,
      "step": 17865
    },
    {
      "entropy": 5.613611793518066,
      "epoch": 1.7735212385867407,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00046898544232289126,
      "loss": 5.3211,
      "mean_token_accuracy": 0.15325204730033876,
      "num_tokens": 39156822.0,
      "step": 17870
    },
    {
      "entropy": 5.587352705001831,
      "epoch": 1.7740174672489082,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046896748889343267,
      "loss": 5.2857,
      "mean_token_accuracy": 0.1554567724466324,
      "num_tokens": 39166693.0,
      "step": 17875
    },
    {
      "entropy": 5.511619520187378,
      "epoch": 1.7745136959110757,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046894953065399836,
      "loss": 5.2741,
      "mean_token_accuracy": 0.15916115790605545,
      "num_tokens": 39176889.0,
      "step": 17880
    },
    {
      "entropy": 5.558032035827637,
      "epoch": 1.7750099245732434,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046893156760503357,
      "loss": 5.2597,
      "mean_token_accuracy": 0.16651064157485962,
      "num_tokens": 39188077.0,
      "step": 17885
    },
    {
      "entropy": 5.535122108459473,
      "epoch": 1.7755061532354108,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004689135997469838,
      "loss": 5.25,
      "mean_token_accuracy": 0.15548657476902009,
      "num_tokens": 39198182.0,
      "step": 17890
    },
    {
      "entropy": 5.6216474056243895,
      "epoch": 1.7760023818975785,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004688956270802945,
      "loss": 5.2392,
      "mean_token_accuracy": 0.16356978416442872,
      "num_tokens": 39208801.0,
      "step": 17895
    },
    {
      "entropy": 5.513303613662719,
      "epoch": 1.776498610559746,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004688776496054117,
      "loss": 5.1193,
      "mean_token_accuracy": 0.16874260902404786,
      "num_tokens": 39220658.0,
      "step": 17900
    },
    {
      "entropy": 5.532590341567993,
      "epoch": 1.7769948392219135,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046885966732278077,
      "loss": 5.3616,
      "mean_token_accuracy": 0.15321645140647888,
      "num_tokens": 39231987.0,
      "step": 17905
    },
    {
      "entropy": 5.581524658203125,
      "epoch": 1.777491067884081,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046884168023284787,
      "loss": 5.2329,
      "mean_token_accuracy": 0.16406987905502318,
      "num_tokens": 39242797.0,
      "step": 17910
    },
    {
      "entropy": 5.561891317367554,
      "epoch": 1.7779872965462484,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046882368833605894,
      "loss": 5.2913,
      "mean_token_accuracy": 0.16215261965990066,
      "num_tokens": 39253621.0,
      "step": 17915
    },
    {
      "entropy": 5.51643476486206,
      "epoch": 1.778483525208416,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046880569163286023,
      "loss": 5.3201,
      "mean_token_accuracy": 0.156054262816906,
      "num_tokens": 39264126.0,
      "step": 17920
    },
    {
      "entropy": 5.542575931549072,
      "epoch": 1.7789797538705836,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046878769012369807,
      "loss": 5.2514,
      "mean_token_accuracy": 0.15714481323957444,
      "num_tokens": 39275118.0,
      "step": 17925
    },
    {
      "entropy": 5.530385971069336,
      "epoch": 1.779475982532751,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046876968380901864,
      "loss": 5.2621,
      "mean_token_accuracy": 0.1567048728466034,
      "num_tokens": 39287237.0,
      "step": 17930
    },
    {
      "entropy": 5.592088985443115,
      "epoch": 1.7799722111949188,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046875167268926865,
      "loss": 5.309,
      "mean_token_accuracy": 0.156399667263031,
      "num_tokens": 39296995.0,
      "step": 17935
    },
    {
      "entropy": 5.548354053497315,
      "epoch": 1.7804684398570862,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004687336567648947,
      "loss": 5.2032,
      "mean_token_accuracy": 0.15589718967676164,
      "num_tokens": 39305980.0,
      "step": 17940
    },
    {
      "entropy": 5.599948930740356,
      "epoch": 1.7809646685192537,
      "grad_norm": 1.375,
      "learning_rate": 0.00046871563603634347,
      "loss": 5.3461,
      "mean_token_accuracy": 0.14938850551843644,
      "num_tokens": 39316494.0,
      "step": 17945
    },
    {
      "entropy": 5.560718536376953,
      "epoch": 1.7814608971814212,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004686976105040619,
      "loss": 5.3208,
      "mean_token_accuracy": 0.15578545480966569,
      "num_tokens": 39327699.0,
      "step": 17950
    },
    {
      "entropy": 5.569830226898193,
      "epoch": 1.7819571258435887,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004686795801684969,
      "loss": 5.3234,
      "mean_token_accuracy": 0.15858250856399536,
      "num_tokens": 39338749.0,
      "step": 17955
    },
    {
      "entropy": 5.555962181091308,
      "epoch": 1.7824533545057561,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004686615450300958,
      "loss": 5.3247,
      "mean_token_accuracy": 0.1622561365365982,
      "num_tokens": 39349747.0,
      "step": 17960
    },
    {
      "entropy": 5.532038974761963,
      "epoch": 1.7829495831679236,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004686435050893056,
      "loss": 5.1666,
      "mean_token_accuracy": 0.16089108884334563,
      "num_tokens": 39360068.0,
      "step": 17965
    },
    {
      "entropy": 5.471302270889282,
      "epoch": 1.7834458118300913,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046862546034657376,
      "loss": 5.2053,
      "mean_token_accuracy": 0.1651167467236519,
      "num_tokens": 39369651.0,
      "step": 17970
    },
    {
      "entropy": 5.548853826522827,
      "epoch": 1.7839420404922588,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004686074108023477,
      "loss": 5.3928,
      "mean_token_accuracy": 0.15544423162937165,
      "num_tokens": 39381354.0,
      "step": 17975
    },
    {
      "entropy": 5.585770130157471,
      "epoch": 1.7844382691544265,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046858935645707506,
      "loss": 5.2495,
      "mean_token_accuracy": 0.15885641276836396,
      "num_tokens": 39391793.0,
      "step": 17980
    },
    {
      "entropy": 5.568853902816772,
      "epoch": 1.784934497816594,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004685712973112035,
      "loss": 5.2654,
      "mean_token_accuracy": 0.16706827729940416,
      "num_tokens": 39402194.0,
      "step": 17985
    },
    {
      "entropy": 5.552650499343872,
      "epoch": 1.7854307264787614,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004685532333651809,
      "loss": 5.2815,
      "mean_token_accuracy": 0.15569668412208557,
      "num_tokens": 39413193.0,
      "step": 17990
    },
    {
      "entropy": 5.509668970108033,
      "epoch": 1.785926955140929,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046853516461945523,
      "loss": 5.2488,
      "mean_token_accuracy": 0.16008808463811874,
      "num_tokens": 39423072.0,
      "step": 17995
    },
    {
      "entropy": 5.5830771923065186,
      "epoch": 1.7864231838030964,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046851709107447446,
      "loss": 5.3677,
      "mean_token_accuracy": 0.15275848507881165,
      "num_tokens": 39435088.0,
      "step": 18000
    },
    {
      "epoch": 1.7864231838030964,
      "eval_entropy": 5.444059413072473,
      "eval_loss": 5.411536693572998,
      "eval_mean_token_accuracy": 0.15608291171294658,
      "eval_num_tokens": 39435088.0,
      "eval_runtime": 27.1297,
      "eval_samples_per_second": 1241.038,
      "eval_steps_per_second": 155.144,
      "step": 18000
    },
    {
      "entropy": 5.541477632522583,
      "epoch": 1.7869194124652639,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004684990127306868,
      "loss": 5.2317,
      "mean_token_accuracy": 0.1579779416322708,
      "num_tokens": 39446129.0,
      "step": 18005
    },
    {
      "entropy": 5.529263687133789,
      "epoch": 1.7874156411274316,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046848092958854065,
      "loss": 5.1997,
      "mean_token_accuracy": 0.17348941415548325,
      "num_tokens": 39458127.0,
      "step": 18010
    },
    {
      "entropy": 5.565672111511231,
      "epoch": 1.787911869789599,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004684628416484843,
      "loss": 5.2514,
      "mean_token_accuracy": 0.163703253865242,
      "num_tokens": 39468204.0,
      "step": 18015
    },
    {
      "entropy": 5.578101587295532,
      "epoch": 1.7884080984517667,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004684447489109663,
      "loss": 5.3467,
      "mean_token_accuracy": 0.15331438034772873,
      "num_tokens": 39479114.0,
      "step": 18020
    },
    {
      "entropy": 5.576656293869019,
      "epoch": 1.7889043271139342,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046842665137643546,
      "loss": 5.279,
      "mean_token_accuracy": 0.1569603145122528,
      "num_tokens": 39489691.0,
      "step": 18025
    },
    {
      "entropy": 5.546512413024902,
      "epoch": 1.7894005557761017,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046840854904534046,
      "loss": 5.3956,
      "mean_token_accuracy": 0.1463721960783005,
      "num_tokens": 39501879.0,
      "step": 18030
    },
    {
      "entropy": 5.592569637298584,
      "epoch": 1.7898967844382692,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004683904419181302,
      "loss": 5.3141,
      "mean_token_accuracy": 0.15871454924345016,
      "num_tokens": 39513413.0,
      "step": 18035
    },
    {
      "entropy": 5.583690643310547,
      "epoch": 1.7903930131004366,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046837232999525366,
      "loss": 5.2516,
      "mean_token_accuracy": 0.15460312217473984,
      "num_tokens": 39522719.0,
      "step": 18040
    },
    {
      "entropy": 5.573954439163208,
      "epoch": 1.7908892417626041,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046835421327716,
      "loss": 5.3051,
      "mean_token_accuracy": 0.15350178629159927,
      "num_tokens": 39532463.0,
      "step": 18045
    },
    {
      "entropy": 5.578861951828003,
      "epoch": 1.7913854704247716,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004683360917642986,
      "loss": 5.2671,
      "mean_token_accuracy": 0.15608745962381362,
      "num_tokens": 39543005.0,
      "step": 18050
    },
    {
      "entropy": 5.599271297454834,
      "epoch": 1.7918816990869393,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046831796545711865,
      "loss": 5.3502,
      "mean_token_accuracy": 0.15490724965929986,
      "num_tokens": 39553468.0,
      "step": 18055
    },
    {
      "entropy": 5.648383617401123,
      "epoch": 1.7923779277491068,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046829983435606973,
      "loss": 5.3981,
      "mean_token_accuracy": 0.15143178552389144,
      "num_tokens": 39563539.0,
      "step": 18060
    },
    {
      "entropy": 5.603395891189575,
      "epoch": 1.7928741564112745,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004682816984616014,
      "loss": 5.3199,
      "mean_token_accuracy": 0.16159234046936036,
      "num_tokens": 39574835.0,
      "step": 18065
    },
    {
      "entropy": 5.572841024398803,
      "epoch": 1.793370385073442,
      "grad_norm": 1.796875,
      "learning_rate": 0.00046826355777416346,
      "loss": 5.2955,
      "mean_token_accuracy": 0.15964069813489914,
      "num_tokens": 39586021.0,
      "step": 18070
    },
    {
      "entropy": 5.574321269989014,
      "epoch": 1.7938666137356094,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046824541229420583,
      "loss": 5.2342,
      "mean_token_accuracy": 0.15801336169242858,
      "num_tokens": 39596333.0,
      "step": 18075
    },
    {
      "entropy": 5.527180290222168,
      "epoch": 1.7943628423977769,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00046822726202217825,
      "loss": 5.2772,
      "mean_token_accuracy": 0.15887489467859267,
      "num_tokens": 39607553.0,
      "step": 18080
    },
    {
      "entropy": 5.537750959396362,
      "epoch": 1.7948590710599444,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000468209106958531,
      "loss": 5.3292,
      "mean_token_accuracy": 0.155597323179245,
      "num_tokens": 39619700.0,
      "step": 18085
    },
    {
      "entropy": 5.572501087188721,
      "epoch": 1.7953552997221118,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004681909471037142,
      "loss": 5.2548,
      "mean_token_accuracy": 0.15917935967445374,
      "num_tokens": 39630623.0,
      "step": 18090
    },
    {
      "entropy": 5.54023847579956,
      "epoch": 1.7958515283842795,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046817278245817815,
      "loss": 5.2613,
      "mean_token_accuracy": 0.1598769873380661,
      "num_tokens": 39641089.0,
      "step": 18095
    },
    {
      "entropy": 5.563070440292359,
      "epoch": 1.796347757046447,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004681546130223735,
      "loss": 5.302,
      "mean_token_accuracy": 0.15580418407917024,
      "num_tokens": 39652381.0,
      "step": 18100
    },
    {
      "entropy": 5.497836065292359,
      "epoch": 1.7968439857086145,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004681364387967505,
      "loss": 5.215,
      "mean_token_accuracy": 0.1569169282913208,
      "num_tokens": 39662536.0,
      "step": 18105
    },
    {
      "entropy": 5.5354756832122805,
      "epoch": 1.7973402143707822,
      "grad_norm": 1.625,
      "learning_rate": 0.0004681182597817601,
      "loss": 5.2789,
      "mean_token_accuracy": 0.16003651469945906,
      "num_tokens": 39673140.0,
      "step": 18110
    },
    {
      "entropy": 5.467613697052002,
      "epoch": 1.7978364430329496,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004681000759778529,
      "loss": 5.2644,
      "mean_token_accuracy": 0.15487031787633895,
      "num_tokens": 39683843.0,
      "step": 18115
    },
    {
      "entropy": 5.530938625335693,
      "epoch": 1.7983326716951171,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004680818873854799,
      "loss": 5.2257,
      "mean_token_accuracy": 0.15817405432462692,
      "num_tokens": 39694178.0,
      "step": 18120
    },
    {
      "entropy": 5.470015525817871,
      "epoch": 1.7988289003572846,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004680636940050922,
      "loss": 5.2202,
      "mean_token_accuracy": 0.16557349413633346,
      "num_tokens": 39704818.0,
      "step": 18125
    },
    {
      "entropy": 5.511974096298218,
      "epoch": 1.799325129019452,
      "grad_norm": 1.21875,
      "learning_rate": 0.000468045495837141,
      "loss": 5.2463,
      "mean_token_accuracy": 0.1520305559039116,
      "num_tokens": 39715599.0,
      "step": 18130
    },
    {
      "entropy": 5.557569074630737,
      "epoch": 1.7998213576816195,
      "grad_norm": 1.375,
      "learning_rate": 0.00046802729288207735,
      "loss": 5.2528,
      "mean_token_accuracy": 0.1622065022587776,
      "num_tokens": 39725736.0,
      "step": 18135
    },
    {
      "entropy": 5.55124979019165,
      "epoch": 1.8003175863437872,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046800908514035285,
      "loss": 5.2899,
      "mean_token_accuracy": 0.15885340571403503,
      "num_tokens": 39735736.0,
      "step": 18140
    },
    {
      "entropy": 5.560364675521851,
      "epoch": 1.8008138150059547,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046799087261241884,
      "loss": 5.3356,
      "mean_token_accuracy": 0.1540846735239029,
      "num_tokens": 39747443.0,
      "step": 18145
    },
    {
      "entropy": 5.576381254196167,
      "epoch": 1.8013100436681224,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004679726552987271,
      "loss": 5.2781,
      "mean_token_accuracy": 0.1565117806196213,
      "num_tokens": 39758128.0,
      "step": 18150
    },
    {
      "entropy": 5.5323628902435305,
      "epoch": 1.80180627233029,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004679544331997294,
      "loss": 5.2111,
      "mean_token_accuracy": 0.1635427564382553,
      "num_tokens": 39769153.0,
      "step": 18155
    },
    {
      "entropy": 5.571730136871338,
      "epoch": 1.8023025009924574,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046793620631587743,
      "loss": 5.3013,
      "mean_token_accuracy": 0.15315420031547547,
      "num_tokens": 39779666.0,
      "step": 18160
    },
    {
      "entropy": 5.523604917526245,
      "epoch": 1.8027987296546248,
      "grad_norm": 1.25,
      "learning_rate": 0.00046791797464762334,
      "loss": 5.3325,
      "mean_token_accuracy": 0.1528678797185421,
      "num_tokens": 39790329.0,
      "step": 18165
    },
    {
      "entropy": 5.591419649124146,
      "epoch": 1.8032949583167923,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046789973819541915,
      "loss": 5.306,
      "mean_token_accuracy": 0.1549484461545944,
      "num_tokens": 39801996.0,
      "step": 18170
    },
    {
      "entropy": 5.607329082489014,
      "epoch": 1.8037911869789598,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004678814969597171,
      "loss": 5.3028,
      "mean_token_accuracy": 0.15222611278295517,
      "num_tokens": 39814064.0,
      "step": 18175
    },
    {
      "entropy": 5.5467527389526365,
      "epoch": 1.8042874156411273,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004678632509409696,
      "loss": 5.2324,
      "mean_token_accuracy": 0.16092788875102998,
      "num_tokens": 39825369.0,
      "step": 18180
    },
    {
      "entropy": 5.478538703918457,
      "epoch": 1.804783644303295,
      "grad_norm": 1.5625,
      "learning_rate": 0.000467845000139629,
      "loss": 5.3078,
      "mean_token_accuracy": 0.15824806690216064,
      "num_tokens": 39835218.0,
      "step": 18185
    },
    {
      "entropy": 5.61581072807312,
      "epoch": 1.8052798729654624,
      "grad_norm": 1.390625,
      "learning_rate": 0.000467826744556148,
      "loss": 5.331,
      "mean_token_accuracy": 0.15529274493455886,
      "num_tokens": 39844706.0,
      "step": 18190
    },
    {
      "entropy": 5.473901414871216,
      "epoch": 1.8057761016276301,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004678084841909792,
      "loss": 5.2592,
      "mean_token_accuracy": 0.1635770246386528,
      "num_tokens": 39856348.0,
      "step": 18195
    },
    {
      "entropy": 5.486716604232788,
      "epoch": 1.8062723302897976,
      "grad_norm": 1.25,
      "learning_rate": 0.00046779021904457555,
      "loss": 5.3105,
      "mean_token_accuracy": 0.1604696422815323,
      "num_tokens": 39866850.0,
      "step": 18200
    },
    {
      "entropy": 5.616256999969482,
      "epoch": 1.806768558951965,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004677719491173898,
      "loss": 5.3431,
      "mean_token_accuracy": 0.15544990003108977,
      "num_tokens": 39877647.0,
      "step": 18205
    },
    {
      "entropy": 5.5884534358978275,
      "epoch": 1.8072647876141326,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046775367440987506,
      "loss": 5.3788,
      "mean_token_accuracy": 0.1568421721458435,
      "num_tokens": 39889172.0,
      "step": 18210
    },
    {
      "entropy": 5.578827142715454,
      "epoch": 1.8077610162763,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004677353949224846,
      "loss": 5.2885,
      "mean_token_accuracy": 0.1596810981631279,
      "num_tokens": 39900445.0,
      "step": 18215
    },
    {
      "entropy": 5.556734418869018,
      "epoch": 1.8082572449384675,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046771711065567167,
      "loss": 5.3031,
      "mean_token_accuracy": 0.1551668033003807,
      "num_tokens": 39912825.0,
      "step": 18220
    },
    {
      "entropy": 5.493354177474975,
      "epoch": 1.8087534736006352,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004676988216098896,
      "loss": 5.2416,
      "mean_token_accuracy": 0.16273412853479385,
      "num_tokens": 39923370.0,
      "step": 18225
    },
    {
      "entropy": 5.642121887207031,
      "epoch": 1.8092497022628027,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004676805277855921,
      "loss": 5.3085,
      "mean_token_accuracy": 0.1543406918644905,
      "num_tokens": 39934724.0,
      "step": 18230
    },
    {
      "entropy": 5.522439479827881,
      "epoch": 1.8097459309249704,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004676622291832326,
      "loss": 5.2396,
      "mean_token_accuracy": 0.15390536636114122,
      "num_tokens": 39947064.0,
      "step": 18235
    },
    {
      "entropy": 5.4929228782653805,
      "epoch": 1.8102421595871379,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046764392580326506,
      "loss": 5.2261,
      "mean_token_accuracy": 0.1592572048306465,
      "num_tokens": 39957649.0,
      "step": 18240
    },
    {
      "entropy": 5.582994842529297,
      "epoch": 1.8107383882493053,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046762561764614316,
      "loss": 5.3056,
      "mean_token_accuracy": 0.15492322593927382,
      "num_tokens": 39968086.0,
      "step": 18245
    },
    {
      "entropy": 5.5377027034759525,
      "epoch": 1.8112346169114728,
      "grad_norm": 1.796875,
      "learning_rate": 0.000467607304712321,
      "loss": 5.2683,
      "mean_token_accuracy": 0.15663823932409288,
      "num_tokens": 39978794.0,
      "step": 18250
    },
    {
      "entropy": 5.54730920791626,
      "epoch": 1.8117308455736403,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046758898700225283,
      "loss": 5.2767,
      "mean_token_accuracy": 0.15789321810007095,
      "num_tokens": 39990383.0,
      "step": 18255
    },
    {
      "entropy": 5.577605819702148,
      "epoch": 1.8122270742358078,
      "grad_norm": 1.25,
      "learning_rate": 0.0004675706645163927,
      "loss": 5.3472,
      "mean_token_accuracy": 0.1573940210044384,
      "num_tokens": 40001900.0,
      "step": 18260
    },
    {
      "entropy": 5.611772775650024,
      "epoch": 1.8127233028979752,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000467552337255195,
      "loss": 5.2633,
      "mean_token_accuracy": 0.1575230985879898,
      "num_tokens": 40012482.0,
      "step": 18265
    },
    {
      "entropy": 5.513954782485962,
      "epoch": 1.813219531560143,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004675340052191144,
      "loss": 5.2712,
      "mean_token_accuracy": 0.15835898220539094,
      "num_tokens": 40023264.0,
      "step": 18270
    },
    {
      "entropy": 5.554391813278198,
      "epoch": 1.8137157602223104,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004675156684086052,
      "loss": 5.349,
      "mean_token_accuracy": 0.15444828495383262,
      "num_tokens": 40034754.0,
      "step": 18275
    },
    {
      "entropy": 5.53561806678772,
      "epoch": 1.814211988884478,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004674973268241224,
      "loss": 5.2429,
      "mean_token_accuracy": 0.16565490067005156,
      "num_tokens": 40045781.0,
      "step": 18280
    },
    {
      "entropy": 5.600417613983154,
      "epoch": 1.8147082175466456,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004674789804661205,
      "loss": 5.316,
      "mean_token_accuracy": 0.1552600458264351,
      "num_tokens": 40055700.0,
      "step": 18285
    },
    {
      "entropy": 5.606537342071533,
      "epoch": 1.815204446208813,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046746062933505477,
      "loss": 5.3645,
      "mean_token_accuracy": 0.15655219107866286,
      "num_tokens": 40066702.0,
      "step": 18290
    },
    {
      "entropy": 5.5764683246612545,
      "epoch": 1.8157006748709805,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004674422734313801,
      "loss": 5.3518,
      "mean_token_accuracy": 0.15960957407951354,
      "num_tokens": 40077856.0,
      "step": 18295
    },
    {
      "entropy": 5.476405143737793,
      "epoch": 1.816196903533148,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046742391275555176,
      "loss": 5.2454,
      "mean_token_accuracy": 0.1615538328886032,
      "num_tokens": 40089205.0,
      "step": 18300
    },
    {
      "entropy": 5.584054803848266,
      "epoch": 1.8166931321953155,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000467405547308025,
      "loss": 5.2817,
      "mean_token_accuracy": 0.15235146284103393,
      "num_tokens": 40100028.0,
      "step": 18305
    },
    {
      "entropy": 5.6102558135986325,
      "epoch": 1.8171893608574832,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004673871770892553,
      "loss": 5.2997,
      "mean_token_accuracy": 0.15416564792394638,
      "num_tokens": 40109635.0,
      "step": 18310
    },
    {
      "entropy": 5.554948520660401,
      "epoch": 1.8176855895196506,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046736880209969814,
      "loss": 5.2803,
      "mean_token_accuracy": 0.15455955266952515,
      "num_tokens": 40120539.0,
      "step": 18315
    },
    {
      "entropy": 5.561210250854492,
      "epoch": 1.8181818181818183,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046735042233980923,
      "loss": 5.3339,
      "mean_token_accuracy": 0.14944813549518585,
      "num_tokens": 40131738.0,
      "step": 18320
    },
    {
      "entropy": 5.677899312973023,
      "epoch": 1.8186780468439858,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046733203781004426,
      "loss": 5.4029,
      "mean_token_accuracy": 0.1549864962697029,
      "num_tokens": 40142800.0,
      "step": 18325
    },
    {
      "entropy": 5.638781118392944,
      "epoch": 1.8191742755061533,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046731364851085923,
      "loss": 5.3505,
      "mean_token_accuracy": 0.1532520778477192,
      "num_tokens": 40153992.0,
      "step": 18330
    },
    {
      "entropy": 5.568453168869018,
      "epoch": 1.8196705041683208,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004672952544427101,
      "loss": 5.275,
      "mean_token_accuracy": 0.1581330418586731,
      "num_tokens": 40165213.0,
      "step": 18335
    },
    {
      "entropy": 5.552844095230102,
      "epoch": 1.8201667328304882,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046727685560605305,
      "loss": 5.3426,
      "mean_token_accuracy": 0.1538911134004593,
      "num_tokens": 40176763.0,
      "step": 18340
    },
    {
      "entropy": 5.553252458572388,
      "epoch": 1.8206629614926557,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004672584520013443,
      "loss": 5.2407,
      "mean_token_accuracy": 0.15332867652177812,
      "num_tokens": 40187958.0,
      "step": 18345
    },
    {
      "entropy": 5.540956306457519,
      "epoch": 1.8211591901548232,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004672400436290402,
      "loss": 5.2552,
      "mean_token_accuracy": 0.15960672348737717,
      "num_tokens": 40197881.0,
      "step": 18350
    },
    {
      "entropy": 5.582522344589234,
      "epoch": 1.821655418816991,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046722163048959725,
      "loss": 5.2346,
      "mean_token_accuracy": 0.1634988933801651,
      "num_tokens": 40209322.0,
      "step": 18355
    },
    {
      "entropy": 5.550030708312988,
      "epoch": 1.8221516474791584,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004672032125834721,
      "loss": 5.2944,
      "mean_token_accuracy": 0.158334568887949,
      "num_tokens": 40222186.0,
      "step": 18360
    },
    {
      "entropy": 5.569630193710327,
      "epoch": 1.822647876141326,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004671847899111214,
      "loss": 5.3192,
      "mean_token_accuracy": 0.15641611069440842,
      "num_tokens": 40232136.0,
      "step": 18365
    },
    {
      "entropy": 5.637604808807373,
      "epoch": 1.8231441048034935,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046716636247300206,
      "loss": 5.3658,
      "mean_token_accuracy": 0.15404367744922637,
      "num_tokens": 40244468.0,
      "step": 18370
    },
    {
      "entropy": 5.583489418029785,
      "epoch": 1.823640333465661,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046714793026957096,
      "loss": 5.2902,
      "mean_token_accuracy": 0.15056217610836028,
      "num_tokens": 40255441.0,
      "step": 18375
    },
    {
      "entropy": 5.494051742553711,
      "epoch": 1.8241365621278285,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004671294933012853,
      "loss": 5.2414,
      "mean_token_accuracy": 0.1537858083844185,
      "num_tokens": 40266109.0,
      "step": 18380
    },
    {
      "entropy": 5.487967300415039,
      "epoch": 1.824632790789996,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046711105156860223,
      "loss": 5.2366,
      "mean_token_accuracy": 0.16049710065126419,
      "num_tokens": 40277635.0,
      "step": 18385
    },
    {
      "entropy": 5.560118818283081,
      "epoch": 1.8251290194521634,
      "grad_norm": 2.140625,
      "learning_rate": 0.00046709260507197896,
      "loss": 5.3828,
      "mean_token_accuracy": 0.15213114395737648,
      "num_tokens": 40288668.0,
      "step": 18390
    },
    {
      "entropy": 5.564276695251465,
      "epoch": 1.8256252481143311,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046707415381187306,
      "loss": 5.2455,
      "mean_token_accuracy": 0.16167444586753846,
      "num_tokens": 40300262.0,
      "step": 18395
    },
    {
      "entropy": 5.551926469802856,
      "epoch": 1.8261214767764986,
      "grad_norm": 1.28125,
      "learning_rate": 0.000467055697788742,
      "loss": 5.3192,
      "mean_token_accuracy": 0.14929328858852386,
      "num_tokens": 40310931.0,
      "step": 18400
    },
    {
      "entropy": 5.5980634689331055,
      "epoch": 1.826617705438666,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00046703723700304345,
      "loss": 5.2735,
      "mean_token_accuracy": 0.15611034631729126,
      "num_tokens": 40321623.0,
      "step": 18405
    },
    {
      "entropy": 5.4955418586730955,
      "epoch": 1.8271139341008338,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004670187714552353,
      "loss": 5.2135,
      "mean_token_accuracy": 0.16074957251548766,
      "num_tokens": 40330951.0,
      "step": 18410
    },
    {
      "entropy": 5.5938342094421385,
      "epoch": 1.8276101627630013,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046700030114577533,
      "loss": 5.3158,
      "mean_token_accuracy": 0.16049137264490126,
      "num_tokens": 40342971.0,
      "step": 18415
    },
    {
      "entropy": 5.553907203674316,
      "epoch": 1.8281063914251687,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004669818260751217,
      "loss": 5.2931,
      "mean_token_accuracy": 0.15734702050685884,
      "num_tokens": 40353837.0,
      "step": 18420
    },
    {
      "entropy": 5.569289636611939,
      "epoch": 1.8286026200873362,
      "grad_norm": 1.25,
      "learning_rate": 0.0004669633462437324,
      "loss": 5.3681,
      "mean_token_accuracy": 0.15694326162338257,
      "num_tokens": 40367389.0,
      "step": 18425
    },
    {
      "entropy": 5.539648342132568,
      "epoch": 1.8290988487495037,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046694486165206573,
      "loss": 5.1641,
      "mean_token_accuracy": 0.16317395865917206,
      "num_tokens": 40376793.0,
      "step": 18430
    },
    {
      "entropy": 5.537055730819702,
      "epoch": 1.8295950774116712,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004669263723005801,
      "loss": 5.2696,
      "mean_token_accuracy": 0.1610516220331192,
      "num_tokens": 40388025.0,
      "step": 18435
    },
    {
      "entropy": 5.556886100769043,
      "epoch": 1.8300913060738389,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046690787818973406,
      "loss": 5.354,
      "mean_token_accuracy": 0.15346518605947496,
      "num_tokens": 40399731.0,
      "step": 18440
    },
    {
      "entropy": 5.588856792449951,
      "epoch": 1.8305875347360063,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046688937931998613,
      "loss": 5.3919,
      "mean_token_accuracy": 0.1574587792158127,
      "num_tokens": 40411078.0,
      "step": 18445
    },
    {
      "entropy": 5.621528816223145,
      "epoch": 1.831083763398174,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004668708756917951,
      "loss": 5.347,
      "mean_token_accuracy": 0.15860870033502578,
      "num_tokens": 40422806.0,
      "step": 18450
    },
    {
      "entropy": 5.715636348724365,
      "epoch": 1.8315799920603415,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046685236730561975,
      "loss": 5.4088,
      "mean_token_accuracy": 0.14916374757885933,
      "num_tokens": 40434041.0,
      "step": 18455
    },
    {
      "entropy": 5.574692726135254,
      "epoch": 1.832076220722509,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046683385416191914,
      "loss": 5.2506,
      "mean_token_accuracy": 0.1624876633286476,
      "num_tokens": 40444658.0,
      "step": 18460
    },
    {
      "entropy": 5.49145359992981,
      "epoch": 1.8325724493846764,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046681533626115226,
      "loss": 5.2595,
      "mean_token_accuracy": 0.15868161618709564,
      "num_tokens": 40455745.0,
      "step": 18465
    },
    {
      "entropy": 5.584494304656983,
      "epoch": 1.833068678046844,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004667968136037784,
      "loss": 5.3664,
      "mean_token_accuracy": 0.15673957169055938,
      "num_tokens": 40468829.0,
      "step": 18470
    },
    {
      "entropy": 5.571448945999146,
      "epoch": 1.8335649067090114,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046677828619025687,
      "loss": 5.2621,
      "mean_token_accuracy": 0.15788576006889343,
      "num_tokens": 40479817.0,
      "step": 18475
    },
    {
      "entropy": 5.617511320114136,
      "epoch": 1.8340611353711789,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046675975402104706,
      "loss": 5.3529,
      "mean_token_accuracy": 0.15174816623330117,
      "num_tokens": 40490398.0,
      "step": 18480
    },
    {
      "entropy": 5.523996353149414,
      "epoch": 1.8345573640333466,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046674121709660855,
      "loss": 5.1876,
      "mean_token_accuracy": 0.16208656877279282,
      "num_tokens": 40500483.0,
      "step": 18485
    },
    {
      "entropy": 5.586969232559204,
      "epoch": 1.835053592695514,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000466722675417401,
      "loss": 5.3424,
      "mean_token_accuracy": 0.15095054656267165,
      "num_tokens": 40511494.0,
      "step": 18490
    },
    {
      "entropy": 5.5352373123168945,
      "epoch": 1.8355498213576817,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046670412898388425,
      "loss": 5.2781,
      "mean_token_accuracy": 0.16742484271526337,
      "num_tokens": 40523573.0,
      "step": 18495
    },
    {
      "entropy": 5.541914606094361,
      "epoch": 1.8360460500198492,
      "grad_norm": 1.96875,
      "learning_rate": 0.00046668557779651825,
      "loss": 5.32,
      "mean_token_accuracy": 0.15719710886478425,
      "num_tokens": 40534904.0,
      "step": 18500
    },
    {
      "entropy": 5.566541385650635,
      "epoch": 1.8365422786820167,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046666702185576295,
      "loss": 5.2953,
      "mean_token_accuracy": 0.15404289662837983,
      "num_tokens": 40545053.0,
      "step": 18505
    },
    {
      "entropy": 5.569765186309814,
      "epoch": 1.8370385073441842,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004666484611620785,
      "loss": 5.2241,
      "mean_token_accuracy": 0.15970046073198318,
      "num_tokens": 40555305.0,
      "step": 18510
    },
    {
      "entropy": 5.540113639831543,
      "epoch": 1.8375347360063516,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004666298957159251,
      "loss": 5.2501,
      "mean_token_accuracy": 0.1546129897236824,
      "num_tokens": 40565611.0,
      "step": 18515
    },
    {
      "entropy": 5.5804023265838625,
      "epoch": 1.8380309646685191,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046661132551776327,
      "loss": 5.3556,
      "mean_token_accuracy": 0.15527749806642532,
      "num_tokens": 40578321.0,
      "step": 18520
    },
    {
      "entropy": 5.588535308837891,
      "epoch": 1.8385271933306868,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004665927505680535,
      "loss": 5.336,
      "mean_token_accuracy": 0.1638243481516838,
      "num_tokens": 40588696.0,
      "step": 18525
    },
    {
      "entropy": 5.54349684715271,
      "epoch": 1.8390234219928543,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046657417086725625,
      "loss": 5.2842,
      "mean_token_accuracy": 0.16013921350240706,
      "num_tokens": 40599699.0,
      "step": 18530
    },
    {
      "entropy": 5.527594327926636,
      "epoch": 1.839519650655022,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004665555864158324,
      "loss": 5.2895,
      "mean_token_accuracy": 0.1608859896659851,
      "num_tokens": 40609706.0,
      "step": 18535
    },
    {
      "entropy": 5.529288721084595,
      "epoch": 1.8400158793171895,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004665369972142428,
      "loss": 5.3137,
      "mean_token_accuracy": 0.16015838235616683,
      "num_tokens": 40619431.0,
      "step": 18540
    },
    {
      "entropy": 5.6026191234588625,
      "epoch": 1.840512107979357,
      "grad_norm": 2.234375,
      "learning_rate": 0.00046651840326294845,
      "loss": 5.2883,
      "mean_token_accuracy": 0.1610789880156517,
      "num_tokens": 40629904.0,
      "step": 18545
    },
    {
      "entropy": 5.561390781402588,
      "epoch": 1.8410083366415244,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004664998045624104,
      "loss": 5.3353,
      "mean_token_accuracy": 0.1535017192363739,
      "num_tokens": 40641266.0,
      "step": 18550
    },
    {
      "entropy": 5.534455633163452,
      "epoch": 1.8415045653036919,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004664812011130897,
      "loss": 5.2798,
      "mean_token_accuracy": 0.15298035889863967,
      "num_tokens": 40652454.0,
      "step": 18555
    },
    {
      "entropy": 5.590737676620483,
      "epoch": 1.8420007939658594,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046646259291544787,
      "loss": 5.258,
      "mean_token_accuracy": 0.15774894431233405,
      "num_tokens": 40664311.0,
      "step": 18560
    },
    {
      "entropy": 5.525270748138428,
      "epoch": 1.8424970226280268,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004664439799699463,
      "loss": 5.1926,
      "mean_token_accuracy": 0.1618000015616417,
      "num_tokens": 40675241.0,
      "step": 18565
    },
    {
      "entropy": 5.5009400844573975,
      "epoch": 1.8429932512901945,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046642536227704665,
      "loss": 5.2722,
      "mean_token_accuracy": 0.15843528509140015,
      "num_tokens": 40685972.0,
      "step": 18570
    },
    {
      "entropy": 5.54360728263855,
      "epoch": 1.843489479952362,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046640673983721037,
      "loss": 5.2815,
      "mean_token_accuracy": 0.15944606214761733,
      "num_tokens": 40696967.0,
      "step": 18575
    },
    {
      "entropy": 5.560307550430298,
      "epoch": 1.8439857086145297,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046638811265089944,
      "loss": 5.2712,
      "mean_token_accuracy": 0.15551171600818633,
      "num_tokens": 40707710.0,
      "step": 18580
    },
    {
      "entropy": 5.590273666381836,
      "epoch": 1.8444819372766972,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004663694807185758,
      "loss": 5.4359,
      "mean_token_accuracy": 0.1540244646370411,
      "num_tokens": 40718586.0,
      "step": 18585
    },
    {
      "entropy": 5.64454197883606,
      "epoch": 1.8449781659388647,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00046635084404070135,
      "loss": 5.3099,
      "mean_token_accuracy": 0.15796590149402617,
      "num_tokens": 40729792.0,
      "step": 18590
    },
    {
      "entropy": 5.683891534805298,
      "epoch": 1.8454743946010321,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004663322026177382,
      "loss": 5.3534,
      "mean_token_accuracy": 0.15034137219190596,
      "num_tokens": 40740974.0,
      "step": 18595
    },
    {
      "entropy": 5.478787422180176,
      "epoch": 1.8459706232631996,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046631355645014875,
      "loss": 5.1357,
      "mean_token_accuracy": 0.1625142365694046,
      "num_tokens": 40750208.0,
      "step": 18600
    },
    {
      "entropy": 5.575337600708008,
      "epoch": 1.846466851925367,
      "grad_norm": 1.375,
      "learning_rate": 0.00046629490553839537,
      "loss": 5.2685,
      "mean_token_accuracy": 0.155897057056427,
      "num_tokens": 40762163.0,
      "step": 18605
    },
    {
      "entropy": 5.600553274154663,
      "epoch": 1.8469630805875348,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004662762498829405,
      "loss": 5.3594,
      "mean_token_accuracy": 0.1540456086397171,
      "num_tokens": 40773028.0,
      "step": 18610
    },
    {
      "entropy": 5.611712837219239,
      "epoch": 1.8474593092497023,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004662575894842469,
      "loss": 5.3315,
      "mean_token_accuracy": 0.148666849732399,
      "num_tokens": 40784939.0,
      "step": 18615
    },
    {
      "entropy": 5.5928404331207275,
      "epoch": 1.84795553791187,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046623892434277713,
      "loss": 5.3036,
      "mean_token_accuracy": 0.15820058286190034,
      "num_tokens": 40796170.0,
      "step": 18620
    },
    {
      "entropy": 5.570671272277832,
      "epoch": 1.8484517665740374,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004662202544589941,
      "loss": 5.2489,
      "mean_token_accuracy": 0.15948697477579116,
      "num_tokens": 40807184.0,
      "step": 18625
    },
    {
      "entropy": 5.498831510543823,
      "epoch": 1.848947995236205,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046620157983336087,
      "loss": 5.2286,
      "mean_token_accuracy": 0.15479253381490707,
      "num_tokens": 40818474.0,
      "step": 18630
    },
    {
      "entropy": 5.540080404281616,
      "epoch": 1.8494442238983724,
      "grad_norm": 2.625,
      "learning_rate": 0.0004661829004663404,
      "loss": 5.2834,
      "mean_token_accuracy": 0.15851295739412308,
      "num_tokens": 40828789.0,
      "step": 18635
    },
    {
      "entropy": 5.556736183166504,
      "epoch": 1.8499404525605398,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004661642163583961,
      "loss": 5.2614,
      "mean_token_accuracy": 0.16417559385299682,
      "num_tokens": 40840564.0,
      "step": 18640
    },
    {
      "entropy": 5.51951036453247,
      "epoch": 1.8504366812227073,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046614552750999093,
      "loss": 5.2264,
      "mean_token_accuracy": 0.1604135274887085,
      "num_tokens": 40852528.0,
      "step": 18645
    },
    {
      "entropy": 5.541305160522461,
      "epoch": 1.8509329098848748,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004661268339215887,
      "loss": 5.2773,
      "mean_token_accuracy": 0.16277666836977006,
      "num_tokens": 40863242.0,
      "step": 18650
    },
    {
      "entropy": 5.516217374801636,
      "epoch": 1.8514291385470425,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004661081355936528,
      "loss": 5.2286,
      "mean_token_accuracy": 0.16523500829935073,
      "num_tokens": 40873721.0,
      "step": 18655
    },
    {
      "entropy": 5.602329254150391,
      "epoch": 1.85192536720921,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000466089432526647,
      "loss": 5.3475,
      "mean_token_accuracy": 0.15494672879576682,
      "num_tokens": 40884921.0,
      "step": 18660
    },
    {
      "entropy": 5.578680801391601,
      "epoch": 1.8524215958713777,
      "grad_norm": 1.46875,
      "learning_rate": 0.00046607072472103493,
      "loss": 5.1979,
      "mean_token_accuracy": 0.15693599432706834,
      "num_tokens": 40896243.0,
      "step": 18665
    },
    {
      "entropy": 5.525571870803833,
      "epoch": 1.8529178245335451,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046605201217728067,
      "loss": 5.3268,
      "mean_token_accuracy": 0.15237598717212678,
      "num_tokens": 40907191.0,
      "step": 18670
    },
    {
      "entropy": 5.593805932998658,
      "epoch": 1.8534140531957126,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046603329489584817,
      "loss": 5.2921,
      "mean_token_accuracy": 0.16315285861492157,
      "num_tokens": 40918737.0,
      "step": 18675
    },
    {
      "entropy": 5.591906547546387,
      "epoch": 1.85391028185788,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004660145728772016,
      "loss": 5.3133,
      "mean_token_accuracy": 0.155313903093338,
      "num_tokens": 40930184.0,
      "step": 18680
    },
    {
      "entropy": 5.495571517944336,
      "epoch": 1.8544065105200476,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004659958461218052,
      "loss": 5.2021,
      "mean_token_accuracy": 0.16240794062614441,
      "num_tokens": 40941483.0,
      "step": 18685
    },
    {
      "entropy": 5.5276322841644285,
      "epoch": 1.854902739182215,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046597711463012344,
      "loss": 5.3104,
      "mean_token_accuracy": 0.1585666060447693,
      "num_tokens": 40952392.0,
      "step": 18690
    },
    {
      "entropy": 5.522467994689942,
      "epoch": 1.8553989678443827,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004659583784026207,
      "loss": 5.2254,
      "mean_token_accuracy": 0.16238734126091003,
      "num_tokens": 40962539.0,
      "step": 18695
    },
    {
      "entropy": 5.531670475006104,
      "epoch": 1.8558951965065502,
      "grad_norm": 1.25,
      "learning_rate": 0.00046593963743976163,
      "loss": 5.1577,
      "mean_token_accuracy": 0.1656871646642685,
      "num_tokens": 40972594.0,
      "step": 18700
    },
    {
      "entropy": 5.580002355575561,
      "epoch": 1.8563914251687177,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000465920891742011,
      "loss": 5.2654,
      "mean_token_accuracy": 0.1532726153731346,
      "num_tokens": 40983460.0,
      "step": 18705
    },
    {
      "entropy": 5.571139717102051,
      "epoch": 1.8568876538308854,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004659021413098337,
      "loss": 5.2777,
      "mean_token_accuracy": 0.1579229936003685,
      "num_tokens": 40993945.0,
      "step": 18710
    },
    {
      "entropy": 5.496938753128052,
      "epoch": 1.8573838824930529,
      "grad_norm": 2.609375,
      "learning_rate": 0.00046588338614369465,
      "loss": 5.2768,
      "mean_token_accuracy": 0.16216108426451684,
      "num_tokens": 41005781.0,
      "step": 18715
    },
    {
      "entropy": 5.6041333198547365,
      "epoch": 1.8578801111552203,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046586462624405897,
      "loss": 5.3853,
      "mean_token_accuracy": 0.15575614720582961,
      "num_tokens": 41017650.0,
      "step": 18720
    },
    {
      "entropy": 5.593478059768676,
      "epoch": 1.8583763398173878,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004658458616113918,
      "loss": 5.2427,
      "mean_token_accuracy": 0.15878477543592454,
      "num_tokens": 41029196.0,
      "step": 18725
    },
    {
      "entropy": 5.561244297027588,
      "epoch": 1.8588725684795553,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004658270922461585,
      "loss": 5.3113,
      "mean_token_accuracy": 0.16090276837348938,
      "num_tokens": 41040204.0,
      "step": 18730
    },
    {
      "entropy": 5.571606254577636,
      "epoch": 1.8593687971417228,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046580831814882446,
      "loss": 5.3016,
      "mean_token_accuracy": 0.15061459243297576,
      "num_tokens": 41050507.0,
      "step": 18735
    },
    {
      "entropy": 5.495503568649292,
      "epoch": 1.8598650258038905,
      "grad_norm": 1.71875,
      "learning_rate": 0.00046578953931985536,
      "loss": 5.2503,
      "mean_token_accuracy": 0.15445465296506883,
      "num_tokens": 41062001.0,
      "step": 18740
    },
    {
      "entropy": 5.58056001663208,
      "epoch": 1.860361254466058,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004657707557597167,
      "loss": 5.2869,
      "mean_token_accuracy": 0.15557732433080673,
      "num_tokens": 41073097.0,
      "step": 18745
    },
    {
      "entropy": 5.590755987167358,
      "epoch": 1.8608574831282256,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046575196746887447,
      "loss": 5.2965,
      "mean_token_accuracy": 0.15638842880725862,
      "num_tokens": 41083883.0,
      "step": 18750
    },
    {
      "entropy": 5.512462854385376,
      "epoch": 1.861353711790393,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046573317444779445,
      "loss": 5.2704,
      "mean_token_accuracy": 0.1648874506354332,
      "num_tokens": 41095506.0,
      "step": 18755
    },
    {
      "entropy": 5.61258978843689,
      "epoch": 1.8618499404525606,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004657143766969427,
      "loss": 5.2977,
      "mean_token_accuracy": 0.16258925646543504,
      "num_tokens": 41106568.0,
      "step": 18760
    },
    {
      "entropy": 5.562077569961548,
      "epoch": 1.862346169114728,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004656955742167853,
      "loss": 5.1981,
      "mean_token_accuracy": 0.15836713463068008,
      "num_tokens": 41117661.0,
      "step": 18765
    },
    {
      "entropy": 5.529874849319458,
      "epoch": 1.8628423977768955,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046567676700778865,
      "loss": 5.2912,
      "mean_token_accuracy": 0.15983612686395646,
      "num_tokens": 41129209.0,
      "step": 18770
    },
    {
      "entropy": 5.554688310623169,
      "epoch": 1.863338626439063,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046565795507041904,
      "loss": 5.3496,
      "mean_token_accuracy": 0.15461396127939225,
      "num_tokens": 41140715.0,
      "step": 18775
    },
    {
      "entropy": 5.608843803405762,
      "epoch": 1.8638348551012305,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000465639138405143,
      "loss": 5.3046,
      "mean_token_accuracy": 0.15713330507278442,
      "num_tokens": 41151080.0,
      "step": 18780
    },
    {
      "entropy": 5.615509796142578,
      "epoch": 1.8643310837633982,
      "grad_norm": 1.25,
      "learning_rate": 0.00046562031701242703,
      "loss": 5.3309,
      "mean_token_accuracy": 0.15880728811025618,
      "num_tokens": 41162816.0,
      "step": 18785
    },
    {
      "entropy": 5.552200365066528,
      "epoch": 1.8648273124255657,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046560149089273803,
      "loss": 5.2318,
      "mean_token_accuracy": 0.16821417212486267,
      "num_tokens": 41173456.0,
      "step": 18790
    },
    {
      "entropy": 5.517711210250854,
      "epoch": 1.8653235410877333,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004655826600465427,
      "loss": 5.1819,
      "mean_token_accuracy": 0.16196999996900557,
      "num_tokens": 41184436.0,
      "step": 18795
    },
    {
      "entropy": 5.513922595977784,
      "epoch": 1.8658197697499008,
      "grad_norm": 1.125,
      "learning_rate": 0.00046556382447430816,
      "loss": 5.2517,
      "mean_token_accuracy": 0.15884799063205718,
      "num_tokens": 41196417.0,
      "step": 18800
    },
    {
      "entropy": 5.554206085205078,
      "epoch": 1.8663159984120683,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046554498417650134,
      "loss": 5.2508,
      "mean_token_accuracy": 0.159711953997612,
      "num_tokens": 41206054.0,
      "step": 18805
    },
    {
      "entropy": 5.504018831253052,
      "epoch": 1.8668122270742358,
      "grad_norm": 1.5,
      "learning_rate": 0.0004655261391535895,
      "loss": 5.279,
      "mean_token_accuracy": 0.15796048194169998,
      "num_tokens": 41217002.0,
      "step": 18810
    },
    {
      "entropy": 5.678374814987182,
      "epoch": 1.8673084557364032,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004655072894060399,
      "loss": 5.4216,
      "mean_token_accuracy": 0.1494823455810547,
      "num_tokens": 41228485.0,
      "step": 18815
    },
    {
      "entropy": 5.620055627822876,
      "epoch": 1.8678046843985707,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004654884349343201,
      "loss": 5.3105,
      "mean_token_accuracy": 0.15747516006231307,
      "num_tokens": 41240600.0,
      "step": 18820
    },
    {
      "entropy": 5.627840614318847,
      "epoch": 1.8683009130607384,
      "grad_norm": 1.375,
      "learning_rate": 0.00046546957573889757,
      "loss": 5.3849,
      "mean_token_accuracy": 0.1516258530318737,
      "num_tokens": 41251870.0,
      "step": 18825
    },
    {
      "entropy": 5.533144760131836,
      "epoch": 1.868797141722906,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046545071182024,
      "loss": 5.3,
      "mean_token_accuracy": 0.154683855175972,
      "num_tokens": 41264679.0,
      "step": 18830
    },
    {
      "entropy": 5.647506046295166,
      "epoch": 1.8692933703850736,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046543184317881516,
      "loss": 5.3771,
      "mean_token_accuracy": 0.15450537353754043,
      "num_tokens": 41275690.0,
      "step": 18835
    },
    {
      "entropy": 5.590131425857544,
      "epoch": 1.869789599047241,
      "grad_norm": 2.4375,
      "learning_rate": 0.000465412969815091,
      "loss": 5.2429,
      "mean_token_accuracy": 0.1634903684258461,
      "num_tokens": 41285118.0,
      "step": 18840
    },
    {
      "entropy": 5.528267574310303,
      "epoch": 1.8702858277094085,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004653940917295354,
      "loss": 5.2517,
      "mean_token_accuracy": 0.16289470046758653,
      "num_tokens": 41296330.0,
      "step": 18845
    },
    {
      "entropy": 5.498303604125977,
      "epoch": 1.870782056371576,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004653752089226166,
      "loss": 5.2638,
      "mean_token_accuracy": 0.15908928960561752,
      "num_tokens": 41307051.0,
      "step": 18850
    },
    {
      "entropy": 5.558881139755249,
      "epoch": 1.8712782850337435,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004653563213948029,
      "loss": 5.2199,
      "mean_token_accuracy": 0.16120738685131072,
      "num_tokens": 41317111.0,
      "step": 18855
    },
    {
      "entropy": 5.5525898933410645,
      "epoch": 1.871774513695911,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046533742914656254,
      "loss": 5.217,
      "mean_token_accuracy": 0.16248553097248078,
      "num_tokens": 41327509.0,
      "step": 18860
    },
    {
      "entropy": 5.54439640045166,
      "epoch": 1.8722707423580784,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046531853217836414,
      "loss": 5.292,
      "mean_token_accuracy": 0.15621367692947388,
      "num_tokens": 41338721.0,
      "step": 18865
    },
    {
      "entropy": 5.602310371398926,
      "epoch": 1.8727669710202461,
      "grad_norm": 1.8984375,
      "learning_rate": 0.00046529963049067623,
      "loss": 5.3403,
      "mean_token_accuracy": 0.15519233494997026,
      "num_tokens": 41349767.0,
      "step": 18870
    },
    {
      "entropy": 5.677579021453857,
      "epoch": 1.8732631996824136,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046528072408396747,
      "loss": 5.4388,
      "mean_token_accuracy": 0.15191676765680312,
      "num_tokens": 41361151.0,
      "step": 18875
    },
    {
      "entropy": 5.5869241714477536,
      "epoch": 1.8737594283445813,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004652618129587069,
      "loss": 5.2538,
      "mean_token_accuracy": 0.15802043527364731,
      "num_tokens": 41372875.0,
      "step": 18880
    },
    {
      "entropy": 5.626903104782104,
      "epoch": 1.8742556570067488,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046524289711536327,
      "loss": 5.3233,
      "mean_token_accuracy": 0.1543267086148262,
      "num_tokens": 41383284.0,
      "step": 18885
    },
    {
      "entropy": 5.535853385925293,
      "epoch": 1.8747518856689163,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004652239765544057,
      "loss": 5.2858,
      "mean_token_accuracy": 0.16177460700273513,
      "num_tokens": 41395389.0,
      "step": 18890
    },
    {
      "entropy": 5.514106702804566,
      "epoch": 1.8752481143310837,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046520505127630346,
      "loss": 5.2359,
      "mean_token_accuracy": 0.15665285140275956,
      "num_tokens": 41406740.0,
      "step": 18895
    },
    {
      "entropy": 5.541116189956665,
      "epoch": 1.8757443429932512,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046518612128152576,
      "loss": 5.2317,
      "mean_token_accuracy": 0.16383861154317855,
      "num_tokens": 41417279.0,
      "step": 18900
    },
    {
      "entropy": 5.509190034866333,
      "epoch": 1.8762405716554187,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004651671865705421,
      "loss": 5.2224,
      "mean_token_accuracy": 0.15601858347654343,
      "num_tokens": 41428574.0,
      "step": 18905
    },
    {
      "entropy": 5.5127284049987795,
      "epoch": 1.8767368003175864,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046514824714382204,
      "loss": 5.1955,
      "mean_token_accuracy": 0.16119302958250045,
      "num_tokens": 41438784.0,
      "step": 18910
    },
    {
      "entropy": 5.561254596710205,
      "epoch": 1.8772330289797539,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004651293030018351,
      "loss": 5.2432,
      "mean_token_accuracy": 0.16104420274496078,
      "num_tokens": 41448481.0,
      "step": 18915
    },
    {
      "entropy": 5.599605989456177,
      "epoch": 1.8777292576419216,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046511035414505115,
      "loss": 5.31,
      "mean_token_accuracy": 0.1605406105518341,
      "num_tokens": 41459097.0,
      "step": 18920
    },
    {
      "entropy": 5.600082635879517,
      "epoch": 1.878225486304089,
      "grad_norm": 1.5,
      "learning_rate": 0.0004650914005739401,
      "loss": 5.3483,
      "mean_token_accuracy": 0.15189947783946992,
      "num_tokens": 41469647.0,
      "step": 18925
    },
    {
      "entropy": 5.543548583984375,
      "epoch": 1.8787217149662565,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004650724422889719,
      "loss": 5.2637,
      "mean_token_accuracy": 0.15662575513124466,
      "num_tokens": 41479826.0,
      "step": 18930
    },
    {
      "entropy": 5.575293779373169,
      "epoch": 1.879217943628424,
      "grad_norm": 1.25,
      "learning_rate": 0.0004650534792906167,
      "loss": 5.2191,
      "mean_token_accuracy": 0.1612800493836403,
      "num_tokens": 41490551.0,
      "step": 18935
    },
    {
      "entropy": 5.534440183639527,
      "epoch": 1.8797141722905915,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004650345115793447,
      "loss": 5.2545,
      "mean_token_accuracy": 0.16557215005159379,
      "num_tokens": 41500908.0,
      "step": 18940
    },
    {
      "entropy": 5.5384808540344235,
      "epoch": 1.880210400952759,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046501553915562644,
      "loss": 5.2194,
      "mean_token_accuracy": 0.16028596460819244,
      "num_tokens": 41511311.0,
      "step": 18945
    },
    {
      "entropy": 5.554424667358399,
      "epoch": 1.8807066296149264,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046499656201993213,
      "loss": 5.2672,
      "mean_token_accuracy": 0.1639294907450676,
      "num_tokens": 41522068.0,
      "step": 18950
    },
    {
      "entropy": 5.5406115531921385,
      "epoch": 1.881202858277094,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046497758017273255,
      "loss": 5.2815,
      "mean_token_accuracy": 0.15442289113998414,
      "num_tokens": 41534099.0,
      "step": 18955
    },
    {
      "entropy": 5.6098557472229,
      "epoch": 1.8816990869392616,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046495859361449834,
      "loss": 5.3385,
      "mean_token_accuracy": 0.15457507371902465,
      "num_tokens": 41545870.0,
      "step": 18960
    },
    {
      "entropy": 5.596281385421753,
      "epoch": 1.8821953156014293,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004649396023457003,
      "loss": 5.3052,
      "mean_token_accuracy": 0.15675653368234635,
      "num_tokens": 41558175.0,
      "step": 18965
    },
    {
      "entropy": 5.522418355941772,
      "epoch": 1.8826915442635968,
      "grad_norm": 1.375,
      "learning_rate": 0.0004649206063668094,
      "loss": 5.2737,
      "mean_token_accuracy": 0.16761212050914764,
      "num_tokens": 41569326.0,
      "step": 18970
    },
    {
      "entropy": 5.550592470169067,
      "epoch": 1.8831877729257642,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004649016056782968,
      "loss": 5.2331,
      "mean_token_accuracy": 0.16291841268539428,
      "num_tokens": 41580065.0,
      "step": 18975
    },
    {
      "entropy": 5.627201509475708,
      "epoch": 1.8836840015879317,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004648826002806334,
      "loss": 5.3629,
      "mean_token_accuracy": 0.15983131825923919,
      "num_tokens": 41590915.0,
      "step": 18980
    },
    {
      "entropy": 5.537559652328492,
      "epoch": 1.8841802302500992,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004648635901742909,
      "loss": 5.2102,
      "mean_token_accuracy": 0.15866461545228958,
      "num_tokens": 41602096.0,
      "step": 18985
    },
    {
      "entropy": 5.550024509429932,
      "epoch": 1.8846764589122667,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004648445753597403,
      "loss": 5.3227,
      "mean_token_accuracy": 0.1558950737118721,
      "num_tokens": 41612589.0,
      "step": 18990
    },
    {
      "entropy": 5.567956352233887,
      "epoch": 1.8851726875744343,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046482555583745336,
      "loss": 5.2147,
      "mean_token_accuracy": 0.15964014381170272,
      "num_tokens": 41624163.0,
      "step": 18995
    },
    {
      "entropy": 5.55069899559021,
      "epoch": 1.8856689162366018,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004648065316079017,
      "loss": 5.2299,
      "mean_token_accuracy": 0.16141905039548873,
      "num_tokens": 41635919.0,
      "step": 19000
    },
    {
      "entropy": 5.528603363037109,
      "epoch": 1.8861651448987695,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046478750267155703,
      "loss": 5.2524,
      "mean_token_accuracy": 0.15777750313282013,
      "num_tokens": 41644723.0,
      "step": 19005
    },
    {
      "entropy": 5.519236755371094,
      "epoch": 1.886661373560937,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046476846902889125,
      "loss": 5.3135,
      "mean_token_accuracy": 0.15854015648365022,
      "num_tokens": 41657137.0,
      "step": 19010
    },
    {
      "entropy": 5.573355388641358,
      "epoch": 1.8871576022231045,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004647494306803764,
      "loss": 5.2861,
      "mean_token_accuracy": 0.1554910346865654,
      "num_tokens": 41667773.0,
      "step": 19015
    },
    {
      "entropy": 5.676171159744262,
      "epoch": 1.887653830885272,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046473038762648446,
      "loss": 5.364,
      "mean_token_accuracy": 0.15431434512138367,
      "num_tokens": 41680549.0,
      "step": 19020
    },
    {
      "entropy": 5.560794162750244,
      "epoch": 1.8881500595474394,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046471133986768784,
      "loss": 5.2401,
      "mean_token_accuracy": 0.16086287647485734,
      "num_tokens": 41689700.0,
      "step": 19025
    },
    {
      "entropy": 5.572253704071045,
      "epoch": 1.888646288209607,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046469228740445874,
      "loss": 5.2908,
      "mean_token_accuracy": 0.15945230200886726,
      "num_tokens": 41699734.0,
      "step": 19030
    },
    {
      "entropy": 5.546799516677856,
      "epoch": 1.8891425168717744,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046467323023726965,
      "loss": 5.295,
      "mean_token_accuracy": 0.15682059824466704,
      "num_tokens": 41710644.0,
      "step": 19035
    },
    {
      "entropy": 5.486372566223144,
      "epoch": 1.889638745533942,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046465416836659313,
      "loss": 5.2122,
      "mean_token_accuracy": 0.16788814216852188,
      "num_tokens": 41721416.0,
      "step": 19040
    },
    {
      "entropy": 5.520211935043335,
      "epoch": 1.8901349741961095,
      "grad_norm": 1.3125,
      "learning_rate": 0.000464635101792902,
      "loss": 5.216,
      "mean_token_accuracy": 0.1577487498521805,
      "num_tokens": 41732139.0,
      "step": 19045
    },
    {
      "entropy": 5.552287626266479,
      "epoch": 1.8906312028582772,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046461603051666883,
      "loss": 5.3329,
      "mean_token_accuracy": 0.15586797297000884,
      "num_tokens": 41744233.0,
      "step": 19050
    },
    {
      "entropy": 5.522658157348633,
      "epoch": 1.8911274315204447,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046459695453836676,
      "loss": 5.247,
      "mean_token_accuracy": 0.16289881914854049,
      "num_tokens": 41755514.0,
      "step": 19055
    },
    {
      "entropy": 5.552477407455444,
      "epoch": 1.8916236601826122,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004645778738584688,
      "loss": 5.2572,
      "mean_token_accuracy": 0.16209413558244706,
      "num_tokens": 41765613.0,
      "step": 19060
    },
    {
      "entropy": 5.579401445388794,
      "epoch": 1.8921198888447797,
      "grad_norm": 1.125,
      "learning_rate": 0.0004645587884774481,
      "loss": 5.3169,
      "mean_token_accuracy": 0.15649330615997314,
      "num_tokens": 41777294.0,
      "step": 19065
    },
    {
      "entropy": 5.488919925689697,
      "epoch": 1.8926161175069471,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004645396983957778,
      "loss": 5.1964,
      "mean_token_accuracy": 0.17000870853662492,
      "num_tokens": 41788278.0,
      "step": 19070
    },
    {
      "entropy": 5.531976652145386,
      "epoch": 1.8931123461691146,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046452060361393147,
      "loss": 5.2894,
      "mean_token_accuracy": 0.15687019303441047,
      "num_tokens": 41798663.0,
      "step": 19075
    },
    {
      "entropy": 5.56573748588562,
      "epoch": 1.8936085748312823,
      "grad_norm": 1.5,
      "learning_rate": 0.0004645015041323825,
      "loss": 5.2435,
      "mean_token_accuracy": 0.16194546967744827,
      "num_tokens": 41808711.0,
      "step": 19080
    },
    {
      "entropy": 5.519096326828003,
      "epoch": 1.8941048034934498,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046448239995160455,
      "loss": 5.2868,
      "mean_token_accuracy": 0.15419536978006362,
      "num_tokens": 41819679.0,
      "step": 19085
    },
    {
      "entropy": 5.554450654983521,
      "epoch": 1.8946010321556173,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004644632910720715,
      "loss": 5.2628,
      "mean_token_accuracy": 0.16118558198213578,
      "num_tokens": 41829649.0,
      "step": 19090
    },
    {
      "entropy": 5.6025361061096195,
      "epoch": 1.895097260817785,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000464444177494257,
      "loss": 5.3301,
      "mean_token_accuracy": 0.15206891894340516,
      "num_tokens": 41842523.0,
      "step": 19095
    },
    {
      "entropy": 5.554118585586548,
      "epoch": 1.8955934894799524,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004644250592186351,
      "loss": 5.2828,
      "mean_token_accuracy": 0.15699696391820908,
      "num_tokens": 41854514.0,
      "step": 19100
    },
    {
      "entropy": 5.5821703433990475,
      "epoch": 1.89608971814212,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046440593624568,
      "loss": 5.2565,
      "mean_token_accuracy": 0.15625796020030974,
      "num_tokens": 41865166.0,
      "step": 19105
    },
    {
      "entropy": 5.54726734161377,
      "epoch": 1.8965859468042874,
      "grad_norm": 1.125,
      "learning_rate": 0.0004643868085758657,
      "loss": 5.2201,
      "mean_token_accuracy": 0.16200809925794601,
      "num_tokens": 41877264.0,
      "step": 19110
    },
    {
      "entropy": 5.532669162750244,
      "epoch": 1.8970821754664549,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046436767620966675,
      "loss": 5.2184,
      "mean_token_accuracy": 0.15917464047670365,
      "num_tokens": 41887345.0,
      "step": 19115
    },
    {
      "entropy": 5.53315954208374,
      "epoch": 1.8975784041286223,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046434853914755744,
      "loss": 5.287,
      "mean_token_accuracy": 0.15705930739641188,
      "num_tokens": 41898877.0,
      "step": 19120
    },
    {
      "entropy": 5.514408826828003,
      "epoch": 1.89807463279079,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046432939739001237,
      "loss": 5.2186,
      "mean_token_accuracy": 0.15997352004051207,
      "num_tokens": 41910593.0,
      "step": 19125
    },
    {
      "entropy": 5.534682273864746,
      "epoch": 1.8985708614529575,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046431025093750613,
      "loss": 5.2576,
      "mean_token_accuracy": 0.16654251366853715,
      "num_tokens": 41920719.0,
      "step": 19130
    },
    {
      "entropy": 5.5554587841033936,
      "epoch": 1.8990670901151252,
      "grad_norm": 1.96875,
      "learning_rate": 0.00046429109979051367,
      "loss": 5.2788,
      "mean_token_accuracy": 0.14829385727643968,
      "num_tokens": 41931630.0,
      "step": 19135
    },
    {
      "entropy": 5.585926628112793,
      "epoch": 1.8995633187772927,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046427194394950985,
      "loss": 5.3232,
      "mean_token_accuracy": 0.15782080441713334,
      "num_tokens": 41943076.0,
      "step": 19140
    },
    {
      "entropy": 5.626083183288574,
      "epoch": 1.9000595474394602,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004642527834149696,
      "loss": 5.2309,
      "mean_token_accuracy": 0.16067853569984436,
      "num_tokens": 41954294.0,
      "step": 19145
    },
    {
      "entropy": 5.502947521209717,
      "epoch": 1.9005557761016276,
      "grad_norm": 1.25,
      "learning_rate": 0.00046423361818736814,
      "loss": 5.1881,
      "mean_token_accuracy": 0.1621055781841278,
      "num_tokens": 41964943.0,
      "step": 19150
    },
    {
      "entropy": 5.468431282043457,
      "epoch": 1.901052004763795,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004642144482671807,
      "loss": 5.233,
      "mean_token_accuracy": 0.16116826832294465,
      "num_tokens": 41975211.0,
      "step": 19155
    },
    {
      "entropy": 5.569794464111328,
      "epoch": 1.9015482334259626,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046419527365488275,
      "loss": 5.2799,
      "mean_token_accuracy": 0.15719657987356186,
      "num_tokens": 41986492.0,
      "step": 19160
    },
    {
      "entropy": 5.568549585342407,
      "epoch": 1.90204446208813,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004641760943509496,
      "loss": 5.2475,
      "mean_token_accuracy": 0.1630413353443146,
      "num_tokens": 41997613.0,
      "step": 19165
    },
    {
      "entropy": 5.504824018478393,
      "epoch": 1.9025406907502977,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000464156910355857,
      "loss": 5.325,
      "mean_token_accuracy": 0.15329516381025315,
      "num_tokens": 42008597.0,
      "step": 19170
    },
    {
      "entropy": 5.496608543395996,
      "epoch": 1.9030369194124652,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046413772167008057,
      "loss": 5.2787,
      "mean_token_accuracy": 0.16236144602298735,
      "num_tokens": 42019698.0,
      "step": 19175
    },
    {
      "entropy": 5.54793062210083,
      "epoch": 1.903533148074633,
      "grad_norm": 1.125,
      "learning_rate": 0.00046411852829409626,
      "loss": 5.1952,
      "mean_token_accuracy": 0.16090109646320344,
      "num_tokens": 42029473.0,
      "step": 19180
    },
    {
      "entropy": 5.562540912628174,
      "epoch": 1.9040293767368004,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046409933022837984,
      "loss": 5.3286,
      "mean_token_accuracy": 0.1539129450917244,
      "num_tokens": 42041494.0,
      "step": 19185
    },
    {
      "entropy": 5.563806581497192,
      "epoch": 1.9045256053989679,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004640801274734076,
      "loss": 5.2788,
      "mean_token_accuracy": 0.1607805833220482,
      "num_tokens": 42051798.0,
      "step": 19190
    },
    {
      "entropy": 5.517984008789062,
      "epoch": 1.9050218340611353,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004640609200296557,
      "loss": 5.2056,
      "mean_token_accuracy": 0.1629386141896248,
      "num_tokens": 42061802.0,
      "step": 19195
    },
    {
      "entropy": 5.543541717529297,
      "epoch": 1.9055180627233028,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004640417078976003,
      "loss": 5.353,
      "mean_token_accuracy": 0.15261506736278535,
      "num_tokens": 42071503.0,
      "step": 19200
    },
    {
      "entropy": 5.444900321960449,
      "epoch": 1.9060142913854703,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004640224910777179,
      "loss": 5.1749,
      "mean_token_accuracy": 0.16232041418552398,
      "num_tokens": 42082104.0,
      "step": 19205
    },
    {
      "entropy": 5.546650075912476,
      "epoch": 1.906510520047638,
      "grad_norm": 1.25,
      "learning_rate": 0.000464003269570485,
      "loss": 5.2522,
      "mean_token_accuracy": 0.15868535339832307,
      "num_tokens": 42092798.0,
      "step": 19210
    },
    {
      "entropy": 5.557008695602417,
      "epoch": 1.9070067487098055,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004639840433763783,
      "loss": 5.2865,
      "mean_token_accuracy": 0.15933801233768463,
      "num_tokens": 42103062.0,
      "step": 19215
    },
    {
      "entropy": 5.548327970504761,
      "epoch": 1.9075029773719732,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046396481249587454,
      "loss": 5.3442,
      "mean_token_accuracy": 0.16002322882413864,
      "num_tokens": 42115062.0,
      "step": 19220
    },
    {
      "entropy": 5.6038836479187015,
      "epoch": 1.9079992060341406,
      "grad_norm": 1.25,
      "learning_rate": 0.0004639455769294506,
      "loss": 5.3232,
      "mean_token_accuracy": 0.1554343193769455,
      "num_tokens": 42125937.0,
      "step": 19225
    },
    {
      "entropy": 5.606100416183471,
      "epoch": 1.9084954346963081,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004639263366775835,
      "loss": 5.2656,
      "mean_token_accuracy": 0.15970194935798646,
      "num_tokens": 42136916.0,
      "step": 19230
    },
    {
      "entropy": 5.5860155582427975,
      "epoch": 1.9089916633584756,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046390709174075036,
      "loss": 5.3257,
      "mean_token_accuracy": 0.1559859812259674,
      "num_tokens": 42148348.0,
      "step": 19235
    },
    {
      "entropy": 5.587312173843384,
      "epoch": 1.909487892020643,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004638878421194284,
      "loss": 5.2159,
      "mean_token_accuracy": 0.16541010439395903,
      "num_tokens": 42157586.0,
      "step": 19240
    },
    {
      "entropy": 5.557260131835937,
      "epoch": 1.9099841206828105,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000463868587814095,
      "loss": 5.2909,
      "mean_token_accuracy": 0.15444362461566924,
      "num_tokens": 42169536.0,
      "step": 19245
    },
    {
      "entropy": 5.5124634265899655,
      "epoch": 1.910480349344978,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004638493288252276,
      "loss": 5.2204,
      "mean_token_accuracy": 0.16480433195829391,
      "num_tokens": 42180024.0,
      "step": 19250
    },
    {
      "entropy": 5.51949028968811,
      "epoch": 1.9109765780071457,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046383006515330374,
      "loss": 5.2285,
      "mean_token_accuracy": 0.1624395951628685,
      "num_tokens": 42191070.0,
      "step": 19255
    },
    {
      "entropy": 5.5651764392852785,
      "epoch": 1.9114728066693132,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004638107967988012,
      "loss": 5.2913,
      "mean_token_accuracy": 0.15982888638973236,
      "num_tokens": 42203643.0,
      "step": 19260
    },
    {
      "entropy": 5.583628177642822,
      "epoch": 1.9119690353314809,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046379152376219775,
      "loss": 5.3263,
      "mean_token_accuracy": 0.15091496407985688,
      "num_tokens": 42214979.0,
      "step": 19265
    },
    {
      "entropy": 5.611435413360596,
      "epoch": 1.9124652639936484,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046377224604397127,
      "loss": 5.2919,
      "mean_token_accuracy": 0.16110427528619767,
      "num_tokens": 42225152.0,
      "step": 19270
    },
    {
      "entropy": 5.583011817932129,
      "epoch": 1.9129614926558158,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004637529636445998,
      "loss": 5.2403,
      "mean_token_accuracy": 0.16228984594345092,
      "num_tokens": 42235151.0,
      "step": 19275
    },
    {
      "entropy": 5.52640585899353,
      "epoch": 1.9134577213179833,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004637336765645617,
      "loss": 5.2334,
      "mean_token_accuracy": 0.16715970635414124,
      "num_tokens": 42245419.0,
      "step": 19280
    },
    {
      "entropy": 5.470816516876221,
      "epoch": 1.9139539499801508,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046371438480433504,
      "loss": 5.1989,
      "mean_token_accuracy": 0.16543122231960297,
      "num_tokens": 42255836.0,
      "step": 19285
    },
    {
      "entropy": 5.527376747131347,
      "epoch": 1.9144501786423183,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004636950883643983,
      "loss": 5.3031,
      "mean_token_accuracy": 0.1570167750120163,
      "num_tokens": 42267490.0,
      "step": 19290
    },
    {
      "entropy": 5.580207443237304,
      "epoch": 1.914946407304486,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046367578724523,
      "loss": 5.2609,
      "mean_token_accuracy": 0.1602252811193466,
      "num_tokens": 42276957.0,
      "step": 19295
    },
    {
      "entropy": 5.544660568237305,
      "epoch": 1.9154426359666534,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004636564814473087,
      "loss": 5.1494,
      "mean_token_accuracy": 0.16603201776742935,
      "num_tokens": 42288469.0,
      "step": 19300
    },
    {
      "entropy": 5.474358701705933,
      "epoch": 1.9159388646288211,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004636371709711131,
      "loss": 5.2043,
      "mean_token_accuracy": 0.17468105703592302,
      "num_tokens": 42299141.0,
      "step": 19305
    },
    {
      "entropy": 5.5234476089477536,
      "epoch": 1.9164350932909886,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046361785581712224,
      "loss": 5.3208,
      "mean_token_accuracy": 0.16260591745376587,
      "num_tokens": 42310651.0,
      "step": 19310
    },
    {
      "entropy": 5.508219766616821,
      "epoch": 1.916931321953156,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000463598535985815,
      "loss": 5.121,
      "mean_token_accuracy": 0.16965116560459137,
      "num_tokens": 42320529.0,
      "step": 19315
    },
    {
      "entropy": 5.50427131652832,
      "epoch": 1.9174275506153236,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004635792114776704,
      "loss": 5.1592,
      "mean_token_accuracy": 0.1655648022890091,
      "num_tokens": 42331216.0,
      "step": 19320
    },
    {
      "entropy": 5.427016115188598,
      "epoch": 1.917923779277491,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004635598822931677,
      "loss": 5.179,
      "mean_token_accuracy": 0.16593878269195556,
      "num_tokens": 42341730.0,
      "step": 19325
    },
    {
      "entropy": 5.548481130599976,
      "epoch": 1.9184200079396585,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046354054843278624,
      "loss": 5.2891,
      "mean_token_accuracy": 0.15692350566387175,
      "num_tokens": 42352643.0,
      "step": 19330
    },
    {
      "entropy": 5.633994054794312,
      "epoch": 1.918916236601826,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046352120989700553,
      "loss": 5.3339,
      "mean_token_accuracy": 0.15505815744400026,
      "num_tokens": 42363979.0,
      "step": 19335
    },
    {
      "entropy": 5.595178270339966,
      "epoch": 1.9194124652639937,
      "grad_norm": 1.28125,
      "learning_rate": 0.000463501866686305,
      "loss": 5.2776,
      "mean_token_accuracy": 0.1574398748576641,
      "num_tokens": 42374104.0,
      "step": 19340
    },
    {
      "entropy": 5.5676422119140625,
      "epoch": 1.9199086939261611,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046348251880116433,
      "loss": 5.254,
      "mean_token_accuracy": 0.15921638906002045,
      "num_tokens": 42383576.0,
      "step": 19345
    },
    {
      "entropy": 5.531959772109985,
      "epoch": 1.9204049225883288,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00046346316624206325,
      "loss": 5.2577,
      "mean_token_accuracy": 0.15887395590543746,
      "num_tokens": 42394625.0,
      "step": 19350
    },
    {
      "entropy": 5.586769533157349,
      "epoch": 1.9209011512504963,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004634438090094819,
      "loss": 5.2113,
      "mean_token_accuracy": 0.16497422903776168,
      "num_tokens": 42404764.0,
      "step": 19355
    },
    {
      "entropy": 5.55294361114502,
      "epoch": 1.9213973799126638,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046342444710390006,
      "loss": 5.2,
      "mean_token_accuracy": 0.16572719663381577,
      "num_tokens": 42414386.0,
      "step": 19360
    },
    {
      "entropy": 5.514994955062866,
      "epoch": 1.9218936085748313,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000463405080525798,
      "loss": 5.2832,
      "mean_token_accuracy": 0.1568116083741188,
      "num_tokens": 42425671.0,
      "step": 19365
    },
    {
      "entropy": 5.541105699539185,
      "epoch": 1.9223898372369987,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004633857092756559,
      "loss": 5.2913,
      "mean_token_accuracy": 0.15727721005678177,
      "num_tokens": 42436365.0,
      "step": 19370
    },
    {
      "entropy": 5.554082298278809,
      "epoch": 1.9228860658991662,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004633663333539542,
      "loss": 5.1783,
      "mean_token_accuracy": 0.15844169855117798,
      "num_tokens": 42447624.0,
      "step": 19375
    },
    {
      "entropy": 5.567620706558228,
      "epoch": 1.923382294561334,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046334695276117324,
      "loss": 5.2613,
      "mean_token_accuracy": 0.15993646085262297,
      "num_tokens": 42458013.0,
      "step": 19380
    },
    {
      "entropy": 5.5469659805297855,
      "epoch": 1.9238785232235014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004633275674977937,
      "loss": 5.3077,
      "mean_token_accuracy": 0.15541875809431077,
      "num_tokens": 42469807.0,
      "step": 19385
    },
    {
      "entropy": 5.499782371520996,
      "epoch": 1.9243747518856689,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004633081775642964,
      "loss": 5.2066,
      "mean_token_accuracy": 0.16460628509521485,
      "num_tokens": 42479240.0,
      "step": 19390
    },
    {
      "entropy": 5.551316785812378,
      "epoch": 1.9248709805478366,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046328878296116196,
      "loss": 5.2836,
      "mean_token_accuracy": 0.16053603440523148,
      "num_tokens": 42489558.0,
      "step": 19395
    },
    {
      "entropy": 5.5680504322052,
      "epoch": 1.925367209210004,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046326938368887146,
      "loss": 5.3057,
      "mean_token_accuracy": 0.1621929720044136,
      "num_tokens": 42501252.0,
      "step": 19400
    },
    {
      "entropy": 5.589534044265747,
      "epoch": 1.9258634378721715,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046324997974790596,
      "loss": 5.3466,
      "mean_token_accuracy": 0.1579672172665596,
      "num_tokens": 42512330.0,
      "step": 19405
    },
    {
      "entropy": 5.587731504440308,
      "epoch": 1.926359666534339,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004632305711387466,
      "loss": 5.3335,
      "mean_token_accuracy": 0.15376077145338057,
      "num_tokens": 42525216.0,
      "step": 19410
    },
    {
      "entropy": 5.553724956512451,
      "epoch": 1.9268558951965065,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046321115786187466,
      "loss": 5.262,
      "mean_token_accuracy": 0.1583865150809288,
      "num_tokens": 42537047.0,
      "step": 19415
    },
    {
      "entropy": 5.523451232910157,
      "epoch": 1.927352123858674,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046319173991777153,
      "loss": 5.2639,
      "mean_token_accuracy": 0.15880534946918487,
      "num_tokens": 42549784.0,
      "step": 19420
    },
    {
      "entropy": 5.532038021087646,
      "epoch": 1.9278483525208416,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004631723173069188,
      "loss": 5.2568,
      "mean_token_accuracy": 0.15634471774101258,
      "num_tokens": 42561079.0,
      "step": 19425
    },
    {
      "entropy": 5.573559856414795,
      "epoch": 1.928344581183009,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004631528900297981,
      "loss": 5.3002,
      "mean_token_accuracy": 0.15225514769554138,
      "num_tokens": 42572531.0,
      "step": 19430
    },
    {
      "entropy": 5.556236124038696,
      "epoch": 1.9288408098451768,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046313345808689113,
      "loss": 5.2279,
      "mean_token_accuracy": 0.16165337860584258,
      "num_tokens": 42583836.0,
      "step": 19435
    },
    {
      "entropy": 5.618405532836914,
      "epoch": 1.9293370385073443,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004631140214786798,
      "loss": 5.3516,
      "mean_token_accuracy": 0.15280528441071511,
      "num_tokens": 42595338.0,
      "step": 19440
    },
    {
      "entropy": 5.556585454940796,
      "epoch": 1.9298332671695118,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046309458020564606,
      "loss": 5.2799,
      "mean_token_accuracy": 0.15833909660577775,
      "num_tokens": 42606804.0,
      "step": 19445
    },
    {
      "entropy": 5.530850410461426,
      "epoch": 1.9303294958316792,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046307513426827214,
      "loss": 5.1893,
      "mean_token_accuracy": 0.16727192997932433,
      "num_tokens": 42616987.0,
      "step": 19450
    },
    {
      "entropy": 5.5574640274047855,
      "epoch": 1.9308257244938467,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046305568366704006,
      "loss": 5.2708,
      "mean_token_accuracy": 0.15454816073179245,
      "num_tokens": 42627633.0,
      "step": 19455
    },
    {
      "entropy": 5.55747880935669,
      "epoch": 1.9313219531560142,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004630362284024322,
      "loss": 5.2485,
      "mean_token_accuracy": 0.15641742795705796,
      "num_tokens": 42638802.0,
      "step": 19460
    },
    {
      "entropy": 5.563270854949951,
      "epoch": 1.9318181818181817,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046301676847493116,
      "loss": 5.2432,
      "mean_token_accuracy": 0.1663563296198845,
      "num_tokens": 42650802.0,
      "step": 19465
    },
    {
      "entropy": 5.5193713188171385,
      "epoch": 1.9323144104803494,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046299730388501935,
      "loss": 5.182,
      "mean_token_accuracy": 0.16349311470985411,
      "num_tokens": 42660290.0,
      "step": 19470
    },
    {
      "entropy": 5.52647213935852,
      "epoch": 1.9328106391425168,
      "grad_norm": 2.390625,
      "learning_rate": 0.00046297783463317957,
      "loss": 5.2871,
      "mean_token_accuracy": 0.1604434758424759,
      "num_tokens": 42671182.0,
      "step": 19475
    },
    {
      "entropy": 5.58445725440979,
      "epoch": 1.9333068678046845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004629583607198945,
      "loss": 5.27,
      "mean_token_accuracy": 0.16437845081090927,
      "num_tokens": 42682499.0,
      "step": 19480
    },
    {
      "entropy": 5.50007176399231,
      "epoch": 1.933803096466852,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046293888214564704,
      "loss": 5.2942,
      "mean_token_accuracy": 0.15744320750236512,
      "num_tokens": 42693728.0,
      "step": 19485
    },
    {
      "entropy": 5.551230239868164,
      "epoch": 1.9342993251290195,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046291939891092036,
      "loss": 5.293,
      "mean_token_accuracy": 0.15233704894781114,
      "num_tokens": 42704877.0,
      "step": 19490
    },
    {
      "entropy": 5.633644342422485,
      "epoch": 1.934795553791187,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004628999110161975,
      "loss": 5.3215,
      "mean_token_accuracy": 0.14997628778219224,
      "num_tokens": 42715787.0,
      "step": 19495
    },
    {
      "entropy": 5.585055637359619,
      "epoch": 1.9352917824533544,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046288041846196174,
      "loss": 5.2771,
      "mean_token_accuracy": 0.15860652327537536,
      "num_tokens": 42727557.0,
      "step": 19500
    },
    {
      "entropy": 5.597351121902466,
      "epoch": 1.935788011115522,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046286092124869647,
      "loss": 5.2738,
      "mean_token_accuracy": 0.1590213358402252,
      "num_tokens": 42737540.0,
      "step": 19505
    },
    {
      "entropy": 5.628708982467652,
      "epoch": 1.9362842397776896,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046284141937688513,
      "loss": 5.3101,
      "mean_token_accuracy": 0.16046586781740188,
      "num_tokens": 42748716.0,
      "step": 19510
    },
    {
      "entropy": 5.62988166809082,
      "epoch": 1.936780468439857,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046282191284701123,
      "loss": 5.3221,
      "mean_token_accuracy": 0.15178160816431047,
      "num_tokens": 42758991.0,
      "step": 19515
    },
    {
      "entropy": 5.618515968322754,
      "epoch": 1.9372766971020248,
      "grad_norm": 1.25,
      "learning_rate": 0.0004628024016595587,
      "loss": 5.3276,
      "mean_token_accuracy": 0.14955669045448303,
      "num_tokens": 42771105.0,
      "step": 19520
    },
    {
      "entropy": 5.573976230621338,
      "epoch": 1.9377729257641922,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046278288581501136,
      "loss": 5.264,
      "mean_token_accuracy": 0.16588084101676942,
      "num_tokens": 42783317.0,
      "step": 19525
    },
    {
      "entropy": 5.640382146835327,
      "epoch": 1.9382691544263597,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046276336531385305,
      "loss": 5.3899,
      "mean_token_accuracy": 0.15519438236951827,
      "num_tokens": 42794998.0,
      "step": 19530
    },
    {
      "entropy": 5.584550189971924,
      "epoch": 1.9387653830885272,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046274384015656785,
      "loss": 5.2593,
      "mean_token_accuracy": 0.16181355118751525,
      "num_tokens": 42806197.0,
      "step": 19535
    },
    {
      "entropy": 5.47781195640564,
      "epoch": 1.9392616117506947,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046272431034363997,
      "loss": 5.16,
      "mean_token_accuracy": 0.15608370304107666,
      "num_tokens": 42816522.0,
      "step": 19540
    },
    {
      "entropy": 5.457324552536011,
      "epoch": 1.9397578404128621,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004627047758755538,
      "loss": 5.1761,
      "mean_token_accuracy": 0.16384976655244826,
      "num_tokens": 42826003.0,
      "step": 19545
    },
    {
      "entropy": 5.558835649490357,
      "epoch": 1.9402540690750296,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004626852367527935,
      "loss": 5.256,
      "mean_token_accuracy": 0.1553269386291504,
      "num_tokens": 42836346.0,
      "step": 19550
    },
    {
      "entropy": 5.5633869647979735,
      "epoch": 1.9407502977371973,
      "grad_norm": 1.125,
      "learning_rate": 0.0004626656929758438,
      "loss": 5.3166,
      "mean_token_accuracy": 0.15767741203308105,
      "num_tokens": 42846296.0,
      "step": 19555
    },
    {
      "entropy": 5.634596109390259,
      "epoch": 1.9412465263993648,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004626461445451894,
      "loss": 5.3337,
      "mean_token_accuracy": 0.1575388044118881,
      "num_tokens": 42856323.0,
      "step": 19560
    },
    {
      "entropy": 5.628345155715943,
      "epoch": 1.9417427550615325,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004626265914613149,
      "loss": 5.325,
      "mean_token_accuracy": 0.15667954534292222,
      "num_tokens": 42866937.0,
      "step": 19565
    },
    {
      "entropy": 5.6041570663452145,
      "epoch": 1.9422389837237,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004626070337247052,
      "loss": 5.2987,
      "mean_token_accuracy": 0.1539609521627426,
      "num_tokens": 42877194.0,
      "step": 19570
    },
    {
      "entropy": 5.576890897750855,
      "epoch": 1.9427352123858674,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004625874713358453,
      "loss": 5.3166,
      "mean_token_accuracy": 0.15063564777374266,
      "num_tokens": 42887741.0,
      "step": 19575
    },
    {
      "entropy": 5.543630647659302,
      "epoch": 1.943231441048035,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046256790429522035,
      "loss": 5.2743,
      "mean_token_accuracy": 0.16902696788311006,
      "num_tokens": 42898195.0,
      "step": 19580
    },
    {
      "entropy": 5.62006516456604,
      "epoch": 1.9437276697102024,
      "grad_norm": 1.375,
      "learning_rate": 0.0004625483326033155,
      "loss": 5.2975,
      "mean_token_accuracy": 0.1603761076927185,
      "num_tokens": 42908852.0,
      "step": 19585
    },
    {
      "entropy": 5.54951958656311,
      "epoch": 1.9442238983723699,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004625287562606162,
      "loss": 5.215,
      "mean_token_accuracy": 0.15960230231285094,
      "num_tokens": 42920166.0,
      "step": 19590
    },
    {
      "entropy": 5.582198238372802,
      "epoch": 1.9447201270345376,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004625091752676078,
      "loss": 5.2883,
      "mean_token_accuracy": 0.15553211271762848,
      "num_tokens": 42931677.0,
      "step": 19595
    },
    {
      "entropy": 5.547936964035034,
      "epoch": 1.945216355696705,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000462489589624776,
      "loss": 5.2567,
      "mean_token_accuracy": 0.15888901203870773,
      "num_tokens": 42942346.0,
      "step": 19600
    },
    {
      "entropy": 5.4947301864624025,
      "epoch": 1.9457125843588727,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004624699993326063,
      "loss": 5.2349,
      "mean_token_accuracy": 0.16136184632778167,
      "num_tokens": 42953821.0,
      "step": 19605
    },
    {
      "entropy": 5.527226209640503,
      "epoch": 1.9462088130210402,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046245040439158456,
      "loss": 5.2329,
      "mean_token_accuracy": 0.1611968994140625,
      "num_tokens": 42964358.0,
      "step": 19610
    },
    {
      "entropy": 5.524620294570923,
      "epoch": 1.9467050416832077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046243080480219673,
      "loss": 5.2278,
      "mean_token_accuracy": 0.15637097954750062,
      "num_tokens": 42976459.0,
      "step": 19615
    },
    {
      "entropy": 5.62258996963501,
      "epoch": 1.9472012703453752,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004624112005649288,
      "loss": 5.3702,
      "mean_token_accuracy": 0.15202370211482047,
      "num_tokens": 42988603.0,
      "step": 19620
    },
    {
      "entropy": 5.53992657661438,
      "epoch": 1.9476974990075426,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046239159168026694,
      "loss": 5.2822,
      "mean_token_accuracy": 0.1613189160823822,
      "num_tokens": 43000262.0,
      "step": 19625
    },
    {
      "entropy": 5.4922832489013675,
      "epoch": 1.94819372766971,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004623719781486974,
      "loss": 5.1974,
      "mean_token_accuracy": 0.16725579500198365,
      "num_tokens": 43009363.0,
      "step": 19630
    },
    {
      "entropy": 5.604364633560181,
      "epoch": 1.9486899563318776,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046235235997070664,
      "loss": 5.2952,
      "mean_token_accuracy": 0.1588408589363098,
      "num_tokens": 43019893.0,
      "step": 19635
    },
    {
      "entropy": 5.506718587875366,
      "epoch": 1.9491861849940453,
      "grad_norm": 1.28125,
      "learning_rate": 0.000462332737146781,
      "loss": 5.2004,
      "mean_token_accuracy": 0.16476341038942338,
      "num_tokens": 43031152.0,
      "step": 19640
    },
    {
      "entropy": 5.532318162918091,
      "epoch": 1.9496824136562128,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004623131096774071,
      "loss": 5.3264,
      "mean_token_accuracy": 0.15356008112430572,
      "num_tokens": 43041247.0,
      "step": 19645
    },
    {
      "entropy": 5.590437316894532,
      "epoch": 1.9501786423183805,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004622934775630717,
      "loss": 5.267,
      "mean_token_accuracy": 0.1534041628241539,
      "num_tokens": 43052065.0,
      "step": 19650
    },
    {
      "entropy": 5.567824649810791,
      "epoch": 1.950674870980548,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046227384080426163,
      "loss": 5.2986,
      "mean_token_accuracy": 0.15645177215337752,
      "num_tokens": 43063217.0,
      "step": 19655
    },
    {
      "entropy": 5.553720855712891,
      "epoch": 1.9511710996427154,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004622541994014639,
      "loss": 5.299,
      "mean_token_accuracy": 0.1620335653424263,
      "num_tokens": 43074463.0,
      "step": 19660
    },
    {
      "entropy": 5.6717315196990965,
      "epoch": 1.9516673283048829,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004622345533551655,
      "loss": 5.3391,
      "mean_token_accuracy": 0.1514980047941208,
      "num_tokens": 43086004.0,
      "step": 19665
    },
    {
      "entropy": 5.582527780532837,
      "epoch": 1.9521635569670504,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004622149026658537,
      "loss": 5.2388,
      "mean_token_accuracy": 0.15898457616567613,
      "num_tokens": 43096973.0,
      "step": 19670
    },
    {
      "entropy": 5.6124194145202635,
      "epoch": 1.9526597856292178,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004621952473340156,
      "loss": 5.3245,
      "mean_token_accuracy": 0.1574496939778328,
      "num_tokens": 43107337.0,
      "step": 19675
    },
    {
      "entropy": 5.5718590259552006,
      "epoch": 1.9531560142913855,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004621755873601388,
      "loss": 5.2368,
      "mean_token_accuracy": 0.16641955375671386,
      "num_tokens": 43118668.0,
      "step": 19680
    },
    {
      "entropy": 5.600230407714844,
      "epoch": 1.953652242953553,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004621559227447108,
      "loss": 5.3359,
      "mean_token_accuracy": 0.15836250334978103,
      "num_tokens": 43129030.0,
      "step": 19685
    },
    {
      "entropy": 5.487634992599487,
      "epoch": 1.9541484716157205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004621362534882192,
      "loss": 5.2056,
      "mean_token_accuracy": 0.1640731617808342,
      "num_tokens": 43139120.0,
      "step": 19690
    },
    {
      "entropy": 5.581285285949707,
      "epoch": 1.9546447002778882,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004621165795911517,
      "loss": 5.3057,
      "mean_token_accuracy": 0.15892233997583388,
      "num_tokens": 43152390.0,
      "step": 19695
    },
    {
      "entropy": 5.5700047492980955,
      "epoch": 1.9551409289400556,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046209690105399623,
      "loss": 5.1079,
      "mean_token_accuracy": 0.16808599084615708,
      "num_tokens": 43162230.0,
      "step": 19700
    },
    {
      "entropy": 5.4614691734313965,
      "epoch": 1.9556371576022231,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046207721787724085,
      "loss": 5.2307,
      "mean_token_accuracy": 0.15783885568380357,
      "num_tokens": 43172751.0,
      "step": 19705
    },
    {
      "entropy": 5.60653281211853,
      "epoch": 1.9561333862643906,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046205753006137347,
      "loss": 5.3322,
      "mean_token_accuracy": 0.1628636121749878,
      "num_tokens": 43183835.0,
      "step": 19710
    },
    {
      "entropy": 5.687164640426635,
      "epoch": 1.956629614926558,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046203783760688245,
      "loss": 5.207,
      "mean_token_accuracy": 0.1584305137395859,
      "num_tokens": 43193178.0,
      "step": 19715
    },
    {
      "entropy": 5.541518115997315,
      "epoch": 1.9571258435887255,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004620181405142561,
      "loss": 5.3726,
      "mean_token_accuracy": 0.1512363851070404,
      "num_tokens": 43204968.0,
      "step": 19720
    },
    {
      "entropy": 5.515860652923584,
      "epoch": 1.9576220722508932,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004619984387839829,
      "loss": 5.3842,
      "mean_token_accuracy": 0.1559639886021614,
      "num_tokens": 43214916.0,
      "step": 19725
    },
    {
      "entropy": 5.70793924331665,
      "epoch": 1.9581183009130607,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00046197873241655133,
      "loss": 5.3694,
      "mean_token_accuracy": 0.15703784078359603,
      "num_tokens": 43226374.0,
      "step": 19730
    },
    {
      "entropy": 5.558064889907837,
      "epoch": 1.9586145295752284,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004619590214124501,
      "loss": 5.2233,
      "mean_token_accuracy": 0.15867508947849274,
      "num_tokens": 43237296.0,
      "step": 19735
    },
    {
      "entropy": 5.640716505050659,
      "epoch": 1.9591107582373959,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046193930577216797,
      "loss": 5.4102,
      "mean_token_accuracy": 0.14489317163825036,
      "num_tokens": 43248220.0,
      "step": 19740
    },
    {
      "entropy": 5.484742879867554,
      "epoch": 1.9596069868995634,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004619195854961939,
      "loss": 5.1773,
      "mean_token_accuracy": 0.1689458578824997,
      "num_tokens": 43259913.0,
      "step": 19745
    },
    {
      "entropy": 5.5831602096557615,
      "epoch": 1.9601032155617308,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046189986058501687,
      "loss": 5.3227,
      "mean_token_accuracy": 0.15096641778945924,
      "num_tokens": 43272317.0,
      "step": 19750
    },
    {
      "entropy": 5.501505565643311,
      "epoch": 1.9605994442238983,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000461880131039126,
      "loss": 5.1883,
      "mean_token_accuracy": 0.162892584502697,
      "num_tokens": 43283170.0,
      "step": 19755
    },
    {
      "entropy": 5.49232964515686,
      "epoch": 1.9610956728860658,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004618603968590107,
      "loss": 5.2591,
      "mean_token_accuracy": 0.16236865147948265,
      "num_tokens": 43293899.0,
      "step": 19760
    },
    {
      "entropy": 5.607998704910278,
      "epoch": 1.9615919015482333,
      "grad_norm": 1.25,
      "learning_rate": 0.00046184065804516007,
      "loss": 5.3147,
      "mean_token_accuracy": 0.16079042553901673,
      "num_tokens": 43305638.0,
      "step": 19765
    },
    {
      "entropy": 5.643455076217651,
      "epoch": 1.962088130210401,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004618209145980638,
      "loss": 5.3709,
      "mean_token_accuracy": 0.15374862924218177,
      "num_tokens": 43315900.0,
      "step": 19770
    },
    {
      "entropy": 5.6645020008087155,
      "epoch": 1.9625843588725684,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004618011665182114,
      "loss": 5.4281,
      "mean_token_accuracy": 0.15237081795930862,
      "num_tokens": 43328464.0,
      "step": 19775
    },
    {
      "entropy": 5.476817989349366,
      "epoch": 1.9630805875347361,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004617814138060926,
      "loss": 5.1284,
      "mean_token_accuracy": 0.1704138398170471,
      "num_tokens": 43339303.0,
      "step": 19780
    },
    {
      "entropy": 5.542270374298096,
      "epoch": 1.9635768161969036,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004617616564621972,
      "loss": 5.357,
      "mean_token_accuracy": 0.15440430641174316,
      "num_tokens": 43349269.0,
      "step": 19785
    },
    {
      "entropy": 5.532397317886352,
      "epoch": 1.964073044859071,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046174189448701516,
      "loss": 5.2159,
      "mean_token_accuracy": 0.16005280762910842,
      "num_tokens": 43361090.0,
      "step": 19790
    },
    {
      "entropy": 5.553050756454468,
      "epoch": 1.9645692735212386,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004617221278810365,
      "loss": 5.2094,
      "mean_token_accuracy": 0.160170316696167,
      "num_tokens": 43373448.0,
      "step": 19795
    },
    {
      "entropy": 5.680231618881225,
      "epoch": 1.965065502183406,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046170235664475147,
      "loss": 5.359,
      "mean_token_accuracy": 0.15345073044300078,
      "num_tokens": 43386202.0,
      "step": 19800
    },
    {
      "entropy": 5.559441614151001,
      "epoch": 1.9655617308455735,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00046168258077865026,
      "loss": 5.2714,
      "mean_token_accuracy": 0.16316028833389282,
      "num_tokens": 43396363.0,
      "step": 19805
    },
    {
      "entropy": 5.488270473480225,
      "epoch": 1.9660579595077412,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004616628002832233,
      "loss": 5.149,
      "mean_token_accuracy": 0.16853403747081758,
      "num_tokens": 43406076.0,
      "step": 19810
    },
    {
      "entropy": 5.575496244430542,
      "epoch": 1.9665541881699087,
      "grad_norm": 1.578125,
      "learning_rate": 0.00046164301515896124,
      "loss": 5.3121,
      "mean_token_accuracy": 0.15993259251117706,
      "num_tokens": 43416226.0,
      "step": 19815
    },
    {
      "entropy": 5.54623327255249,
      "epoch": 1.9670504168320764,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004616232254063544,
      "loss": 5.1467,
      "mean_token_accuracy": 0.16562157422304152,
      "num_tokens": 43427364.0,
      "step": 19820
    },
    {
      "entropy": 5.5268641948699955,
      "epoch": 1.9675466454942439,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046160343102589384,
      "loss": 5.2748,
      "mean_token_accuracy": 0.16566108018159867,
      "num_tokens": 43437854.0,
      "step": 19825
    },
    {
      "entropy": 5.559770584106445,
      "epoch": 1.9680428741564113,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004615836320180702,
      "loss": 5.3193,
      "mean_token_accuracy": 0.15895950347185134,
      "num_tokens": 43449370.0,
      "step": 19830
    },
    {
      "entropy": 5.5148192882537845,
      "epoch": 1.9685391028185788,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046156382838337455,
      "loss": 5.2681,
      "mean_token_accuracy": 0.16693799644708635,
      "num_tokens": 43461631.0,
      "step": 19835
    },
    {
      "entropy": 5.561821126937867,
      "epoch": 1.9690353314807463,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046154402012229807,
      "loss": 5.2456,
      "mean_token_accuracy": 0.15486990809440612,
      "num_tokens": 43472357.0,
      "step": 19840
    },
    {
      "entropy": 5.544839811325073,
      "epoch": 1.9695315601429138,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046152420723533175,
      "loss": 5.2144,
      "mean_token_accuracy": 0.16121936291456224,
      "num_tokens": 43483642.0,
      "step": 19845
    },
    {
      "entropy": 5.610239028930664,
      "epoch": 1.9700277888050812,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046150438972296706,
      "loss": 5.3056,
      "mean_token_accuracy": 0.16251853853464127,
      "num_tokens": 43495136.0,
      "step": 19850
    },
    {
      "entropy": 5.504488754272461,
      "epoch": 1.970524017467249,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046148456758569536,
      "loss": 5.2739,
      "mean_token_accuracy": 0.15714021623134614,
      "num_tokens": 43506376.0,
      "step": 19855
    },
    {
      "entropy": 5.543992042541504,
      "epoch": 1.9710202461294164,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046146474082400823,
      "loss": 5.2446,
      "mean_token_accuracy": 0.1634073555469513,
      "num_tokens": 43516658.0,
      "step": 19860
    },
    {
      "entropy": 5.566338014602661,
      "epoch": 1.971516474791584,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046144490943839735,
      "loss": 5.2107,
      "mean_token_accuracy": 0.1634721502661705,
      "num_tokens": 43526907.0,
      "step": 19865
    },
    {
      "entropy": 5.5744983673095705,
      "epoch": 1.9720127034537516,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004614250734293544,
      "loss": 5.3007,
      "mean_token_accuracy": 0.15490227788686753,
      "num_tokens": 43537739.0,
      "step": 19870
    },
    {
      "entropy": 5.469098234176636,
      "epoch": 1.972508932115919,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00046140523279737146,
      "loss": 5.2616,
      "mean_token_accuracy": 0.15898085832595826,
      "num_tokens": 43549240.0,
      "step": 19875
    },
    {
      "entropy": 5.545454597473144,
      "epoch": 1.9730051607780865,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004613853875429403,
      "loss": 5.2101,
      "mean_token_accuracy": 0.157559597492218,
      "num_tokens": 43559857.0,
      "step": 19880
    },
    {
      "entropy": 5.504265546798706,
      "epoch": 1.973501389440254,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004613655376665532,
      "loss": 5.2056,
      "mean_token_accuracy": 0.162647145986557,
      "num_tokens": 43570262.0,
      "step": 19885
    },
    {
      "entropy": 5.61243634223938,
      "epoch": 1.9739976181024215,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004613456831687024,
      "loss": 5.2951,
      "mean_token_accuracy": 0.15344480574131011,
      "num_tokens": 43581965.0,
      "step": 19890
    },
    {
      "entropy": 5.556191635131836,
      "epoch": 1.9744938467645892,
      "grad_norm": 1.875,
      "learning_rate": 0.00046132582404988004,
      "loss": 5.2408,
      "mean_token_accuracy": 0.16310577243566513,
      "num_tokens": 43592728.0,
      "step": 19895
    },
    {
      "entropy": 5.533427810668945,
      "epoch": 1.9749900754267566,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046130596031057885,
      "loss": 5.2159,
      "mean_token_accuracy": 0.15735997259616852,
      "num_tokens": 43602437.0,
      "step": 19900
    },
    {
      "entropy": 5.589560699462891,
      "epoch": 1.9754863040889243,
      "grad_norm": 2.09375,
      "learning_rate": 0.00046128609195129135,
      "loss": 5.3282,
      "mean_token_accuracy": 0.15575515180826188,
      "num_tokens": 43613115.0,
      "step": 19905
    },
    {
      "entropy": 5.649006605148315,
      "epoch": 1.9759825327510918,
      "grad_norm": 5.03125,
      "learning_rate": 0.0004612662189725101,
      "loss": 5.3157,
      "mean_token_accuracy": 0.1565030500292778,
      "num_tokens": 43623700.0,
      "step": 19910
    },
    {
      "entropy": 5.626991653442383,
      "epoch": 1.9764787614132593,
      "grad_norm": 1.234375,
      "learning_rate": 0.000461246341374728,
      "loss": 5.319,
      "mean_token_accuracy": 0.15817543119192123,
      "num_tokens": 43633712.0,
      "step": 19915
    },
    {
      "entropy": 5.553853368759155,
      "epoch": 1.9769749900754268,
      "grad_norm": 1.671875,
      "learning_rate": 0.00046122645915843796,
      "loss": 5.3525,
      "mean_token_accuracy": 0.15670206993818284,
      "num_tokens": 43643705.0,
      "step": 19920
    },
    {
      "entropy": 5.588766241073609,
      "epoch": 1.9774712187375942,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046120657232413305,
      "loss": 5.3033,
      "mean_token_accuracy": 0.16170886009931565,
      "num_tokens": 43655376.0,
      "step": 19925
    },
    {
      "entropy": 5.552834987640381,
      "epoch": 1.9779674473997617,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004611866808723064,
      "loss": 5.2135,
      "mean_token_accuracy": 0.16469504684209824,
      "num_tokens": 43666329.0,
      "step": 19930
    },
    {
      "entropy": 5.502244710922241,
      "epoch": 1.9784636760619292,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046116678480345117,
      "loss": 5.2317,
      "mean_token_accuracy": 0.163393534719944,
      "num_tokens": 43676415.0,
      "step": 19935
    },
    {
      "entropy": 5.5006256103515625,
      "epoch": 1.9789599047240969,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046114688411806083,
      "loss": 5.215,
      "mean_token_accuracy": 0.16491199284791946,
      "num_tokens": 43686997.0,
      "step": 19940
    },
    {
      "entropy": 5.52919373512268,
      "epoch": 1.9794561333862644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046112697881662895,
      "loss": 5.2002,
      "mean_token_accuracy": 0.16448879688978196,
      "num_tokens": 43697342.0,
      "step": 19945
    },
    {
      "entropy": 5.614354038238526,
      "epoch": 1.979952362048432,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046110706889964905,
      "loss": 5.3057,
      "mean_token_accuracy": 0.15996172577142714,
      "num_tokens": 43707876.0,
      "step": 19950
    },
    {
      "entropy": 5.543427801132202,
      "epoch": 1.9804485907105995,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004610871543676149,
      "loss": 5.1722,
      "mean_token_accuracy": 0.16604748368263245,
      "num_tokens": 43718268.0,
      "step": 19955
    },
    {
      "entropy": 5.54981164932251,
      "epoch": 1.980944819372767,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046106723522102024,
      "loss": 5.3296,
      "mean_token_accuracy": 0.15599067956209184,
      "num_tokens": 43729853.0,
      "step": 19960
    },
    {
      "entropy": 5.570429086685181,
      "epoch": 1.9814410480349345,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046104731146035904,
      "loss": 5.2863,
      "mean_token_accuracy": 0.15822838395833969,
      "num_tokens": 43739886.0,
      "step": 19965
    },
    {
      "entropy": 5.599892044067383,
      "epoch": 1.981937276697102,
      "grad_norm": 1.75,
      "learning_rate": 0.0004610273830861255,
      "loss": 5.2834,
      "mean_token_accuracy": 0.15635916590690613,
      "num_tokens": 43750536.0,
      "step": 19970
    },
    {
      "entropy": 5.571971797943116,
      "epoch": 1.9824335053592694,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004610074500988137,
      "loss": 5.2713,
      "mean_token_accuracy": 0.15873189121484757,
      "num_tokens": 43763192.0,
      "step": 19975
    },
    {
      "entropy": 5.6035479545593265,
      "epoch": 1.9829297340214371,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004609875124989179,
      "loss": 5.289,
      "mean_token_accuracy": 0.16080703735351562,
      "num_tokens": 43774117.0,
      "step": 19980
    },
    {
      "entropy": 5.536799383163452,
      "epoch": 1.9834259626836046,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004609675702869326,
      "loss": 5.2164,
      "mean_token_accuracy": 0.1617936059832573,
      "num_tokens": 43784708.0,
      "step": 19985
    },
    {
      "entropy": 5.583167266845703,
      "epoch": 1.983922191345772,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004609476234633523,
      "loss": 5.3037,
      "mean_token_accuracy": 0.16122852861881257,
      "num_tokens": 43796767.0,
      "step": 19990
    },
    {
      "entropy": 5.597214126586914,
      "epoch": 1.9844184200079398,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046092767202867156,
      "loss": 5.3915,
      "mean_token_accuracy": 0.14806200191378593,
      "num_tokens": 43807203.0,
      "step": 19995
    },
    {
      "entropy": 5.550383806228638,
      "epoch": 1.9849146486701073,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046090771598338525,
      "loss": 5.2789,
      "mean_token_accuracy": 0.1605715796351433,
      "num_tokens": 43817123.0,
      "step": 20000
    },
    {
      "entropy": 5.556264877319336,
      "epoch": 1.9854108773322747,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004608877553279881,
      "loss": 5.2423,
      "mean_token_accuracy": 0.16583844274282455,
      "num_tokens": 43828625.0,
      "step": 20005
    },
    {
      "entropy": 5.563394021987915,
      "epoch": 1.9859071059944422,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004608677900629753,
      "loss": 5.212,
      "mean_token_accuracy": 0.16050174310803414,
      "num_tokens": 43840012.0,
      "step": 20010
    },
    {
      "entropy": 5.4338477611541744,
      "epoch": 1.9864033346566097,
      "grad_norm": 1.6875,
      "learning_rate": 0.00046084782018884177,
      "loss": 5.2143,
      "mean_token_accuracy": 0.16733813285827637,
      "num_tokens": 43851253.0,
      "step": 20015
    },
    {
      "entropy": 5.615855884552002,
      "epoch": 1.9868995633187772,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046082784570608266,
      "loss": 5.3282,
      "mean_token_accuracy": 0.16017450243234635,
      "num_tokens": 43864389.0,
      "step": 20020
    },
    {
      "entropy": 5.5969099521636965,
      "epoch": 1.9873957919809448,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046080786661519354,
      "loss": 5.3272,
      "mean_token_accuracy": 0.15491539388895034,
      "num_tokens": 43876446.0,
      "step": 20025
    },
    {
      "entropy": 5.558966541290284,
      "epoch": 1.9878920206431123,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004607878829166696,
      "loss": 5.2798,
      "mean_token_accuracy": 0.15505143851041794,
      "num_tokens": 43888318.0,
      "step": 20030
    },
    {
      "entropy": 5.509389591217041,
      "epoch": 1.98838824930528,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046076789461100653,
      "loss": 5.2165,
      "mean_token_accuracy": 0.16668246686458588,
      "num_tokens": 43898358.0,
      "step": 20035
    },
    {
      "entropy": 5.578261995315552,
      "epoch": 1.9888844779674475,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004607479016987001,
      "loss": 5.2753,
      "mean_token_accuracy": 0.1581972911953926,
      "num_tokens": 43909831.0,
      "step": 20040
    },
    {
      "entropy": 5.572990226745605,
      "epoch": 1.989380706629615,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004607279041802458,
      "loss": 5.2338,
      "mean_token_accuracy": 0.16282927393913268,
      "num_tokens": 43919648.0,
      "step": 20045
    },
    {
      "entropy": 5.503210973739624,
      "epoch": 1.9898769352917824,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046070790205613967,
      "loss": 5.2891,
      "mean_token_accuracy": 0.1574067711830139,
      "num_tokens": 43930563.0,
      "step": 20050
    },
    {
      "entropy": 5.5998293399810795,
      "epoch": 1.99037316395395,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046068789532687784,
      "loss": 5.3572,
      "mean_token_accuracy": 0.15028757005929946,
      "num_tokens": 43942490.0,
      "step": 20055
    },
    {
      "entropy": 5.520554637908935,
      "epoch": 1.9908693926161174,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004606678839929563,
      "loss": 5.1305,
      "mean_token_accuracy": 0.17117466777563095,
      "num_tokens": 43952187.0,
      "step": 20060
    },
    {
      "entropy": 5.497833299636841,
      "epoch": 1.9913656212782849,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004606478680548713,
      "loss": 5.2004,
      "mean_token_accuracy": 0.16442528367042542,
      "num_tokens": 43964203.0,
      "step": 20065
    },
    {
      "entropy": 5.534693908691406,
      "epoch": 1.9918618499404526,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046062784751311917,
      "loss": 5.254,
      "mean_token_accuracy": 0.16327286809682845,
      "num_tokens": 43974048.0,
      "step": 20070
    },
    {
      "entropy": 5.61813907623291,
      "epoch": 1.99235807860262,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046060782236819647,
      "loss": 5.2812,
      "mean_token_accuracy": 0.16077876538038255,
      "num_tokens": 43984469.0,
      "step": 20075
    },
    {
      "entropy": 5.539441728591919,
      "epoch": 1.9928543072647877,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004605877926205996,
      "loss": 5.2969,
      "mean_token_accuracy": 0.15771833509206773,
      "num_tokens": 43995494.0,
      "step": 20080
    },
    {
      "entropy": 5.537288999557495,
      "epoch": 1.9933505359269552,
      "grad_norm": 1.125,
      "learning_rate": 0.0004605677582708256,
      "loss": 5.2131,
      "mean_token_accuracy": 0.16007684022188187,
      "num_tokens": 44005869.0,
      "step": 20085
    },
    {
      "entropy": 5.645225286483765,
      "epoch": 1.9938467645891227,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004605477193193709,
      "loss": 5.2737,
      "mean_token_accuracy": 0.1607801914215088,
      "num_tokens": 44016956.0,
      "step": 20090
    },
    {
      "entropy": 5.622159814834594,
      "epoch": 1.9943429932512902,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004605276757667326,
      "loss": 5.3517,
      "mean_token_accuracy": 0.14640547037124635,
      "num_tokens": 44029524.0,
      "step": 20095
    },
    {
      "entropy": 5.5533052444458,
      "epoch": 1.9948392219134576,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004605076276134078,
      "loss": 5.2166,
      "mean_token_accuracy": 0.1647270768880844,
      "num_tokens": 44039523.0,
      "step": 20100
    },
    {
      "entropy": 5.637494230270386,
      "epoch": 1.9953354505756251,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046048757485989345,
      "loss": 5.3454,
      "mean_token_accuracy": 0.15358293429017067,
      "num_tokens": 44050538.0,
      "step": 20105
    },
    {
      "entropy": 5.620696258544922,
      "epoch": 1.9958316792377928,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000460467517506687,
      "loss": 5.3289,
      "mean_token_accuracy": 0.15161218345165253,
      "num_tokens": 44061108.0,
      "step": 20110
    },
    {
      "entropy": 5.519739294052124,
      "epoch": 1.9963279078999603,
      "grad_norm": 2.125,
      "learning_rate": 0.0004604474555542858,
      "loss": 5.2503,
      "mean_token_accuracy": 0.16396655291318893,
      "num_tokens": 44072706.0,
      "step": 20115
    },
    {
      "entropy": 5.541420650482178,
      "epoch": 1.996824136562128,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004604273890031872,
      "loss": 5.2424,
      "mean_token_accuracy": 0.15618920475244522,
      "num_tokens": 44082697.0,
      "step": 20120
    },
    {
      "entropy": 5.52349624633789,
      "epoch": 1.9973203652242955,
      "grad_norm": 1.640625,
      "learning_rate": 0.000460407317853889,
      "loss": 5.2236,
      "mean_token_accuracy": 0.1690586432814598,
      "num_tokens": 44093265.0,
      "step": 20125
    },
    {
      "entropy": 5.534213018417359,
      "epoch": 1.997816593886463,
      "grad_norm": 1.375,
      "learning_rate": 0.0004603872421068888,
      "loss": 5.2239,
      "mean_token_accuracy": 0.156887885928154,
      "num_tokens": 44103535.0,
      "step": 20130
    },
    {
      "entropy": 5.498982858657837,
      "epoch": 1.9983128225486304,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004603671617626845,
      "loss": 5.1875,
      "mean_token_accuracy": 0.16251697540283203,
      "num_tokens": 44113903.0,
      "step": 20135
    },
    {
      "entropy": 5.582382154464722,
      "epoch": 1.9988090512107979,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046034707682177393,
      "loss": 5.2488,
      "mean_token_accuracy": 0.1646324709057808,
      "num_tokens": 44125148.0,
      "step": 20140
    },
    {
      "entropy": 5.681245517730713,
      "epoch": 1.9993052798729654,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004603269872846554,
      "loss": 5.3546,
      "mean_token_accuracy": 0.14991891086101533,
      "num_tokens": 44136563.0,
      "step": 20145
    },
    {
      "entropy": 5.675927591323853,
      "epoch": 1.9998015085351328,
      "grad_norm": 1.96875,
      "learning_rate": 0.00046030689315182676,
      "loss": 5.3163,
      "mean_token_accuracy": 0.16035114228725433,
      "num_tokens": 44147521.0,
      "step": 20150
    },
    {
      "entropy": 5.600184440612793,
      "epoch": 2.0002977371973003,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004602867944237866,
      "loss": 5.3,
      "mean_token_accuracy": 0.15758753716945648,
      "num_tokens": 44158977.0,
      "step": 20155
    },
    {
      "entropy": 5.606082487106323,
      "epoch": 2.0007939658594682,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046026669110103315,
      "loss": 5.2708,
      "mean_token_accuracy": 0.16173941045999526,
      "num_tokens": 44170986.0,
      "step": 20160
    },
    {
      "entropy": 5.679245281219482,
      "epoch": 2.0012901945216357,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046024658318406493,
      "loss": 5.328,
      "mean_token_accuracy": 0.15520637035369872,
      "num_tokens": 44180936.0,
      "step": 20165
    },
    {
      "entropy": 5.656619358062744,
      "epoch": 2.001786423183803,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004602264706733807,
      "loss": 5.3395,
      "mean_token_accuracy": 0.15634195506572723,
      "num_tokens": 44191741.0,
      "step": 20170
    },
    {
      "entropy": 5.5566819190979,
      "epoch": 2.0022826518459707,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004602063535694791,
      "loss": 5.1913,
      "mean_token_accuracy": 0.15801966041326523,
      "num_tokens": 44202310.0,
      "step": 20175
    },
    {
      "entropy": 5.56564416885376,
      "epoch": 2.002778880508138,
      "grad_norm": 1.21875,
      "learning_rate": 0.000460186231872859,
      "loss": 5.2312,
      "mean_token_accuracy": 0.16265376508235932,
      "num_tokens": 44213514.0,
      "step": 20180
    },
    {
      "entropy": 5.592441129684448,
      "epoch": 2.0032751091703056,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046016610558401936,
      "loss": 5.285,
      "mean_token_accuracy": 0.15426699221134185,
      "num_tokens": 44224718.0,
      "step": 20185
    },
    {
      "entropy": 5.585710763931274,
      "epoch": 2.003771337832473,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046014597470345935,
      "loss": 5.3036,
      "mean_token_accuracy": 0.16195093244314193,
      "num_tokens": 44235999.0,
      "step": 20190
    },
    {
      "entropy": 5.569272136688232,
      "epoch": 2.0042675664946406,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004601258392316781,
      "loss": 5.1478,
      "mean_token_accuracy": 0.16161946952342987,
      "num_tokens": 44246323.0,
      "step": 20195
    },
    {
      "entropy": 5.527615118026733,
      "epoch": 2.0047637951568085,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046010569916917496,
      "loss": 5.1679,
      "mean_token_accuracy": 0.1683705687522888,
      "num_tokens": 44256323.0,
      "step": 20200
    },
    {
      "entropy": 5.587560081481934,
      "epoch": 2.005260023818976,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004600855545164493,
      "loss": 5.1886,
      "mean_token_accuracy": 0.1600828468799591,
      "num_tokens": 44266845.0,
      "step": 20205
    },
    {
      "entropy": 5.570067834854126,
      "epoch": 2.0057562524811434,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004600654052740008,
      "loss": 5.2791,
      "mean_token_accuracy": 0.1638893373310566,
      "num_tokens": 44277861.0,
      "step": 20210
    },
    {
      "entropy": 5.534161710739136,
      "epoch": 2.006252481143311,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004600452514423289,
      "loss": 5.2321,
      "mean_token_accuracy": 0.16352788358926773,
      "num_tokens": 44288511.0,
      "step": 20215
    },
    {
      "entropy": 5.526432800292969,
      "epoch": 2.0067487098054784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004600250930219336,
      "loss": 5.2983,
      "mean_token_accuracy": 0.1505104884505272,
      "num_tokens": 44299752.0,
      "step": 20220
    },
    {
      "entropy": 5.538737058639526,
      "epoch": 2.007244938467646,
      "grad_norm": 1.125,
      "learning_rate": 0.00046000493001331465,
      "loss": 5.1431,
      "mean_token_accuracy": 0.16525601595640182,
      "num_tokens": 44309443.0,
      "step": 20225
    },
    {
      "entropy": 5.591734266281128,
      "epoch": 2.0077411671298133,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045998476241697214,
      "loss": 5.1997,
      "mean_token_accuracy": 0.16392762809991837,
      "num_tokens": 44320030.0,
      "step": 20230
    },
    {
      "entropy": 5.455964517593384,
      "epoch": 2.008237395791981,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004599645902334061,
      "loss": 5.1087,
      "mean_token_accuracy": 0.16695154458284378,
      "num_tokens": 44330077.0,
      "step": 20235
    },
    {
      "entropy": 5.422861194610595,
      "epoch": 2.0087336244541483,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045994441346311677,
      "loss": 5.1936,
      "mean_token_accuracy": 0.1672942101955414,
      "num_tokens": 44340336.0,
      "step": 20240
    },
    {
      "entropy": 5.564670085906982,
      "epoch": 2.009229853116316,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004599242321066045,
      "loss": 5.1772,
      "mean_token_accuracy": 0.16238430589437486,
      "num_tokens": 44350704.0,
      "step": 20245
    },
    {
      "entropy": 5.570580005645752,
      "epoch": 2.0097260817784837,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045990404616436976,
      "loss": 5.2095,
      "mean_token_accuracy": 0.16575430035591127,
      "num_tokens": 44360541.0,
      "step": 20250
    },
    {
      "entropy": 5.494733047485352,
      "epoch": 2.010222310440651,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004598838556369131,
      "loss": 5.1327,
      "mean_token_accuracy": 0.16695835590362548,
      "num_tokens": 44372132.0,
      "step": 20255
    },
    {
      "entropy": 5.453971815109253,
      "epoch": 2.0107185391028186,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045986366052473525,
      "loss": 5.1298,
      "mean_token_accuracy": 0.17009460777044297,
      "num_tokens": 44383594.0,
      "step": 20260
    },
    {
      "entropy": 5.439764928817749,
      "epoch": 2.011214767764986,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004598434608283369,
      "loss": 5.1444,
      "mean_token_accuracy": 0.16489843279123306,
      "num_tokens": 44394101.0,
      "step": 20265
    },
    {
      "entropy": 5.586856889724731,
      "epoch": 2.0117109964271536,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004598232565482191,
      "loss": 5.2576,
      "mean_token_accuracy": 0.16091257333755493,
      "num_tokens": 44405670.0,
      "step": 20270
    },
    {
      "entropy": 5.608515310287475,
      "epoch": 2.012207225089321,
      "grad_norm": 1.125,
      "learning_rate": 0.00045980304768488274,
      "loss": 5.226,
      "mean_token_accuracy": 0.16530393064022064,
      "num_tokens": 44417748.0,
      "step": 20275
    },
    {
      "entropy": 5.595344305038452,
      "epoch": 2.0127034537514885,
      "grad_norm": 1.25,
      "learning_rate": 0.00045978283423882906,
      "loss": 5.1964,
      "mean_token_accuracy": 0.1615544766187668,
      "num_tokens": 44429648.0,
      "step": 20280
    },
    {
      "entropy": 5.500770044326782,
      "epoch": 2.0131996824136564,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045976261621055915,
      "loss": 5.1981,
      "mean_token_accuracy": 0.1623966097831726,
      "num_tokens": 44440867.0,
      "step": 20285
    },
    {
      "entropy": 5.559764575958252,
      "epoch": 2.013695911075824,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045974239360057457,
      "loss": 5.2552,
      "mean_token_accuracy": 0.15873301178216934,
      "num_tokens": 44452981.0,
      "step": 20290
    },
    {
      "entropy": 5.562725019454956,
      "epoch": 2.0141921397379914,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045972216640937676,
      "loss": 5.1724,
      "mean_token_accuracy": 0.16409362256526946,
      "num_tokens": 44464243.0,
      "step": 20295
    },
    {
      "entropy": 5.54937047958374,
      "epoch": 2.014688368400159,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004597019346374672,
      "loss": 5.2436,
      "mean_token_accuracy": 0.16327481120824813,
      "num_tokens": 44474376.0,
      "step": 20300
    },
    {
      "entropy": 5.539242601394653,
      "epoch": 2.0151845970623263,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004596816982853477,
      "loss": 5.1376,
      "mean_token_accuracy": 0.1741255536675453,
      "num_tokens": 44485686.0,
      "step": 20305
    },
    {
      "entropy": 5.468685293197632,
      "epoch": 2.015680825724494,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045966145735352,
      "loss": 5.1513,
      "mean_token_accuracy": 0.16923225820064544,
      "num_tokens": 44496096.0,
      "step": 20310
    },
    {
      "entropy": 5.606292819976806,
      "epoch": 2.0161770543866613,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045964121184248604,
      "loss": 5.2432,
      "mean_token_accuracy": 0.15729447305202485,
      "num_tokens": 44507555.0,
      "step": 20315
    },
    {
      "entropy": 5.5583514213562015,
      "epoch": 2.0166732830488288,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004596209617527479,
      "loss": 5.2252,
      "mean_token_accuracy": 0.16259772181510926,
      "num_tokens": 44519288.0,
      "step": 20320
    },
    {
      "entropy": 5.606596422195435,
      "epoch": 2.0171695117109962,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004596007070848078,
      "loss": 5.252,
      "mean_token_accuracy": 0.15295133143663406,
      "num_tokens": 44530146.0,
      "step": 20325
    },
    {
      "entropy": 5.571588850021362,
      "epoch": 2.017665740373164,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045958044783916795,
      "loss": 5.2425,
      "mean_token_accuracy": 0.15861518159508706,
      "num_tokens": 44541411.0,
      "step": 20330
    },
    {
      "entropy": 5.505281162261963,
      "epoch": 2.0181619690353316,
      "grad_norm": 1.25,
      "learning_rate": 0.0004595601840163306,
      "loss": 5.2115,
      "mean_token_accuracy": 0.1560682088136673,
      "num_tokens": 44552556.0,
      "step": 20335
    },
    {
      "entropy": 5.522555875778198,
      "epoch": 2.018658197697499,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045953991561679855,
      "loss": 5.21,
      "mean_token_accuracy": 0.16347582638263702,
      "num_tokens": 44563052.0,
      "step": 20340
    },
    {
      "entropy": 5.5466235160827635,
      "epoch": 2.0191544263596666,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004595196426410742,
      "loss": 5.167,
      "mean_token_accuracy": 0.16464006155729294,
      "num_tokens": 44574020.0,
      "step": 20345
    },
    {
      "entropy": 5.536630392074585,
      "epoch": 2.019650655021834,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045949936508966015,
      "loss": 5.1417,
      "mean_token_accuracy": 0.16943299323320388,
      "num_tokens": 44586261.0,
      "step": 20350
    },
    {
      "entropy": 5.5580822944641115,
      "epoch": 2.0201468836840015,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004594790829630596,
      "loss": 5.1804,
      "mean_token_accuracy": 0.16060441732406616,
      "num_tokens": 44596391.0,
      "step": 20355
    },
    {
      "entropy": 5.47349910736084,
      "epoch": 2.020643112346169,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004594587962617752,
      "loss": 5.1268,
      "mean_token_accuracy": 0.16499797403812408,
      "num_tokens": 44606263.0,
      "step": 20360
    },
    {
      "entropy": 5.530084133148193,
      "epoch": 2.0211393410083365,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004594385049863102,
      "loss": 5.2138,
      "mean_token_accuracy": 0.1611294388771057,
      "num_tokens": 44617332.0,
      "step": 20365
    },
    {
      "entropy": 5.591156148910523,
      "epoch": 2.021635569670504,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00045941820913716766,
      "loss": 5.2629,
      "mean_token_accuracy": 0.16769322007894516,
      "num_tokens": 44629784.0,
      "step": 20370
    },
    {
      "entropy": 5.59921236038208,
      "epoch": 2.022131798332672,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045939790871485093,
      "loss": 5.3295,
      "mean_token_accuracy": 0.15286279171705247,
      "num_tokens": 44641398.0,
      "step": 20375
    },
    {
      "entropy": 5.486858272552491,
      "epoch": 2.0226280269948393,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00045937760371986334,
      "loss": 5.1017,
      "mean_token_accuracy": 0.17234388142824172,
      "num_tokens": 44652592.0,
      "step": 20380
    },
    {
      "entropy": 5.572465801239014,
      "epoch": 2.023124255657007,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00045935729415270857,
      "loss": 5.173,
      "mean_token_accuracy": 0.1743335708975792,
      "num_tokens": 44663896.0,
      "step": 20385
    },
    {
      "entropy": 5.657465553283691,
      "epoch": 2.0236204843191743,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004593369800138901,
      "loss": 5.31,
      "mean_token_accuracy": 0.1547352910041809,
      "num_tokens": 44675004.0,
      "step": 20390
    },
    {
      "entropy": 5.587868785858154,
      "epoch": 2.0241167129813418,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004593166613039118,
      "loss": 5.2531,
      "mean_token_accuracy": 0.1637282118201256,
      "num_tokens": 44686778.0,
      "step": 20395
    },
    {
      "entropy": 5.484924840927124,
      "epoch": 2.0246129416435092,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045929633802327743,
      "loss": 5.1964,
      "mean_token_accuracy": 0.1675669014453888,
      "num_tokens": 44697600.0,
      "step": 20400
    },
    {
      "entropy": 5.491256952285767,
      "epoch": 2.0251091703056767,
      "grad_norm": 1.390625,
      "learning_rate": 0.000459276010172491,
      "loss": 5.2454,
      "mean_token_accuracy": 0.15665214359760285,
      "num_tokens": 44707453.0,
      "step": 20405
    },
    {
      "entropy": 5.562164258956909,
      "epoch": 2.025605398967844,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045925567775205656,
      "loss": 5.3039,
      "mean_token_accuracy": 0.15227348133921623,
      "num_tokens": 44718351.0,
      "step": 20410
    },
    {
      "entropy": 5.576249408721924,
      "epoch": 2.026101627630012,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045923534076247846,
      "loss": 5.162,
      "mean_token_accuracy": 0.16872986555099487,
      "num_tokens": 44728884.0,
      "step": 20415
    },
    {
      "entropy": 5.55685510635376,
      "epoch": 2.0265978562921796,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045921499920426083,
      "loss": 5.1934,
      "mean_token_accuracy": 0.1638472631573677,
      "num_tokens": 44740769.0,
      "step": 20420
    },
    {
      "entropy": 5.563490629196167,
      "epoch": 2.027094084954347,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045919465307790817,
      "loss": 5.3262,
      "mean_token_accuracy": 0.15794071108102797,
      "num_tokens": 44753094.0,
      "step": 20425
    },
    {
      "entropy": 5.440780782699585,
      "epoch": 2.0275903136165145,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045917430238392505,
      "loss": 5.0728,
      "mean_token_accuracy": 0.16766185462474822,
      "num_tokens": 44762996.0,
      "step": 20430
    },
    {
      "entropy": 5.490050458908081,
      "epoch": 2.028086542278682,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000459153947122816,
      "loss": 5.2387,
      "mean_token_accuracy": 0.16412266939878464,
      "num_tokens": 44775343.0,
      "step": 20435
    },
    {
      "entropy": 5.501114368438721,
      "epoch": 2.0285827709408495,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004591335872950859,
      "loss": 5.1046,
      "mean_token_accuracy": 0.16950987726449968,
      "num_tokens": 44785383.0,
      "step": 20440
    },
    {
      "entropy": 5.539332103729248,
      "epoch": 2.029078999603017,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004591132229012397,
      "loss": 5.1706,
      "mean_token_accuracy": 0.16302753239870071,
      "num_tokens": 44796805.0,
      "step": 20445
    },
    {
      "entropy": 5.541423368453979,
      "epoch": 2.0295752282651844,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045909285394178217,
      "loss": 5.1702,
      "mean_token_accuracy": 0.16544023156166077,
      "num_tokens": 44806800.0,
      "step": 20450
    },
    {
      "entropy": 5.561867761611938,
      "epoch": 2.030071456927352,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004590724804172187,
      "loss": 5.3078,
      "mean_token_accuracy": 0.15559129267930985,
      "num_tokens": 44818745.0,
      "step": 20455
    },
    {
      "entropy": 5.474311780929566,
      "epoch": 2.03056768558952,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004590521023280543,
      "loss": 5.1525,
      "mean_token_accuracy": 0.1628002017736435,
      "num_tokens": 44829825.0,
      "step": 20460
    },
    {
      "entropy": 5.488186597824097,
      "epoch": 2.0310639142516873,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045903171967479425,
      "loss": 5.1622,
      "mean_token_accuracy": 0.1672174945473671,
      "num_tokens": 44841742.0,
      "step": 20465
    },
    {
      "entropy": 5.537072038650512,
      "epoch": 2.031560142913855,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045901133245794426,
      "loss": 5.2758,
      "mean_token_accuracy": 0.1610562339425087,
      "num_tokens": 44853019.0,
      "step": 20470
    },
    {
      "entropy": 5.528987884521484,
      "epoch": 2.0320563715760223,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004589909406780096,
      "loss": 5.1286,
      "mean_token_accuracy": 0.16717736721038817,
      "num_tokens": 44863147.0,
      "step": 20475
    },
    {
      "entropy": 5.557825422286987,
      "epoch": 2.0325526002381897,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045897054433549617,
      "loss": 5.1908,
      "mean_token_accuracy": 0.16563311368227004,
      "num_tokens": 44873638.0,
      "step": 20480
    },
    {
      "entropy": 5.513948917388916,
      "epoch": 2.033048828900357,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045895014343090963,
      "loss": 5.206,
      "mean_token_accuracy": 0.1619486093521118,
      "num_tokens": 44885752.0,
      "step": 20485
    },
    {
      "entropy": 5.49658432006836,
      "epoch": 2.0335450575625247,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045892973796475594,
      "loss": 5.1656,
      "mean_token_accuracy": 0.16546135544776916,
      "num_tokens": 44897471.0,
      "step": 20490
    },
    {
      "entropy": 5.6036765575408936,
      "epoch": 2.034041286224692,
      "grad_norm": 1.25,
      "learning_rate": 0.000458909327937541,
      "loss": 5.2442,
      "mean_token_accuracy": 0.15526221692562103,
      "num_tokens": 44907236.0,
      "step": 20495
    },
    {
      "entropy": 5.536808109283447,
      "epoch": 2.03453751488686,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045888891334977105,
      "loss": 5.1495,
      "mean_token_accuracy": 0.1700774297118187,
      "num_tokens": 44918103.0,
      "step": 20500
    },
    {
      "entropy": 5.457612419128418,
      "epoch": 2.0350337435490276,
      "grad_norm": 1.25,
      "learning_rate": 0.0004588684942019524,
      "loss": 5.2052,
      "mean_token_accuracy": 0.16572725474834443,
      "num_tokens": 44928263.0,
      "step": 20505
    },
    {
      "entropy": 5.51563835144043,
      "epoch": 2.035529972211195,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045884807049459116,
      "loss": 5.1988,
      "mean_token_accuracy": 0.16300869584083558,
      "num_tokens": 44939487.0,
      "step": 20510
    },
    {
      "entropy": 5.572490310668945,
      "epoch": 2.0360262008733625,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000458827642228194,
      "loss": 5.2744,
      "mean_token_accuracy": 0.1586801141500473,
      "num_tokens": 44949749.0,
      "step": 20515
    },
    {
      "entropy": 5.55518593788147,
      "epoch": 2.03652242953553,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045880720940326744,
      "loss": 5.2019,
      "mean_token_accuracy": 0.1622742995619774,
      "num_tokens": 44960347.0,
      "step": 20520
    },
    {
      "entropy": 5.604697561264038,
      "epoch": 2.0370186581976975,
      "grad_norm": 1.375,
      "learning_rate": 0.00045878677202031804,
      "loss": 5.2653,
      "mean_token_accuracy": 0.15569096952676773,
      "num_tokens": 44970420.0,
      "step": 20525
    },
    {
      "entropy": 5.528841400146485,
      "epoch": 2.037514886859865,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004587663300798529,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1619316518306732,
      "num_tokens": 44980888.0,
      "step": 20530
    },
    {
      "entropy": 5.5475481986999515,
      "epoch": 2.0380111155220324,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004587458835823786,
      "loss": 5.2081,
      "mean_token_accuracy": 0.15744587033987045,
      "num_tokens": 44992721.0,
      "step": 20535
    },
    {
      "entropy": 5.472412729263306,
      "epoch": 2.0385073441842,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045872543252840236,
      "loss": 5.1665,
      "mean_token_accuracy": 0.1697659209370613,
      "num_tokens": 45003603.0,
      "step": 20540
    },
    {
      "entropy": 5.527600574493408,
      "epoch": 2.039003572846368,
      "grad_norm": 1.375,
      "learning_rate": 0.00045870497691843134,
      "loss": 5.2036,
      "mean_token_accuracy": 0.1719418928027153,
      "num_tokens": 45014864.0,
      "step": 20545
    },
    {
      "entropy": 5.549706888198853,
      "epoch": 2.0394998015085353,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004586845167529727,
      "loss": 5.2314,
      "mean_token_accuracy": 0.16088565140962602,
      "num_tokens": 45027766.0,
      "step": 20550
    },
    {
      "entropy": 5.522126340866089,
      "epoch": 2.0399960301707027,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045866405203253384,
      "loss": 5.1854,
      "mean_token_accuracy": 0.16440579146146775,
      "num_tokens": 45039524.0,
      "step": 20555
    },
    {
      "entropy": 5.542201995849609,
      "epoch": 2.04049225883287,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004586435827576223,
      "loss": 5.248,
      "mean_token_accuracy": 0.15908483266830445,
      "num_tokens": 45050848.0,
      "step": 20560
    },
    {
      "entropy": 5.533735036849976,
      "epoch": 2.0409884874950377,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004586231089287456,
      "loss": 5.1688,
      "mean_token_accuracy": 0.16072193533182144,
      "num_tokens": 45061229.0,
      "step": 20565
    },
    {
      "entropy": 5.5018878936767575,
      "epoch": 2.041484716157205,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004586026305464115,
      "loss": 5.0955,
      "mean_token_accuracy": 0.16884349882602692,
      "num_tokens": 45070466.0,
      "step": 20570
    },
    {
      "entropy": 5.526498079299927,
      "epoch": 2.0419809448193726,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045858214761112777,
      "loss": 5.2152,
      "mean_token_accuracy": 0.16152973026037215,
      "num_tokens": 45081839.0,
      "step": 20575
    },
    {
      "entropy": 5.554927349090576,
      "epoch": 2.04247717348154,
      "grad_norm": 1.46875,
      "learning_rate": 0.00045856166012340235,
      "loss": 5.2802,
      "mean_token_accuracy": 0.15758807808160782,
      "num_tokens": 45092458.0,
      "step": 20580
    },
    {
      "entropy": 5.584849834442139,
      "epoch": 2.0429734021437076,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004585411680837433,
      "loss": 5.2488,
      "mean_token_accuracy": 0.15862326174974442,
      "num_tokens": 45104629.0,
      "step": 20585
    },
    {
      "entropy": 5.551599550247192,
      "epoch": 2.0434696308058755,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045852067149265875,
      "loss": 5.1065,
      "mean_token_accuracy": 0.16229032278060912,
      "num_tokens": 45115556.0,
      "step": 20590
    },
    {
      "entropy": 5.504107189178467,
      "epoch": 2.043965859468043,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000458500170350657,
      "loss": 5.1643,
      "mean_token_accuracy": 0.1620946079492569,
      "num_tokens": 45126489.0,
      "step": 20595
    },
    {
      "entropy": 5.559096574783325,
      "epoch": 2.0444620881302105,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045847966465824654,
      "loss": 5.2936,
      "mean_token_accuracy": 0.1546989932656288,
      "num_tokens": 45138882.0,
      "step": 20600
    },
    {
      "entropy": 5.5856325149536135,
      "epoch": 2.044958316792378,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004584591544159356,
      "loss": 5.2324,
      "mean_token_accuracy": 0.15873403549194337,
      "num_tokens": 45149941.0,
      "step": 20605
    },
    {
      "entropy": 5.588236522674561,
      "epoch": 2.0454545454545454,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004584386396242331,
      "loss": 5.254,
      "mean_token_accuracy": 0.1593601018190384,
      "num_tokens": 45162267.0,
      "step": 20610
    },
    {
      "entropy": 5.557562017440796,
      "epoch": 2.045950774116713,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045841812028364756,
      "loss": 5.332,
      "mean_token_accuracy": 0.1568589389324188,
      "num_tokens": 45174444.0,
      "step": 20615
    },
    {
      "entropy": 5.543388700485229,
      "epoch": 2.0464470027788804,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045839759639468774,
      "loss": 5.1931,
      "mean_token_accuracy": 0.16252980381250381,
      "num_tokens": 45183957.0,
      "step": 20620
    },
    {
      "entropy": 5.571543550491333,
      "epoch": 2.046943231441048,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004583770679578628,
      "loss": 5.2876,
      "mean_token_accuracy": 0.16071171313524246,
      "num_tokens": 45195607.0,
      "step": 20625
    },
    {
      "entropy": 5.508136558532715,
      "epoch": 2.0474394601032158,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004583565349736818,
      "loss": 5.1733,
      "mean_token_accuracy": 0.16412870734930038,
      "num_tokens": 45205627.0,
      "step": 20630
    },
    {
      "entropy": 5.51350588798523,
      "epoch": 2.0479356887653832,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004583359974426538,
      "loss": 5.218,
      "mean_token_accuracy": 0.15771169513463973,
      "num_tokens": 45216331.0,
      "step": 20635
    },
    {
      "entropy": 5.541222286224365,
      "epoch": 2.0484319174275507,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045831545536528806,
      "loss": 5.2535,
      "mean_token_accuracy": 0.1544589191675186,
      "num_tokens": 45227826.0,
      "step": 20640
    },
    {
      "entropy": 5.540256309509277,
      "epoch": 2.048928146089718,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045829490874209404,
      "loss": 5.1664,
      "mean_token_accuracy": 0.16322572380304337,
      "num_tokens": 45238093.0,
      "step": 20645
    },
    {
      "entropy": 5.492119598388672,
      "epoch": 2.0494243747518857,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045827435757358136,
      "loss": 5.1943,
      "mean_token_accuracy": 0.1709095522761345,
      "num_tokens": 45248660.0,
      "step": 20650
    },
    {
      "entropy": 5.557211589813233,
      "epoch": 2.049920603414053,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004582538018602594,
      "loss": 5.2792,
      "mean_token_accuracy": 0.15434218645095826,
      "num_tokens": 45261059.0,
      "step": 20655
    },
    {
      "entropy": 5.543549633026123,
      "epoch": 2.0504168320762206,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004582332416026381,
      "loss": 5.1735,
      "mean_token_accuracy": 0.16620728522539138,
      "num_tokens": 45272210.0,
      "step": 20660
    },
    {
      "entropy": 5.5868370056152346,
      "epoch": 2.050913060738388,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045821267680122725,
      "loss": 5.2709,
      "mean_token_accuracy": 0.16197322458028793,
      "num_tokens": 45283488.0,
      "step": 20665
    },
    {
      "entropy": 5.5113507270812985,
      "epoch": 2.0514092894005556,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004581921074565368,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1690952315926552,
      "num_tokens": 45292824.0,
      "step": 20670
    },
    {
      "entropy": 5.481184720993042,
      "epoch": 2.0519055180627235,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004581715335690768,
      "loss": 5.1887,
      "mean_token_accuracy": 0.1644715890288353,
      "num_tokens": 45303102.0,
      "step": 20675
    },
    {
      "entropy": 5.537046909332275,
      "epoch": 2.052401746724891,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045815095513935747,
      "loss": 5.1159,
      "mean_token_accuracy": 0.16017683893442153,
      "num_tokens": 45313825.0,
      "step": 20680
    },
    {
      "entropy": 5.532138061523438,
      "epoch": 2.0528979753870584,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004581303721678891,
      "loss": 5.1448,
      "mean_token_accuracy": 0.16867225170135497,
      "num_tokens": 45325789.0,
      "step": 20685
    },
    {
      "entropy": 5.501030445098877,
      "epoch": 2.053394204049226,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004581097846551821,
      "loss": 5.2506,
      "mean_token_accuracy": 0.15569623708724975,
      "num_tokens": 45337630.0,
      "step": 20690
    },
    {
      "entropy": 5.583867073059082,
      "epoch": 2.0538904327113934,
      "grad_norm": 1.25,
      "learning_rate": 0.000458089192601747,
      "loss": 5.2713,
      "mean_token_accuracy": 0.1565987065434456,
      "num_tokens": 45349815.0,
      "step": 20695
    },
    {
      "entropy": 5.556465339660645,
      "epoch": 2.054386661373561,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045806859600809455,
      "loss": 5.1972,
      "mean_token_accuracy": 0.15700705498456954,
      "num_tokens": 45360741.0,
      "step": 20700
    },
    {
      "entropy": 5.465679788589478,
      "epoch": 2.0548828900357283,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004580479948747353,
      "loss": 5.0883,
      "mean_token_accuracy": 0.17203922420740128,
      "num_tokens": 45371258.0,
      "step": 20705
    },
    {
      "entropy": 5.521376466751098,
      "epoch": 2.055379118697896,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045802738920218015,
      "loss": 5.198,
      "mean_token_accuracy": 0.15884949862957,
      "num_tokens": 45381914.0,
      "step": 20710
    },
    {
      "entropy": 5.491108369827271,
      "epoch": 2.0558753473600637,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004580067789909403,
      "loss": 5.2072,
      "mean_token_accuracy": 0.16488956063985824,
      "num_tokens": 45393159.0,
      "step": 20715
    },
    {
      "entropy": 5.495648288726807,
      "epoch": 2.056371576022231,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045798616424152654,
      "loss": 5.092,
      "mean_token_accuracy": 0.16621140241622925,
      "num_tokens": 45403076.0,
      "step": 20720
    },
    {
      "entropy": 5.5028883934021,
      "epoch": 2.0568678046843987,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045796554495445023,
      "loss": 5.2275,
      "mean_token_accuracy": 0.16430277228355408,
      "num_tokens": 45412962.0,
      "step": 20725
    },
    {
      "entropy": 5.512819576263428,
      "epoch": 2.057364033346566,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004579449211302228,
      "loss": 5.1774,
      "mean_token_accuracy": 0.16650424003601075,
      "num_tokens": 45423620.0,
      "step": 20730
    },
    {
      "entropy": 5.515007066726684,
      "epoch": 2.0578602620087336,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004579242927693554,
      "loss": 5.2069,
      "mean_token_accuracy": 0.15541243255138398,
      "num_tokens": 45435901.0,
      "step": 20735
    },
    {
      "entropy": 5.516304349899292,
      "epoch": 2.058356490670901,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004579036598723598,
      "loss": 5.1261,
      "mean_token_accuracy": 0.17095738500356675,
      "num_tokens": 45445409.0,
      "step": 20740
    },
    {
      "entropy": 5.614623785018921,
      "epoch": 2.0588527193330686,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045788302243974745,
      "loss": 5.2781,
      "mean_token_accuracy": 0.16579465121030806,
      "num_tokens": 45455794.0,
      "step": 20745
    },
    {
      "entropy": 5.508283710479736,
      "epoch": 2.059348947995236,
      "grad_norm": 1.125,
      "learning_rate": 0.0004578623804720302,
      "loss": 5.1575,
      "mean_token_accuracy": 0.16608809381723405,
      "num_tokens": 45465922.0,
      "step": 20750
    },
    {
      "entropy": 5.49040412902832,
      "epoch": 2.0598451766574035,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004578417339697201,
      "loss": 5.2443,
      "mean_token_accuracy": 0.16000977158546448,
      "num_tokens": 45476314.0,
      "step": 20755
    },
    {
      "entropy": 5.553521633148193,
      "epoch": 2.0603414053195714,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045782108293332884,
      "loss": 5.2069,
      "mean_token_accuracy": 0.1581800699234009,
      "num_tokens": 45487232.0,
      "step": 20760
    },
    {
      "entropy": 5.6199277400970455,
      "epoch": 2.060837633981739,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004578004273633687,
      "loss": 5.2485,
      "mean_token_accuracy": 0.16299643963575364,
      "num_tokens": 45498368.0,
      "step": 20765
    },
    {
      "entropy": 5.488735008239746,
      "epoch": 2.0613338626439064,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045777976726035194,
      "loss": 5.1943,
      "mean_token_accuracy": 0.16334162205457686,
      "num_tokens": 45507600.0,
      "step": 20770
    },
    {
      "entropy": 5.554404544830322,
      "epoch": 2.061830091306074,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004577591026247907,
      "loss": 5.281,
      "mean_token_accuracy": 0.15446316003799437,
      "num_tokens": 45518594.0,
      "step": 20775
    },
    {
      "entropy": 5.547867393493652,
      "epoch": 2.0623263199682413,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004577384334571976,
      "loss": 5.1711,
      "mean_token_accuracy": 0.1568750560283661,
      "num_tokens": 45529929.0,
      "step": 20780
    },
    {
      "entropy": 5.51588191986084,
      "epoch": 2.062822548630409,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004577177597580851,
      "loss": 5.1648,
      "mean_token_accuracy": 0.16691308319568635,
      "num_tokens": 45540228.0,
      "step": 20785
    },
    {
      "entropy": 5.441077756881714,
      "epoch": 2.0633187772925763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045769708152796593,
      "loss": 5.1341,
      "mean_token_accuracy": 0.16816219538450242,
      "num_tokens": 45552026.0,
      "step": 20790
    },
    {
      "entropy": 5.501677227020264,
      "epoch": 2.0638150059547438,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045767639876735276,
      "loss": 5.2472,
      "mean_token_accuracy": 0.1628403529524803,
      "num_tokens": 45563579.0,
      "step": 20795
    },
    {
      "entropy": 5.636795377731323,
      "epoch": 2.0643112346169117,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004576557114767585,
      "loss": 5.2818,
      "mean_token_accuracy": 0.15780508369207383,
      "num_tokens": 45573800.0,
      "step": 20800
    },
    {
      "entropy": 5.525663089752197,
      "epoch": 2.064807463279079,
      "grad_norm": 1.375,
      "learning_rate": 0.00045763501965669627,
      "loss": 5.1635,
      "mean_token_accuracy": 0.16766264885663987,
      "num_tokens": 45583758.0,
      "step": 20805
    },
    {
      "entropy": 5.4944792747497555,
      "epoch": 2.0653036919412466,
      "grad_norm": 1.171875,
      "learning_rate": 0.000457614323307679,
      "loss": 5.1633,
      "mean_token_accuracy": 0.16491268575191498,
      "num_tokens": 45595154.0,
      "step": 20810
    },
    {
      "entropy": 5.63830156326294,
      "epoch": 2.065799920603414,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045759362243022,
      "loss": 5.2953,
      "mean_token_accuracy": 0.15973795354366302,
      "num_tokens": 45606557.0,
      "step": 20815
    },
    {
      "entropy": 5.486002111434937,
      "epoch": 2.0662961492655816,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004575729170248327,
      "loss": 5.104,
      "mean_token_accuracy": 0.1728888213634491,
      "num_tokens": 45615531.0,
      "step": 20820
    },
    {
      "entropy": 5.4309654712677,
      "epoch": 2.066792377927749,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004575522070920305,
      "loss": 5.1287,
      "mean_token_accuracy": 0.1654913529753685,
      "num_tokens": 45626512.0,
      "step": 20825
    },
    {
      "entropy": 5.584114027023316,
      "epoch": 2.0672886065899165,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045753149263232687,
      "loss": 5.2511,
      "mean_token_accuracy": 0.15532525032758712,
      "num_tokens": 45638005.0,
      "step": 20830
    },
    {
      "entropy": 5.555320692062378,
      "epoch": 2.067784835252084,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004575107736462355,
      "loss": 5.2246,
      "mean_token_accuracy": 0.15916720032691956,
      "num_tokens": 45647740.0,
      "step": 20835
    },
    {
      "entropy": 5.492320251464844,
      "epoch": 2.0682810639142515,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004574900501342703,
      "loss": 5.1679,
      "mean_token_accuracy": 0.16960099041461946,
      "num_tokens": 45659021.0,
      "step": 20840
    },
    {
      "entropy": 5.5014458179473875,
      "epoch": 2.0687772925764194,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004574693220969451,
      "loss": 5.2383,
      "mean_token_accuracy": 0.15968297570943832,
      "num_tokens": 45669413.0,
      "step": 20845
    },
    {
      "entropy": 5.594769716262817,
      "epoch": 2.069273521238587,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004574485895347738,
      "loss": 5.2979,
      "mean_token_accuracy": 0.16021818816661834,
      "num_tokens": 45680719.0,
      "step": 20850
    },
    {
      "entropy": 5.488606309890747,
      "epoch": 2.0697697499007544,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045742785244827066,
      "loss": 5.1203,
      "mean_token_accuracy": 0.16773884296417235,
      "num_tokens": 45690393.0,
      "step": 20855
    },
    {
      "entropy": 5.535280990600586,
      "epoch": 2.070265978562922,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004574071108379499,
      "loss": 5.2543,
      "mean_token_accuracy": 0.16247047930955888,
      "num_tokens": 45701304.0,
      "step": 20860
    },
    {
      "entropy": 5.536807870864868,
      "epoch": 2.0707622072250893,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004573863647043259,
      "loss": 5.0947,
      "mean_token_accuracy": 0.16732445508241653,
      "num_tokens": 45711892.0,
      "step": 20865
    },
    {
      "entropy": 5.488884115219117,
      "epoch": 2.0712584358872568,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004573656140479129,
      "loss": 5.1956,
      "mean_token_accuracy": 0.16110664010047912,
      "num_tokens": 45723693.0,
      "step": 20870
    },
    {
      "entropy": 5.452740716934204,
      "epoch": 2.0717546645494243,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004573448588692258,
      "loss": 5.1535,
      "mean_token_accuracy": 0.16377484798431396,
      "num_tokens": 45734392.0,
      "step": 20875
    },
    {
      "entropy": 5.44530930519104,
      "epoch": 2.0722508932115917,
      "grad_norm": 1.21875,
      "learning_rate": 0.000457324099168779,
      "loss": 5.0946,
      "mean_token_accuracy": 0.16803934723138808,
      "num_tokens": 45745732.0,
      "step": 20880
    },
    {
      "entropy": 5.452918195724488,
      "epoch": 2.0727471218737596,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004573033349470874,
      "loss": 5.1642,
      "mean_token_accuracy": 0.16819653809070587,
      "num_tokens": 45756671.0,
      "step": 20885
    },
    {
      "entropy": 5.57900333404541,
      "epoch": 2.073243350535927,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004572825662046661,
      "loss": 5.2488,
      "mean_token_accuracy": 0.15882737934589386,
      "num_tokens": 45767685.0,
      "step": 20890
    },
    {
      "entropy": 5.597635698318482,
      "epoch": 2.0737395791980946,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004572617929420298,
      "loss": 5.1619,
      "mean_token_accuracy": 0.16326408684253693,
      "num_tokens": 45777251.0,
      "step": 20895
    },
    {
      "entropy": 5.546514320373535,
      "epoch": 2.074235807860262,
      "grad_norm": 1.453125,
      "learning_rate": 0.00045724101515969375,
      "loss": 5.2524,
      "mean_token_accuracy": 0.1614864856004715,
      "num_tokens": 45790037.0,
      "step": 20900
    },
    {
      "entropy": 5.537281131744384,
      "epoch": 2.0747320365224295,
      "grad_norm": 1.984375,
      "learning_rate": 0.00045722023285817327,
      "loss": 5.2847,
      "mean_token_accuracy": 0.15543804913759232,
      "num_tokens": 45800803.0,
      "step": 20905
    },
    {
      "entropy": 5.616736125946045,
      "epoch": 2.075228265184597,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004571994460379837,
      "loss": 5.2335,
      "mean_token_accuracy": 0.151889069378376,
      "num_tokens": 45812447.0,
      "step": 20910
    },
    {
      "entropy": 5.532973527908325,
      "epoch": 2.0757244938467645,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045717865469964045,
      "loss": 5.1763,
      "mean_token_accuracy": 0.1602273553609848,
      "num_tokens": 45822172.0,
      "step": 20915
    },
    {
      "entropy": 5.586142301559448,
      "epoch": 2.076220722508932,
      "grad_norm": 1.46875,
      "learning_rate": 0.00045715785884365914,
      "loss": 5.3152,
      "mean_token_accuracy": 0.15965129062533379,
      "num_tokens": 45832583.0,
      "step": 20920
    },
    {
      "entropy": 5.626015090942383,
      "epoch": 2.0767169511710994,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045713705847055537,
      "loss": 5.3201,
      "mean_token_accuracy": 0.1588379755616188,
      "num_tokens": 45842640.0,
      "step": 20925
    },
    {
      "entropy": 5.605521583557129,
      "epoch": 2.0772131798332674,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004571162535808451,
      "loss": 5.3142,
      "mean_token_accuracy": 0.1557700365781784,
      "num_tokens": 45854618.0,
      "step": 20930
    },
    {
      "entropy": 5.50201416015625,
      "epoch": 2.077709408495435,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004570954441750441,
      "loss": 5.1942,
      "mean_token_accuracy": 0.16383085697889327,
      "num_tokens": 45866130.0,
      "step": 20935
    },
    {
      "entropy": 5.545937824249267,
      "epoch": 2.0782056371576023,
      "grad_norm": 1.125,
      "learning_rate": 0.0004570746302536685,
      "loss": 5.2106,
      "mean_token_accuracy": 0.1592772200703621,
      "num_tokens": 45877187.0,
      "step": 20940
    },
    {
      "entropy": 5.534173440933228,
      "epoch": 2.07870186581977,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045705381181723444,
      "loss": 5.2677,
      "mean_token_accuracy": 0.16101457476615905,
      "num_tokens": 45889557.0,
      "step": 20945
    },
    {
      "entropy": 5.55206618309021,
      "epoch": 2.0791980944819373,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045703298886625804,
      "loss": 5.2373,
      "mean_token_accuracy": 0.15666287988424302,
      "num_tokens": 45900860.0,
      "step": 20950
    },
    {
      "entropy": 5.539902210235596,
      "epoch": 2.0796943231441047,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004570121614012558,
      "loss": 5.2313,
      "mean_token_accuracy": 0.16480804681777955,
      "num_tokens": 45912231.0,
      "step": 20955
    },
    {
      "entropy": 5.576876878738403,
      "epoch": 2.080190551806272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045699132942274425,
      "loss": 5.2963,
      "mean_token_accuracy": 0.15818197503685952,
      "num_tokens": 45922722.0,
      "step": 20960
    },
    {
      "entropy": 5.476846647262573,
      "epoch": 2.0806867804684397,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045697049293123984,
      "loss": 5.1791,
      "mean_token_accuracy": 0.16387999653816224,
      "num_tokens": 45932444.0,
      "step": 20965
    },
    {
      "entropy": 5.549612522125244,
      "epoch": 2.0811830091306076,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004569496519272593,
      "loss": 5.207,
      "mean_token_accuracy": 0.1610811397433281,
      "num_tokens": 45942826.0,
      "step": 20970
    },
    {
      "entropy": 5.568046188354492,
      "epoch": 2.081679237792775,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045692880641131945,
      "loss": 5.2142,
      "mean_token_accuracy": 0.1639574110507965,
      "num_tokens": 45953935.0,
      "step": 20975
    },
    {
      "entropy": 5.570403909683227,
      "epoch": 2.0821754664549426,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004569079563839372,
      "loss": 5.2921,
      "mean_token_accuracy": 0.15010600984096528,
      "num_tokens": 45964693.0,
      "step": 20980
    },
    {
      "entropy": 5.593636798858642,
      "epoch": 2.08267169511711,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045688710184562965,
      "loss": 5.2185,
      "mean_token_accuracy": 0.15937648266553878,
      "num_tokens": 45976481.0,
      "step": 20985
    },
    {
      "entropy": 5.587125015258789,
      "epoch": 2.0831679237792775,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045686624279691387,
      "loss": 5.3015,
      "mean_token_accuracy": 0.1583019584417343,
      "num_tokens": 45988879.0,
      "step": 20990
    },
    {
      "entropy": 5.4982716083526615,
      "epoch": 2.083664152441445,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004568453792383072,
      "loss": 5.1643,
      "mean_token_accuracy": 0.1689223200082779,
      "num_tokens": 45999332.0,
      "step": 20995
    },
    {
      "entropy": 5.543160724639892,
      "epoch": 2.0841603811036125,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000456824511170327,
      "loss": 5.2112,
      "mean_token_accuracy": 0.16398333609104157,
      "num_tokens": 46009841.0,
      "step": 21000
    },
    {
      "epoch": 2.0841603811036125,
      "eval_entropy": 5.381959172432596,
      "eval_loss": 5.365278720855713,
      "eval_mean_token_accuracy": 0.16085563439997097,
      "eval_num_tokens": 46009841.0,
      "eval_runtime": 26.5058,
      "eval_samples_per_second": 1270.249,
      "eval_steps_per_second": 158.795,
      "step": 21000
    },
    {
      "entropy": 5.488643646240234,
      "epoch": 2.08465660976578,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004568036385934906,
      "loss": 5.1413,
      "mean_token_accuracy": 0.15987980365753174,
      "num_tokens": 46020678.0,
      "step": 21005
    },
    {
      "entropy": 5.534518384933472,
      "epoch": 2.0851528384279474,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045678276150831586,
      "loss": 5.2131,
      "mean_token_accuracy": 0.1616602897644043,
      "num_tokens": 46032036.0,
      "step": 21010
    },
    {
      "entropy": 5.606665515899659,
      "epoch": 2.0856490670901153,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004567618799153202,
      "loss": 5.227,
      "mean_token_accuracy": 0.16197765618562698,
      "num_tokens": 46042675.0,
      "step": 21015
    },
    {
      "entropy": 5.47002763748169,
      "epoch": 2.086145295752283,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045674099381502167,
      "loss": 5.2183,
      "mean_token_accuracy": 0.16180895566940307,
      "num_tokens": 46053616.0,
      "step": 21020
    },
    {
      "entropy": 5.534269189834594,
      "epoch": 2.0866415244144503,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004567201032079381,
      "loss": 5.246,
      "mean_token_accuracy": 0.16192087158560753,
      "num_tokens": 46063755.0,
      "step": 21025
    },
    {
      "entropy": 5.578156852722168,
      "epoch": 2.0871377530766178,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045669920809458754,
      "loss": 5.2282,
      "mean_token_accuracy": 0.16132556647062302,
      "num_tokens": 46074459.0,
      "step": 21030
    },
    {
      "entropy": 5.655478477478027,
      "epoch": 2.0876339817387852,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004566783084754881,
      "loss": 5.3059,
      "mean_token_accuracy": 0.15718813836574555,
      "num_tokens": 46085053.0,
      "step": 21035
    },
    {
      "entropy": 5.462538385391236,
      "epoch": 2.0881302104009527,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045665740435115815,
      "loss": 5.1065,
      "mean_token_accuracy": 0.17380682826042176,
      "num_tokens": 46094697.0,
      "step": 21040
    },
    {
      "entropy": 5.476847505569458,
      "epoch": 2.08862643906312,
      "grad_norm": 1.34375,
      "learning_rate": 0.000456636495722116,
      "loss": 5.2493,
      "mean_token_accuracy": 0.15844349563121796,
      "num_tokens": 46105617.0,
      "step": 21045
    },
    {
      "entropy": 5.584053087234497,
      "epoch": 2.0891226677252877,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004566155825888801,
      "loss": 5.215,
      "mean_token_accuracy": 0.16283852010965347,
      "num_tokens": 46115450.0,
      "step": 21050
    },
    {
      "entropy": 5.5484058380126955,
      "epoch": 2.089618896387455,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004565946649519691,
      "loss": 5.2667,
      "mean_token_accuracy": 0.15976429283618926,
      "num_tokens": 46126511.0,
      "step": 21055
    },
    {
      "entropy": 5.518594408035279,
      "epoch": 2.090115125049623,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00045657374281190177,
      "loss": 5.1599,
      "mean_token_accuracy": 0.16629549711942673,
      "num_tokens": 46136294.0,
      "step": 21060
    },
    {
      "entropy": 5.485678815841675,
      "epoch": 2.0906113537117905,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004565528161691969,
      "loss": 5.2406,
      "mean_token_accuracy": 0.15420478731393814,
      "num_tokens": 46148926.0,
      "step": 21065
    },
    {
      "entropy": 5.588468408584594,
      "epoch": 2.091107582373958,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004565318850243733,
      "loss": 5.2378,
      "mean_token_accuracy": 0.16202940791845322,
      "num_tokens": 46160130.0,
      "step": 21070
    },
    {
      "entropy": 5.519721221923828,
      "epoch": 2.0916038110361255,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004565109493779502,
      "loss": 5.1607,
      "mean_token_accuracy": 0.16774796396493913,
      "num_tokens": 46169919.0,
      "step": 21075
    },
    {
      "entropy": 5.556796407699585,
      "epoch": 2.092100039698293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045649000923044664,
      "loss": 5.2625,
      "mean_token_accuracy": 0.15870760679244994,
      "num_tokens": 46181186.0,
      "step": 21080
    },
    {
      "entropy": 5.513354110717773,
      "epoch": 2.0925962683604604,
      "grad_norm": 1.25,
      "learning_rate": 0.00045646906458238196,
      "loss": 5.231,
      "mean_token_accuracy": 0.15894316285848617,
      "num_tokens": 46192172.0,
      "step": 21085
    },
    {
      "entropy": 5.529216480255127,
      "epoch": 2.093092497022628,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004564481154342754,
      "loss": 5.2281,
      "mean_token_accuracy": 0.15265561789274215,
      "num_tokens": 46202678.0,
      "step": 21090
    },
    {
      "entropy": 5.558067893981933,
      "epoch": 2.0935887256847954,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004564271617866467,
      "loss": 5.2221,
      "mean_token_accuracy": 0.16110694855451585,
      "num_tokens": 46212213.0,
      "step": 21095
    },
    {
      "entropy": 5.6192930221557615,
      "epoch": 2.0940849543469633,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045640620364001524,
      "loss": 5.2679,
      "mean_token_accuracy": 0.15876856744289397,
      "num_tokens": 46223044.0,
      "step": 21100
    },
    {
      "entropy": 5.609292459487915,
      "epoch": 2.0945811830091308,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045638524099490086,
      "loss": 5.2676,
      "mean_token_accuracy": 0.15792862921953202,
      "num_tokens": 46236008.0,
      "step": 21105
    },
    {
      "entropy": 5.504161548614502,
      "epoch": 2.0950774116712982,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045636427385182334,
      "loss": 5.2174,
      "mean_token_accuracy": 0.16276352107524872,
      "num_tokens": 46247631.0,
      "step": 21110
    },
    {
      "entropy": 5.514677095413208,
      "epoch": 2.0955736403334657,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045634330221130265,
      "loss": 5.1694,
      "mean_token_accuracy": 0.16236639767885208,
      "num_tokens": 46258294.0,
      "step": 21115
    },
    {
      "entropy": 5.5268806457519535,
      "epoch": 2.096069868995633,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004563223260738589,
      "loss": 5.1864,
      "mean_token_accuracy": 0.1579926997423172,
      "num_tokens": 46269611.0,
      "step": 21120
    },
    {
      "entropy": 5.532867956161499,
      "epoch": 2.0965660976578007,
      "grad_norm": 1.203125,
      "learning_rate": 0.000456301345440012,
      "loss": 5.2739,
      "mean_token_accuracy": 0.167898927628994,
      "num_tokens": 46280731.0,
      "step": 21125
    },
    {
      "entropy": 5.537222146987915,
      "epoch": 2.097062326319968,
      "grad_norm": 1.25,
      "learning_rate": 0.0004562803603102825,
      "loss": 5.1856,
      "mean_token_accuracy": 0.15862118750810622,
      "num_tokens": 46290226.0,
      "step": 21130
    },
    {
      "entropy": 5.579013824462891,
      "epoch": 2.0975585549821356,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045625937068519075,
      "loss": 5.1918,
      "mean_token_accuracy": 0.16458752006292343,
      "num_tokens": 46300388.0,
      "step": 21135
    },
    {
      "entropy": 5.52675085067749,
      "epoch": 2.098054783644303,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004562383765652571,
      "loss": 5.2658,
      "mean_token_accuracy": 0.1637671858072281,
      "num_tokens": 46312214.0,
      "step": 21140
    },
    {
      "entropy": 5.539980554580689,
      "epoch": 2.098551012306471,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004562173779510023,
      "loss": 5.2822,
      "mean_token_accuracy": 0.1559859037399292,
      "num_tokens": 46322728.0,
      "step": 21145
    },
    {
      "entropy": 5.520312643051147,
      "epoch": 2.0990472409686385,
      "grad_norm": 1.15625,
      "learning_rate": 0.000456196374842947,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1690568670630455,
      "num_tokens": 46334169.0,
      "step": 21150
    },
    {
      "entropy": 5.582124280929565,
      "epoch": 2.099543469630806,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045617536724161203,
      "loss": 5.1682,
      "mean_token_accuracy": 0.1611720710992813,
      "num_tokens": 46344601.0,
      "step": 21155
    },
    {
      "entropy": 5.526064538955689,
      "epoch": 2.1000396982929734,
      "grad_norm": 1.25,
      "learning_rate": 0.00045615435514751834,
      "loss": 5.2531,
      "mean_token_accuracy": 0.15899643450975418,
      "num_tokens": 46355633.0,
      "step": 21160
    },
    {
      "entropy": 5.5381440162658695,
      "epoch": 2.100535926955141,
      "grad_norm": 1.09375,
      "learning_rate": 0.000456133338561187,
      "loss": 5.288,
      "mean_token_accuracy": 0.15262553989887237,
      "num_tokens": 46365981.0,
      "step": 21165
    },
    {
      "entropy": 5.558678913116455,
      "epoch": 2.1010321556173084,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004561123174831391,
      "loss": 5.2623,
      "mean_token_accuracy": 0.16583374366164208,
      "num_tokens": 46376664.0,
      "step": 21170
    },
    {
      "entropy": 5.558452796936035,
      "epoch": 2.101528384279476,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004560912919138962,
      "loss": 5.1538,
      "mean_token_accuracy": 0.16957743167877198,
      "num_tokens": 46387918.0,
      "step": 21175
    },
    {
      "entropy": 5.615236473083496,
      "epoch": 2.1020246129416433,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045607026185397924,
      "loss": 5.2083,
      "mean_token_accuracy": 0.16338768750429153,
      "num_tokens": 46398660.0,
      "step": 21180
    },
    {
      "entropy": 5.516957521438599,
      "epoch": 2.102520841603811,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004560492273039101,
      "loss": 5.1489,
      "mean_token_accuracy": 0.16603241711854935,
      "num_tokens": 46409365.0,
      "step": 21185
    },
    {
      "entropy": 5.496510553359985,
      "epoch": 2.1030170702659787,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045602818826421024,
      "loss": 5.1517,
      "mean_token_accuracy": 0.169710723310709,
      "num_tokens": 46419457.0,
      "step": 21190
    },
    {
      "entropy": 5.54266939163208,
      "epoch": 2.103513298928146,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004560071447354013,
      "loss": 5.2383,
      "mean_token_accuracy": 0.16801748275756836,
      "num_tokens": 46430189.0,
      "step": 21195
    },
    {
      "entropy": 5.466357040405273,
      "epoch": 2.1040095275903137,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004559860967180053,
      "loss": 5.1962,
      "mean_token_accuracy": 0.16440754532814025,
      "num_tokens": 46440555.0,
      "step": 21200
    },
    {
      "entropy": 5.597494125366211,
      "epoch": 2.104505756252481,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004559650442125441,
      "loss": 5.2277,
      "mean_token_accuracy": 0.15916885882616044,
      "num_tokens": 46451582.0,
      "step": 21205
    },
    {
      "entropy": 5.654547071456909,
      "epoch": 2.1050019849146486,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004559439872195397,
      "loss": 5.2293,
      "mean_token_accuracy": 0.16196174621582032,
      "num_tokens": 46463614.0,
      "step": 21210
    },
    {
      "entropy": 5.556007766723633,
      "epoch": 2.105498213576816,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004559229257395144,
      "loss": 5.2411,
      "mean_token_accuracy": 0.15803247392177583,
      "num_tokens": 46474279.0,
      "step": 21215
    },
    {
      "entropy": 5.579813480377197,
      "epoch": 2.1059944422389836,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045590185977299025,
      "loss": 5.3486,
      "mean_token_accuracy": 0.15755492597818374,
      "num_tokens": 46485531.0,
      "step": 21220
    },
    {
      "entropy": 5.508579111099243,
      "epoch": 2.106490670901151,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004558807893204899,
      "loss": 5.1665,
      "mean_token_accuracy": 0.17068906724452973,
      "num_tokens": 46496197.0,
      "step": 21225
    },
    {
      "entropy": 5.527390480041504,
      "epoch": 2.106986899563319,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004558597143825357,
      "loss": 5.2413,
      "mean_token_accuracy": 0.16019395589828492,
      "num_tokens": 46506759.0,
      "step": 21230
    },
    {
      "entropy": 5.540473794937133,
      "epoch": 2.1074831282254864,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004558386349596503,
      "loss": 5.2421,
      "mean_token_accuracy": 0.16243263632059096,
      "num_tokens": 46518956.0,
      "step": 21235
    },
    {
      "entropy": 5.624525690078736,
      "epoch": 2.107979356887654,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004558175510523564,
      "loss": 5.3041,
      "mean_token_accuracy": 0.16450520604848862,
      "num_tokens": 46530109.0,
      "step": 21240
    },
    {
      "entropy": 5.602064847946167,
      "epoch": 2.1084755855498214,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045579646266117696,
      "loss": 5.1707,
      "mean_token_accuracy": 0.15978098809719085,
      "num_tokens": 46542206.0,
      "step": 21245
    },
    {
      "entropy": 5.585670757293701,
      "epoch": 2.108971814211989,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045577536978663475,
      "loss": 5.3443,
      "mean_token_accuracy": 0.16080426946282386,
      "num_tokens": 46554143.0,
      "step": 21250
    },
    {
      "entropy": 5.4875880718231205,
      "epoch": 2.1094680428741563,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004557542724292529,
      "loss": 5.1513,
      "mean_token_accuracy": 0.16471897810697556,
      "num_tokens": 46565697.0,
      "step": 21255
    },
    {
      "entropy": 5.55903697013855,
      "epoch": 2.109964271536324,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004557331705895545,
      "loss": 5.2243,
      "mean_token_accuracy": 0.15747759789228438,
      "num_tokens": 46576358.0,
      "step": 21260
    },
    {
      "entropy": 5.553475046157837,
      "epoch": 2.1104605001984913,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004557120642680631,
      "loss": 5.2583,
      "mean_token_accuracy": 0.15618784576654435,
      "num_tokens": 46589422.0,
      "step": 21265
    },
    {
      "entropy": 5.517782258987427,
      "epoch": 2.1109567288606588,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004556909534653018,
      "loss": 5.2266,
      "mean_token_accuracy": 0.16710186749696732,
      "num_tokens": 46599623.0,
      "step": 21270
    },
    {
      "entropy": 5.536098909378052,
      "epoch": 2.1114529575228267,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004556698381817942,
      "loss": 5.1485,
      "mean_token_accuracy": 0.1684466689825058,
      "num_tokens": 46609321.0,
      "step": 21275
    },
    {
      "entropy": 5.482399654388428,
      "epoch": 2.111949186184994,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000455648718418064,
      "loss": 5.22,
      "mean_token_accuracy": 0.15936898738145827,
      "num_tokens": 46620265.0,
      "step": 21280
    },
    {
      "entropy": 5.543705940246582,
      "epoch": 2.1124454148471616,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004556275941746348,
      "loss": 5.2419,
      "mean_token_accuracy": 0.1634519502520561,
      "num_tokens": 46630776.0,
      "step": 21285
    },
    {
      "entropy": 5.46487021446228,
      "epoch": 2.112941643509329,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045560646545203044,
      "loss": 5.1497,
      "mean_token_accuracy": 0.16810588389635087,
      "num_tokens": 46641382.0,
      "step": 21290
    },
    {
      "entropy": 5.50220422744751,
      "epoch": 2.1134378721714966,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045558533225077496,
      "loss": 5.1407,
      "mean_token_accuracy": 0.16555486172437667,
      "num_tokens": 46653589.0,
      "step": 21295
    },
    {
      "entropy": 5.513012409210205,
      "epoch": 2.113934100833664,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045556419457139235,
      "loss": 5.1845,
      "mean_token_accuracy": 0.1680415764451027,
      "num_tokens": 46664595.0,
      "step": 21300
    },
    {
      "entropy": 5.576735782623291,
      "epoch": 2.1144303294958315,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045554305241440684,
      "loss": 5.2176,
      "mean_token_accuracy": 0.16234654188156128,
      "num_tokens": 46675453.0,
      "step": 21305
    },
    {
      "entropy": 5.475623846054077,
      "epoch": 2.114926558157999,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004555219057803426,
      "loss": 5.1171,
      "mean_token_accuracy": 0.1705949991941452,
      "num_tokens": 46685668.0,
      "step": 21310
    },
    {
      "entropy": 5.535132312774659,
      "epoch": 2.115422786820167,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004555007546697242,
      "loss": 5.227,
      "mean_token_accuracy": 0.15985895693302155,
      "num_tokens": 46696144.0,
      "step": 21315
    },
    {
      "entropy": 5.550120687484741,
      "epoch": 2.1159190154823344,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045547959908307585,
      "loss": 5.2162,
      "mean_token_accuracy": 0.1597440555691719,
      "num_tokens": 46707892.0,
      "step": 21320
    },
    {
      "entropy": 5.557758760452271,
      "epoch": 2.116415244144502,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004554584390209225,
      "loss": 5.2336,
      "mean_token_accuracy": 0.16337419003248216,
      "num_tokens": 46721099.0,
      "step": 21325
    },
    {
      "entropy": 5.574937582015991,
      "epoch": 2.1169114728066694,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004554372744837886,
      "loss": 5.2452,
      "mean_token_accuracy": 0.157948100566864,
      "num_tokens": 46732974.0,
      "step": 21330
    },
    {
      "entropy": 5.551499557495117,
      "epoch": 2.117407701468837,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004554161054721992,
      "loss": 5.2708,
      "mean_token_accuracy": 0.16198687851428986,
      "num_tokens": 46743148.0,
      "step": 21335
    },
    {
      "entropy": 5.542793416976929,
      "epoch": 2.1179039301310043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004553949319866791,
      "loss": 5.1863,
      "mean_token_accuracy": 0.16880941540002822,
      "num_tokens": 46754214.0,
      "step": 21340
    },
    {
      "entropy": 5.527360486984253,
      "epoch": 2.118400158793172,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045537375402775334,
      "loss": 5.2104,
      "mean_token_accuracy": 0.15876018404960632,
      "num_tokens": 46764112.0,
      "step": 21345
    },
    {
      "entropy": 5.5020825386047365,
      "epoch": 2.1188963874553393,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045535257159594724,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1700740173459053,
      "num_tokens": 46775202.0,
      "step": 21350
    },
    {
      "entropy": 5.640245389938355,
      "epoch": 2.1193926161175067,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000455331384691786,
      "loss": 5.2197,
      "mean_token_accuracy": 0.163314650952816,
      "num_tokens": 46786557.0,
      "step": 21355
    },
    {
      "entropy": 5.543274354934693,
      "epoch": 2.1198888447796747,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004553101933157949,
      "loss": 5.1807,
      "mean_token_accuracy": 0.1572815239429474,
      "num_tokens": 46797082.0,
      "step": 21360
    },
    {
      "entropy": 5.4770409107208256,
      "epoch": 2.120385073441842,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045528899746849965,
      "loss": 5.1952,
      "mean_token_accuracy": 0.1660243973135948,
      "num_tokens": 46806679.0,
      "step": 21365
    },
    {
      "entropy": 5.603643941879272,
      "epoch": 2.1208813021040096,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004552677971504257,
      "loss": 5.3417,
      "mean_token_accuracy": 0.15247157365083694,
      "num_tokens": 46818169.0,
      "step": 21370
    },
    {
      "entropy": 5.708225059509277,
      "epoch": 2.121377530766177,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004552465923620988,
      "loss": 5.3824,
      "mean_token_accuracy": 0.15579296350479127,
      "num_tokens": 46829552.0,
      "step": 21375
    },
    {
      "entropy": 5.61361608505249,
      "epoch": 2.1218737594283446,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004552253831040448,
      "loss": 5.2683,
      "mean_token_accuracy": 0.16531425714492798,
      "num_tokens": 46840808.0,
      "step": 21380
    },
    {
      "entropy": 5.554231643676758,
      "epoch": 2.122369988090512,
      "grad_norm": 1.125,
      "learning_rate": 0.00045520416937678973,
      "loss": 5.2169,
      "mean_token_accuracy": 0.15674423426389694,
      "num_tokens": 46852066.0,
      "step": 21385
    },
    {
      "entropy": 5.49283995628357,
      "epoch": 2.1228662167526795,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045518295118085955,
      "loss": 5.1544,
      "mean_token_accuracy": 0.16174201294779778,
      "num_tokens": 46863163.0,
      "step": 21390
    },
    {
      "entropy": 5.545247554779053,
      "epoch": 2.123362445414847,
      "grad_norm": 1.40625,
      "learning_rate": 0.00045516172851678036,
      "loss": 5.2073,
      "mean_token_accuracy": 0.1601942151784897,
      "num_tokens": 46873246.0,
      "step": 21395
    },
    {
      "entropy": 5.503837728500367,
      "epoch": 2.123858674077015,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045514050138507857,
      "loss": 5.2254,
      "mean_token_accuracy": 0.1627745360136032,
      "num_tokens": 46883881.0,
      "step": 21400
    },
    {
      "entropy": 5.565494108200073,
      "epoch": 2.1243549027391824,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004551192697862805,
      "loss": 5.2528,
      "mean_token_accuracy": 0.16093432903289795,
      "num_tokens": 46894907.0,
      "step": 21405
    },
    {
      "entropy": 5.489614582061767,
      "epoch": 2.12485113140135,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004550980337209127,
      "loss": 5.1456,
      "mean_token_accuracy": 0.16111299991607667,
      "num_tokens": 46905220.0,
      "step": 21410
    },
    {
      "entropy": 5.456166982650757,
      "epoch": 2.1253473600635173,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045507679318950166,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1681951880455017,
      "num_tokens": 46916328.0,
      "step": 21415
    },
    {
      "entropy": 5.695725011825561,
      "epoch": 2.125843588725685,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004550555481925743,
      "loss": 5.3862,
      "mean_token_accuracy": 0.14913859814405442,
      "num_tokens": 46927450.0,
      "step": 21420
    },
    {
      "entropy": 5.498643636703491,
      "epoch": 2.1263398173878523,
      "grad_norm": 3.109375,
      "learning_rate": 0.00045503429873065723,
      "loss": 5.1176,
      "mean_token_accuracy": 0.17038396894931793,
      "num_tokens": 46936758.0,
      "step": 21425
    },
    {
      "entropy": 5.4107115268707275,
      "epoch": 2.1268360460500197,
      "grad_norm": 1.25,
      "learning_rate": 0.0004550130448042775,
      "loss": 5.0777,
      "mean_token_accuracy": 0.17154116779565812,
      "num_tokens": 46945776.0,
      "step": 21430
    },
    {
      "entropy": 5.48002815246582,
      "epoch": 2.127332274712187,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004549917864139622,
      "loss": 5.1976,
      "mean_token_accuracy": 0.16928635388612748,
      "num_tokens": 46955621.0,
      "step": 21435
    },
    {
      "entropy": 5.629693078994751,
      "epoch": 2.1278285033743547,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004549705235602384,
      "loss": 5.308,
      "mean_token_accuracy": 0.16463774144649507,
      "num_tokens": 46966890.0,
      "step": 21440
    },
    {
      "entropy": 5.574399709701538,
      "epoch": 2.1283247320365226,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004549492562436334,
      "loss": 5.3127,
      "mean_token_accuracy": 0.16169525384902955,
      "num_tokens": 46978380.0,
      "step": 21445
    },
    {
      "entropy": 5.531002855300903,
      "epoch": 2.12882096069869,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045492798446467465,
      "loss": 5.1332,
      "mean_token_accuracy": 0.16725665479898452,
      "num_tokens": 46988286.0,
      "step": 21450
    },
    {
      "entropy": 5.47766900062561,
      "epoch": 2.1293171893608576,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045490670822388957,
      "loss": 5.1384,
      "mean_token_accuracy": 0.16732918322086335,
      "num_tokens": 46997795.0,
      "step": 21455
    },
    {
      "entropy": 5.419619607925415,
      "epoch": 2.129813418023025,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004548854275218059,
      "loss": 5.0268,
      "mean_token_accuracy": 0.17517706155776977,
      "num_tokens": 47007970.0,
      "step": 21460
    },
    {
      "entropy": 5.4493101119995115,
      "epoch": 2.1303096466851925,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00045486414235895115,
      "loss": 5.177,
      "mean_token_accuracy": 0.16520348638296128,
      "num_tokens": 47019548.0,
      "step": 21465
    },
    {
      "entropy": 5.5276265144348145,
      "epoch": 2.13080587534736,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004548428527358532,
      "loss": 5.2651,
      "mean_token_accuracy": 0.16383299082517624,
      "num_tokens": 47029568.0,
      "step": 21470
    },
    {
      "entropy": 5.58453402519226,
      "epoch": 2.1313021040095275,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004548215586530401,
      "loss": 5.2852,
      "mean_token_accuracy": 0.15513922572135924,
      "num_tokens": 47042048.0,
      "step": 21475
    },
    {
      "entropy": 5.494825744628907,
      "epoch": 2.131798332671695,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00045480026011103975,
      "loss": 5.1108,
      "mean_token_accuracy": 0.16455139666795732,
      "num_tokens": 47052099.0,
      "step": 21480
    },
    {
      "entropy": 5.522879505157471,
      "epoch": 2.132294561333863,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045477895711038045,
      "loss": 5.1698,
      "mean_token_accuracy": 0.1625725194811821,
      "num_tokens": 47063310.0,
      "step": 21485
    },
    {
      "entropy": 5.546034288406372,
      "epoch": 2.1327907899960303,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004547576496515904,
      "loss": 5.186,
      "mean_token_accuracy": 0.16305418461561202,
      "num_tokens": 47073545.0,
      "step": 21490
    },
    {
      "entropy": 5.542877960205078,
      "epoch": 2.133287018658198,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045473633773519795,
      "loss": 5.2556,
      "mean_token_accuracy": 0.16517556309700013,
      "num_tokens": 47085456.0,
      "step": 21495
    },
    {
      "entropy": 5.5548415184021,
      "epoch": 2.1337832473203653,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045471502136173173,
      "loss": 5.2333,
      "mean_token_accuracy": 0.16435377299785614,
      "num_tokens": 47096590.0,
      "step": 21500
    },
    {
      "entropy": 5.583862018585205,
      "epoch": 2.1342794759825328,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004546937005317201,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1687924399971962,
      "num_tokens": 47107876.0,
      "step": 21505
    },
    {
      "entropy": 5.515149927139282,
      "epoch": 2.1347757046447002,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045467237524569193,
      "loss": 5.2089,
      "mean_token_accuracy": 0.1659271240234375,
      "num_tokens": 47117662.0,
      "step": 21510
    },
    {
      "entropy": 5.482282924652099,
      "epoch": 2.1352719333068677,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045465104550417604,
      "loss": 5.2263,
      "mean_token_accuracy": 0.16182601004838942,
      "num_tokens": 47129283.0,
      "step": 21515
    },
    {
      "entropy": 5.593205308914184,
      "epoch": 2.135768161969035,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045462971130770123,
      "loss": 5.3015,
      "mean_token_accuracy": 0.15734051018953324,
      "num_tokens": 47140824.0,
      "step": 21520
    },
    {
      "entropy": 5.547849798202515,
      "epoch": 2.1362643906312027,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004546083726567968,
      "loss": 5.1815,
      "mean_token_accuracy": 0.16654270738363267,
      "num_tokens": 47151733.0,
      "step": 21525
    },
    {
      "entropy": 5.533311605453491,
      "epoch": 2.1367606192933706,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004545870295519917,
      "loss": 5.2739,
      "mean_token_accuracy": 0.15523533821105956,
      "num_tokens": 47162846.0,
      "step": 21530
    },
    {
      "entropy": 5.565865898132325,
      "epoch": 2.137256847955538,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045456568199381524,
      "loss": 5.1761,
      "mean_token_accuracy": 0.16610772013664246,
      "num_tokens": 47174969.0,
      "step": 21535
    },
    {
      "entropy": 5.594806718826294,
      "epoch": 2.1377530766177055,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045454432998279676,
      "loss": 5.2721,
      "mean_token_accuracy": 0.1595334604382515,
      "num_tokens": 47186903.0,
      "step": 21540
    },
    {
      "entropy": 5.547672080993652,
      "epoch": 2.138249305279873,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004545229735194658,
      "loss": 5.2257,
      "mean_token_accuracy": 0.16228191703557968,
      "num_tokens": 47197716.0,
      "step": 21545
    },
    {
      "entropy": 5.511367559432983,
      "epoch": 2.1387455339420405,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045450161260435196,
      "loss": 5.1613,
      "mean_token_accuracy": 0.16345838457345963,
      "num_tokens": 47208623.0,
      "step": 21550
    },
    {
      "entropy": 5.464579772949219,
      "epoch": 2.139241762604208,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004544802472379849,
      "loss": 5.1752,
      "mean_token_accuracy": 0.16824949085712432,
      "num_tokens": 47218248.0,
      "step": 21555
    },
    {
      "entropy": 5.550894021987915,
      "epoch": 2.1397379912663754,
      "grad_norm": 1.375,
      "learning_rate": 0.0004544588774208945,
      "loss": 5.3087,
      "mean_token_accuracy": 0.15754791498184204,
      "num_tokens": 47232028.0,
      "step": 21560
    },
    {
      "entropy": 5.604903364181519,
      "epoch": 2.140234219928543,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045443750315361055,
      "loss": 5.1712,
      "mean_token_accuracy": 0.16541116088628768,
      "num_tokens": 47241976.0,
      "step": 21565
    },
    {
      "entropy": 5.543795156478882,
      "epoch": 2.140730448590711,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004544161244366632,
      "loss": 5.1774,
      "mean_token_accuracy": 0.1695006251335144,
      "num_tokens": 47253228.0,
      "step": 21570
    },
    {
      "entropy": 5.483796787261963,
      "epoch": 2.1412266772528783,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004543947412705826,
      "loss": 5.1904,
      "mean_token_accuracy": 0.16227948069572448,
      "num_tokens": 47265210.0,
      "step": 21575
    },
    {
      "entropy": 5.4924109935760494,
      "epoch": 2.1417229059150458,
      "grad_norm": 1.921875,
      "learning_rate": 0.00045437335365589905,
      "loss": 5.2138,
      "mean_token_accuracy": 0.16159106642007828,
      "num_tokens": 47275723.0,
      "step": 21580
    },
    {
      "entropy": 5.578125619888306,
      "epoch": 2.1422191345772132,
      "grad_norm": 1.65625,
      "learning_rate": 0.00045435196159314276,
      "loss": 5.2527,
      "mean_token_accuracy": 0.16295116245746613,
      "num_tokens": 47286469.0,
      "step": 21585
    },
    {
      "entropy": 5.616263341903687,
      "epoch": 2.1427153632393807,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045433056508284425,
      "loss": 5.2686,
      "mean_token_accuracy": 0.15944778621196748,
      "num_tokens": 47296461.0,
      "step": 21590
    },
    {
      "entropy": 5.524826860427856,
      "epoch": 2.143211591901548,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004543091641255342,
      "loss": 5.1613,
      "mean_token_accuracy": 0.1688230276107788,
      "num_tokens": 47307217.0,
      "step": 21595
    },
    {
      "entropy": 5.539480447769165,
      "epoch": 2.1437078205637157,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004542877587217432,
      "loss": 5.2915,
      "mean_token_accuracy": 0.1543130487203598,
      "num_tokens": 47317041.0,
      "step": 21600
    },
    {
      "entropy": 5.5477211475372314,
      "epoch": 2.144204049225883,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00045426634887200224,
      "loss": 5.2104,
      "mean_token_accuracy": 0.1630730912089348,
      "num_tokens": 47328185.0,
      "step": 21605
    },
    {
      "entropy": 5.5892256736755375,
      "epoch": 2.1447002778880506,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000454244934576842,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1557512938976288,
      "num_tokens": 47339984.0,
      "step": 21610
    },
    {
      "entropy": 5.56279935836792,
      "epoch": 2.1451965065502185,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00045422351583679365,
      "loss": 5.2017,
      "mean_token_accuracy": 0.16832885146141052,
      "num_tokens": 47350911.0,
      "step": 21615
    },
    {
      "entropy": 5.557290315628052,
      "epoch": 2.145692735212386,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004542020926523883,
      "loss": 5.1947,
      "mean_token_accuracy": 0.15839567631483079,
      "num_tokens": 47361703.0,
      "step": 21620
    },
    {
      "entropy": 5.50635175704956,
      "epoch": 2.1461889638745535,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004541806650241572,
      "loss": 5.1651,
      "mean_token_accuracy": 0.15979110300540925,
      "num_tokens": 47372878.0,
      "step": 21625
    },
    {
      "entropy": 5.490881872177124,
      "epoch": 2.146685192536721,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004541592329526317,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1596708342432976,
      "num_tokens": 47383056.0,
      "step": 21630
    },
    {
      "entropy": 5.508893394470215,
      "epoch": 2.1471814211988884,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004541377964383434,
      "loss": 5.2213,
      "mean_token_accuracy": 0.1614010900259018,
      "num_tokens": 47394160.0,
      "step": 21635
    },
    {
      "entropy": 5.543125247955322,
      "epoch": 2.147677649861056,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004541163554818237,
      "loss": 5.1929,
      "mean_token_accuracy": 0.15970841646194459,
      "num_tokens": 47404861.0,
      "step": 21640
    },
    {
      "entropy": 5.55666880607605,
      "epoch": 2.1481738785232234,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045409491008360423,
      "loss": 5.2467,
      "mean_token_accuracy": 0.16449418663978577,
      "num_tokens": 47415236.0,
      "step": 21645
    },
    {
      "entropy": 5.582776975631714,
      "epoch": 2.148670107185391,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004540734602442169,
      "loss": 5.2207,
      "mean_token_accuracy": 0.1621472045779228,
      "num_tokens": 47425789.0,
      "step": 21650
    },
    {
      "entropy": 5.531851482391358,
      "epoch": 2.149166335847559,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004540520059641937,
      "loss": 5.0878,
      "mean_token_accuracy": 0.1685101330280304,
      "num_tokens": 47436236.0,
      "step": 21655
    },
    {
      "entropy": 5.457549619674682,
      "epoch": 2.1496625645097263,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004540305472440666,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1710662618279457,
      "num_tokens": 47447041.0,
      "step": 21660
    },
    {
      "entropy": 5.512644672393799,
      "epoch": 2.1501587931718937,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004540090840843677,
      "loss": 5.238,
      "mean_token_accuracy": 0.15955309420824051,
      "num_tokens": 47458001.0,
      "step": 21665
    },
    {
      "entropy": 5.45636568069458,
      "epoch": 2.150655021834061,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004539876164856292,
      "loss": 5.1163,
      "mean_token_accuracy": 0.1688752591609955,
      "num_tokens": 47468621.0,
      "step": 21670
    },
    {
      "entropy": 5.486571264266968,
      "epoch": 2.1511512504962287,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004539661444483835,
      "loss": 5.2544,
      "mean_token_accuracy": 0.16284046769142152,
      "num_tokens": 47478957.0,
      "step": 21675
    },
    {
      "entropy": 5.535067939758301,
      "epoch": 2.151647479158396,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004539446679731631,
      "loss": 5.1645,
      "mean_token_accuracy": 0.16432086527347564,
      "num_tokens": 47489640.0,
      "step": 21680
    },
    {
      "entropy": 5.587317323684692,
      "epoch": 2.1521437078205636,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004539231870605005,
      "loss": 5.3195,
      "mean_token_accuracy": 0.1592220976948738,
      "num_tokens": 47501581.0,
      "step": 21685
    },
    {
      "entropy": 5.564262771606446,
      "epoch": 2.152639936482731,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045390170171092845,
      "loss": 5.2424,
      "mean_token_accuracy": 0.16339317560195923,
      "num_tokens": 47513605.0,
      "step": 21690
    },
    {
      "entropy": 5.592100524902344,
      "epoch": 2.1531361651448986,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045388021192497956,
      "loss": 5.238,
      "mean_token_accuracy": 0.16385203301906587,
      "num_tokens": 47523784.0,
      "step": 21695
    },
    {
      "entropy": 5.498303174972534,
      "epoch": 2.1536323938070665,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000453858717703187,
      "loss": 5.1157,
      "mean_token_accuracy": 0.16363901197910308,
      "num_tokens": 47534643.0,
      "step": 21700
    },
    {
      "entropy": 5.477455997467041,
      "epoch": 2.154128622469234,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004538372190460836,
      "loss": 5.2638,
      "mean_token_accuracy": 0.16033129543066024,
      "num_tokens": 47545920.0,
      "step": 21705
    },
    {
      "entropy": 5.522467422485351,
      "epoch": 2.1546248511314015,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004538157159542025,
      "loss": 5.1759,
      "mean_token_accuracy": 0.1609536215662956,
      "num_tokens": 47557024.0,
      "step": 21710
    },
    {
      "entropy": 5.571152448654175,
      "epoch": 2.155121079793569,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004537942084280771,
      "loss": 5.1506,
      "mean_token_accuracy": 0.16375865787267685,
      "num_tokens": 47567531.0,
      "step": 21715
    },
    {
      "entropy": 5.537459373474121,
      "epoch": 2.1556173084557364,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045377269646824036,
      "loss": 5.2838,
      "mean_token_accuracy": 0.1622534692287445,
      "num_tokens": 47578642.0,
      "step": 21720
    },
    {
      "entropy": 5.523290157318115,
      "epoch": 2.156113537117904,
      "grad_norm": 1.328125,
      "learning_rate": 0.00045375118007522607,
      "loss": 5.3356,
      "mean_token_accuracy": 0.15656960755586624,
      "num_tokens": 47589346.0,
      "step": 21725
    },
    {
      "entropy": 5.489192962646484,
      "epoch": 2.1566097657800714,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004537296592495678,
      "loss": 5.1235,
      "mean_token_accuracy": 0.16788423508405687,
      "num_tokens": 47599689.0,
      "step": 21730
    },
    {
      "entropy": 5.573242998123169,
      "epoch": 2.157105994442239,
      "grad_norm": 1.265625,
      "learning_rate": 0.000453708133991799,
      "loss": 5.2306,
      "mean_token_accuracy": 0.16168343126773835,
      "num_tokens": 47609214.0,
      "step": 21735
    },
    {
      "entropy": 5.566438102722168,
      "epoch": 2.1576022231044063,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045368660430245364,
      "loss": 5.1828,
      "mean_token_accuracy": 0.1608782559633255,
      "num_tokens": 47620119.0,
      "step": 21740
    },
    {
      "entropy": 5.529528856277466,
      "epoch": 2.158098451766574,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004536650701820654,
      "loss": 5.144,
      "mean_token_accuracy": 0.16849304884672164,
      "num_tokens": 47630085.0,
      "step": 21745
    },
    {
      "entropy": 5.516789102554322,
      "epoch": 2.1585946804287417,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004536435316311685,
      "loss": 5.146,
      "mean_token_accuracy": 0.16701103001832962,
      "num_tokens": 47641457.0,
      "step": 21750
    },
    {
      "entropy": 5.500120401382446,
      "epoch": 2.159090909090909,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045362198865029696,
      "loss": 5.1681,
      "mean_token_accuracy": 0.16073959767818452,
      "num_tokens": 47652269.0,
      "step": 21755
    },
    {
      "entropy": 5.544623184204101,
      "epoch": 2.1595871377530766,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045360044123998495,
      "loss": 5.1981,
      "mean_token_accuracy": 0.16373511403799057,
      "num_tokens": 47664601.0,
      "step": 21760
    },
    {
      "entropy": 5.5800323486328125,
      "epoch": 2.160083366415244,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00045357888940076685,
      "loss": 5.3421,
      "mean_token_accuracy": 0.1561278894543648,
      "num_tokens": 47676187.0,
      "step": 21765
    },
    {
      "entropy": 5.451880311965942,
      "epoch": 2.1605795950774116,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004535573331331771,
      "loss": 5.1123,
      "mean_token_accuracy": 0.17143600583076476,
      "num_tokens": 47687057.0,
      "step": 21770
    },
    {
      "entropy": 5.509826374053955,
      "epoch": 2.161075823739579,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045353577243775035,
      "loss": 5.1681,
      "mean_token_accuracy": 0.16035470217466355,
      "num_tokens": 47699074.0,
      "step": 21775
    },
    {
      "entropy": 5.545927572250366,
      "epoch": 2.1615720524017465,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004535142073150211,
      "loss": 5.1649,
      "mean_token_accuracy": 0.16737963557243346,
      "num_tokens": 47710400.0,
      "step": 21780
    },
    {
      "entropy": 5.479594707489014,
      "epoch": 2.162068281063914,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004534926377655241,
      "loss": 5.0482,
      "mean_token_accuracy": 0.1769162595272064,
      "num_tokens": 47720900.0,
      "step": 21785
    },
    {
      "entropy": 5.490664958953857,
      "epoch": 2.162564509726082,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045347106378979435,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17007680535316466,
      "num_tokens": 47730702.0,
      "step": 21790
    },
    {
      "entropy": 5.477607774734497,
      "epoch": 2.1630607383882494,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045344948538836674,
      "loss": 5.1466,
      "mean_token_accuracy": 0.16512856632471085,
      "num_tokens": 47740451.0,
      "step": 21795
    },
    {
      "entropy": 5.527210330963134,
      "epoch": 2.163556967050417,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045342790256177646,
      "loss": 5.2298,
      "mean_token_accuracy": 0.16097091287374496,
      "num_tokens": 47752285.0,
      "step": 21800
    },
    {
      "entropy": 5.557173204421997,
      "epoch": 2.1640531957125844,
      "grad_norm": 1.25,
      "learning_rate": 0.0004534063153105587,
      "loss": 5.1972,
      "mean_token_accuracy": 0.164613838493824,
      "num_tokens": 47761895.0,
      "step": 21805
    },
    {
      "entropy": 5.545401382446289,
      "epoch": 2.164549424374752,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004533847236352487,
      "loss": 5.1214,
      "mean_token_accuracy": 0.17162588834762574,
      "num_tokens": 47771718.0,
      "step": 21810
    },
    {
      "entropy": 5.457294893264771,
      "epoch": 2.1650456530369193,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000453363127536382,
      "loss": 5.1558,
      "mean_token_accuracy": 0.16674934774637223,
      "num_tokens": 47782183.0,
      "step": 21815
    },
    {
      "entropy": 5.457601070404053,
      "epoch": 2.165541881699087,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045334152701449407,
      "loss": 5.119,
      "mean_token_accuracy": 0.16741530746221542,
      "num_tokens": 47792190.0,
      "step": 21820
    },
    {
      "entropy": 5.592886257171631,
      "epoch": 2.1660381103612543,
      "grad_norm": 1.125,
      "learning_rate": 0.0004533199220701205,
      "loss": 5.3017,
      "mean_token_accuracy": 0.15461400896310806,
      "num_tokens": 47803245.0,
      "step": 21825
    },
    {
      "entropy": 5.558335494995117,
      "epoch": 2.166534339023422,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045329831270379714,
      "loss": 5.2491,
      "mean_token_accuracy": 0.15587470084428787,
      "num_tokens": 47814605.0,
      "step": 21830
    },
    {
      "entropy": 5.542298793792725,
      "epoch": 2.1670305676855897,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045327669891605975,
      "loss": 5.2337,
      "mean_token_accuracy": 0.16080724596977233,
      "num_tokens": 47825945.0,
      "step": 21835
    },
    {
      "entropy": 5.5601531028747555,
      "epoch": 2.167526796347757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004532550807074445,
      "loss": 5.2423,
      "mean_token_accuracy": 0.15991494804620743,
      "num_tokens": 47838119.0,
      "step": 21840
    },
    {
      "entropy": 5.547828102111817,
      "epoch": 2.1680230250099246,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004532334580784873,
      "loss": 5.2174,
      "mean_token_accuracy": 0.1640021800994873,
      "num_tokens": 47849560.0,
      "step": 21845
    },
    {
      "entropy": 5.597617673873901,
      "epoch": 2.168519253672092,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004532118310297244,
      "loss": 5.2275,
      "mean_token_accuracy": 0.15216896533966065,
      "num_tokens": 47860036.0,
      "step": 21850
    },
    {
      "entropy": 5.484221744537353,
      "epoch": 2.1690154823342596,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004531901995616921,
      "loss": 5.1024,
      "mean_token_accuracy": 0.16708443462848663,
      "num_tokens": 47870826.0,
      "step": 21855
    },
    {
      "entropy": 5.47931227684021,
      "epoch": 2.169511710996427,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004531685636749268,
      "loss": 5.2388,
      "mean_token_accuracy": 0.15682356655597687,
      "num_tokens": 47881415.0,
      "step": 21860
    },
    {
      "entropy": 5.612077903747559,
      "epoch": 2.1700079396585945,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004531469233699651,
      "loss": 5.2352,
      "mean_token_accuracy": 0.1622293084859848,
      "num_tokens": 47893343.0,
      "step": 21865
    },
    {
      "entropy": 5.548490238189697,
      "epoch": 2.170504168320762,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004531252786473436,
      "loss": 5.2031,
      "mean_token_accuracy": 0.15819638669490815,
      "num_tokens": 47904871.0,
      "step": 21870
    },
    {
      "entropy": 5.4975332736969,
      "epoch": 2.17100039698293,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004531036295075989,
      "loss": 5.1212,
      "mean_token_accuracy": 0.16638021320104598,
      "num_tokens": 47915367.0,
      "step": 21875
    },
    {
      "entropy": 5.522042226791382,
      "epoch": 2.1714966256450974,
      "grad_norm": 1.546875,
      "learning_rate": 0.00045308197595126797,
      "loss": 5.1568,
      "mean_token_accuracy": 0.16746782064437865,
      "num_tokens": 47925987.0,
      "step": 21880
    },
    {
      "entropy": 5.585354280471802,
      "epoch": 2.171992854307265,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004530603179788879,
      "loss": 5.2579,
      "mean_token_accuracy": 0.16244084984064103,
      "num_tokens": 47937688.0,
      "step": 21885
    },
    {
      "entropy": 5.522144556045532,
      "epoch": 2.1724890829694323,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004530386555909955,
      "loss": 5.1877,
      "mean_token_accuracy": 0.16773975342512132,
      "num_tokens": 47948550.0,
      "step": 21890
    },
    {
      "entropy": 5.51338791847229,
      "epoch": 2.1729853116316,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045301698878812813,
      "loss": 5.1737,
      "mean_token_accuracy": 0.16802124977111815,
      "num_tokens": 47957982.0,
      "step": 21895
    },
    {
      "entropy": 5.572063541412353,
      "epoch": 2.1734815402937673,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004529953175708231,
      "loss": 5.2541,
      "mean_token_accuracy": 0.16334800720214843,
      "num_tokens": 47969562.0,
      "step": 21900
    },
    {
      "entropy": 5.531816291809082,
      "epoch": 2.1739777689559348,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045297364193961763,
      "loss": 5.2434,
      "mean_token_accuracy": 0.16115805506706238,
      "num_tokens": 47981323.0,
      "step": 21905
    },
    {
      "entropy": 5.535302019119262,
      "epoch": 2.1744739976181022,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045295196189504934,
      "loss": 5.1743,
      "mean_token_accuracy": 0.16044152975082399,
      "num_tokens": 47991454.0,
      "step": 21910
    },
    {
      "entropy": 5.49794020652771,
      "epoch": 2.17497022628027,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045293027743765595,
      "loss": 5.0851,
      "mean_token_accuracy": 0.16724717915058135,
      "num_tokens": 48001300.0,
      "step": 21915
    },
    {
      "entropy": 5.5530232906341555,
      "epoch": 2.1754664549424376,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045290858856797503,
      "loss": 5.1924,
      "mean_token_accuracy": 0.16431782096624375,
      "num_tokens": 48012608.0,
      "step": 21920
    },
    {
      "entropy": 5.530513906478882,
      "epoch": 2.175962683604605,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045288689528654444,
      "loss": 5.1792,
      "mean_token_accuracy": 0.16402978748083114,
      "num_tokens": 48022120.0,
      "step": 21925
    },
    {
      "entropy": 5.552199935913086,
      "epoch": 2.1764589122667726,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045286519759390217,
      "loss": 5.1833,
      "mean_token_accuracy": 0.16256005316972733,
      "num_tokens": 48032629.0,
      "step": 21930
    },
    {
      "entropy": 5.491131782531738,
      "epoch": 2.17695514092894,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004528434954905863,
      "loss": 5.1151,
      "mean_token_accuracy": 0.165934157371521,
      "num_tokens": 48042934.0,
      "step": 21935
    },
    {
      "entropy": 5.522381544113159,
      "epoch": 2.1774513695911075,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004528217889771349,
      "loss": 5.1802,
      "mean_token_accuracy": 0.16735552996397018,
      "num_tokens": 48052825.0,
      "step": 21940
    },
    {
      "entropy": 5.581136560440063,
      "epoch": 2.177947598253275,
      "grad_norm": 1.25,
      "learning_rate": 0.00045280007805408636,
      "loss": 5.27,
      "mean_token_accuracy": 0.15402856469154358,
      "num_tokens": 48064565.0,
      "step": 21945
    },
    {
      "entropy": 5.616187906265258,
      "epoch": 2.1784438269154425,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045277836272197895,
      "loss": 5.2392,
      "mean_token_accuracy": 0.16292775571346282,
      "num_tokens": 48076006.0,
      "step": 21950
    },
    {
      "entropy": 5.6470366477966305,
      "epoch": 2.17894005557761,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045275664298135134,
      "loss": 5.3443,
      "mean_token_accuracy": 0.15871218740940093,
      "num_tokens": 48087110.0,
      "step": 21955
    },
    {
      "entropy": 5.547079706192017,
      "epoch": 2.179436284239778,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004527349188327419,
      "loss": 5.2385,
      "mean_token_accuracy": 0.1657652124762535,
      "num_tokens": 48098894.0,
      "step": 21960
    },
    {
      "entropy": 5.557986259460449,
      "epoch": 2.1799325129019453,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004527131902766895,
      "loss": 5.2371,
      "mean_token_accuracy": 0.16266757547855376,
      "num_tokens": 48109729.0,
      "step": 21965
    },
    {
      "entropy": 5.557261371612549,
      "epoch": 2.180428741564113,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004526914573137329,
      "loss": 5.2348,
      "mean_token_accuracy": 0.15736823081970214,
      "num_tokens": 48120345.0,
      "step": 21970
    },
    {
      "entropy": 5.5306089401245115,
      "epoch": 2.1809249702262803,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045266971994441116,
      "loss": 5.1628,
      "mean_token_accuracy": 0.16652027517557144,
      "num_tokens": 48130185.0,
      "step": 21975
    },
    {
      "entropy": 5.504667139053344,
      "epoch": 2.1814211988884478,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000452647978169263,
      "loss": 5.1315,
      "mean_token_accuracy": 0.16570262461900712,
      "num_tokens": 48141020.0,
      "step": 21980
    },
    {
      "entropy": 5.508136987686157,
      "epoch": 2.1819174275506152,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045262623198882797,
      "loss": 5.2329,
      "mean_token_accuracy": 0.1676996737718582,
      "num_tokens": 48153361.0,
      "step": 21985
    },
    {
      "entropy": 5.488990068435669,
      "epoch": 2.1824136562127827,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004526044814036451,
      "loss": 5.1728,
      "mean_token_accuracy": 0.16374772787094116,
      "num_tokens": 48163604.0,
      "step": 21990
    },
    {
      "entropy": 5.539148807525635,
      "epoch": 2.18290988487495,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045258272641425377,
      "loss": 5.1838,
      "mean_token_accuracy": 0.1634184941649437,
      "num_tokens": 48174732.0,
      "step": 21995
    },
    {
      "entropy": 5.527225112915039,
      "epoch": 2.183406113537118,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004525609670211934,
      "loss": 5.1557,
      "mean_token_accuracy": 0.16717542856931686,
      "num_tokens": 48187082.0,
      "step": 22000
    },
    {
      "entropy": 5.541254568099975,
      "epoch": 2.1839023421992856,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045253920322500373,
      "loss": 5.1649,
      "mean_token_accuracy": 0.166192664206028,
      "num_tokens": 48196547.0,
      "step": 22005
    },
    {
      "entropy": 5.499629878997803,
      "epoch": 2.184398570861453,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045251743502622445,
      "loss": 5.2165,
      "mean_token_accuracy": 0.16471272557973862,
      "num_tokens": 48207404.0,
      "step": 22010
    },
    {
      "entropy": 5.476947402954101,
      "epoch": 2.1848947995236205,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004524956624253952,
      "loss": 5.1888,
      "mean_token_accuracy": 0.16860360503196717,
      "num_tokens": 48217093.0,
      "step": 22015
    },
    {
      "entropy": 5.574565124511719,
      "epoch": 2.185391028185788,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045247388542305595,
      "loss": 5.2286,
      "mean_token_accuracy": 0.16294972151517867,
      "num_tokens": 48228093.0,
      "step": 22020
    },
    {
      "entropy": 5.51366548538208,
      "epoch": 2.1858872568479555,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004524521040197469,
      "loss": 5.1767,
      "mean_token_accuracy": 0.16623015999794005,
      "num_tokens": 48238834.0,
      "step": 22025
    },
    {
      "entropy": 5.565222454071045,
      "epoch": 2.186383485510123,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045243031821600794,
      "loss": 5.2371,
      "mean_token_accuracy": 0.169087977707386,
      "num_tokens": 48250782.0,
      "step": 22030
    },
    {
      "entropy": 5.462993240356445,
      "epoch": 2.1868797141722904,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004524085280123795,
      "loss": 5.1069,
      "mean_token_accuracy": 0.16611941009759904,
      "num_tokens": 48261711.0,
      "step": 22035
    },
    {
      "entropy": 5.5165441036224365,
      "epoch": 2.187375942834458,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004523867334094018,
      "loss": 5.2277,
      "mean_token_accuracy": 0.16026346236467362,
      "num_tokens": 48273505.0,
      "step": 22040
    },
    {
      "entropy": 5.46373405456543,
      "epoch": 2.187872171496626,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004523649344076154,
      "loss": 5.1273,
      "mean_token_accuracy": 0.16246480643749237,
      "num_tokens": 48284639.0,
      "step": 22045
    },
    {
      "entropy": 5.530734872817993,
      "epoch": 2.1883684001587933,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045234313100756077,
      "loss": 5.2357,
      "mean_token_accuracy": 0.1596030905842781,
      "num_tokens": 48296275.0,
      "step": 22050
    },
    {
      "entropy": 5.486393642425537,
      "epoch": 2.188864628820961,
      "grad_norm": 1.125,
      "learning_rate": 0.00045232132320977854,
      "loss": 5.1111,
      "mean_token_accuracy": 0.1689975827932358,
      "num_tokens": 48306859.0,
      "step": 22055
    },
    {
      "entropy": 5.504120588302612,
      "epoch": 2.1893608574831283,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004522995110148097,
      "loss": 5.2573,
      "mean_token_accuracy": 0.16436937153339387,
      "num_tokens": 48317561.0,
      "step": 22060
    },
    {
      "entropy": 5.515422582626343,
      "epoch": 2.1898570861452957,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000452277694423195,
      "loss": 5.2712,
      "mean_token_accuracy": 0.15568673312664033,
      "num_tokens": 48327814.0,
      "step": 22065
    },
    {
      "entropy": 5.568010139465332,
      "epoch": 2.190353314807463,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004522558734354755,
      "loss": 5.1423,
      "mean_token_accuracy": 0.167154560983181,
      "num_tokens": 48339658.0,
      "step": 22070
    },
    {
      "entropy": 5.501924705505371,
      "epoch": 2.1908495434696307,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045223404805219225,
      "loss": 5.2213,
      "mean_token_accuracy": 0.16437961757183076,
      "num_tokens": 48352115.0,
      "step": 22075
    },
    {
      "entropy": 5.538412714004517,
      "epoch": 2.191345772131798,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045221221827388655,
      "loss": 5.1846,
      "mean_token_accuracy": 0.16611179262399672,
      "num_tokens": 48362378.0,
      "step": 22080
    },
    {
      "entropy": 5.513823699951172,
      "epoch": 2.191842000793966,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004521903841010997,
      "loss": 5.1361,
      "mean_token_accuracy": 0.16875235587358475,
      "num_tokens": 48373035.0,
      "step": 22085
    },
    {
      "entropy": 5.486190938949585,
      "epoch": 2.1923382294561335,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004521685455343731,
      "loss": 5.1858,
      "mean_token_accuracy": 0.16872251629829407,
      "num_tokens": 48383737.0,
      "step": 22090
    },
    {
      "entropy": 5.477408218383789,
      "epoch": 2.192834458118301,
      "grad_norm": 1.125,
      "learning_rate": 0.00045214670257424835,
      "loss": 5.1766,
      "mean_token_accuracy": 0.1658860757946968,
      "num_tokens": 48394547.0,
      "step": 22095
    },
    {
      "entropy": 5.527464771270752,
      "epoch": 2.1933306867804685,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004521248552212671,
      "loss": 5.2403,
      "mean_token_accuracy": 0.16284438222646713,
      "num_tokens": 48404906.0,
      "step": 22100
    },
    {
      "entropy": 5.5140691757202145,
      "epoch": 2.193826915442636,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004521030034759711,
      "loss": 5.1798,
      "mean_token_accuracy": 0.16307101547718048,
      "num_tokens": 48416425.0,
      "step": 22105
    },
    {
      "entropy": 5.496731376647949,
      "epoch": 2.1943231441048034,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004520811473389023,
      "loss": 5.1177,
      "mean_token_accuracy": 0.1661939650774002,
      "num_tokens": 48426967.0,
      "step": 22110
    },
    {
      "entropy": 5.541968250274659,
      "epoch": 2.194819372766971,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045205928681060255,
      "loss": 5.2773,
      "mean_token_accuracy": 0.16380477845668792,
      "num_tokens": 48439720.0,
      "step": 22115
    },
    {
      "entropy": 5.551459217071534,
      "epoch": 2.1953156014291384,
      "grad_norm": 1.25,
      "learning_rate": 0.000452037421891614,
      "loss": 5.2351,
      "mean_token_accuracy": 0.16155776977539063,
      "num_tokens": 48449949.0,
      "step": 22120
    },
    {
      "entropy": 5.4712121963500975,
      "epoch": 2.195811830091306,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045201555258247896,
      "loss": 5.1271,
      "mean_token_accuracy": 0.1685774564743042,
      "num_tokens": 48462515.0,
      "step": 22125
    },
    {
      "entropy": 5.481185245513916,
      "epoch": 2.196308058753474,
      "grad_norm": 1.25,
      "learning_rate": 0.0004519936788837396,
      "loss": 5.1451,
      "mean_token_accuracy": 0.16681773364543914,
      "num_tokens": 48471984.0,
      "step": 22130
    },
    {
      "entropy": 5.5798540115356445,
      "epoch": 2.1968042874156413,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045197180079593836,
      "loss": 5.2536,
      "mean_token_accuracy": 0.16004066467285155,
      "num_tokens": 48482814.0,
      "step": 22135
    },
    {
      "entropy": 5.5615898132324215,
      "epoch": 2.1973005160778087,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045194991831961777,
      "loss": 5.2014,
      "mean_token_accuracy": 0.1671997055411339,
      "num_tokens": 48493670.0,
      "step": 22140
    },
    {
      "entropy": 5.498005151748657,
      "epoch": 2.197796744739976,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004519280314553206,
      "loss": 5.1935,
      "mean_token_accuracy": 0.1573366791009903,
      "num_tokens": 48504758.0,
      "step": 22145
    },
    {
      "entropy": 5.553024768829346,
      "epoch": 2.1982929734021437,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004519061402035894,
      "loss": 5.2077,
      "mean_token_accuracy": 0.15679626315832138,
      "num_tokens": 48515846.0,
      "step": 22150
    },
    {
      "entropy": 5.520708227157593,
      "epoch": 2.198789202064311,
      "grad_norm": 1.375,
      "learning_rate": 0.00045188424456496723,
      "loss": 5.201,
      "mean_token_accuracy": 0.16654880791902543,
      "num_tokens": 48526136.0,
      "step": 22155
    },
    {
      "entropy": 5.571712160110474,
      "epoch": 2.1992854307264786,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045186234453999693,
      "loss": 5.289,
      "mean_token_accuracy": 0.16040649265050888,
      "num_tokens": 48536031.0,
      "step": 22160
    },
    {
      "entropy": 5.431867456436157,
      "epoch": 2.199781659388646,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045184044012922154,
      "loss": 5.1079,
      "mean_token_accuracy": 0.1679813578724861,
      "num_tokens": 48546965.0,
      "step": 22165
    },
    {
      "entropy": 5.508588743209839,
      "epoch": 2.200277888050814,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045181853133318425,
      "loss": 5.1958,
      "mean_token_accuracy": 0.16278156042098998,
      "num_tokens": 48557637.0,
      "step": 22170
    },
    {
      "entropy": 5.529969072341919,
      "epoch": 2.2007741167129815,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004517966181524285,
      "loss": 5.2324,
      "mean_token_accuracy": 0.1644562378525734,
      "num_tokens": 48568578.0,
      "step": 22175
    },
    {
      "entropy": 5.6090991497039795,
      "epoch": 2.201270345375149,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004517747005874975,
      "loss": 5.3154,
      "mean_token_accuracy": 0.15978675931692124,
      "num_tokens": 48581782.0,
      "step": 22180
    },
    {
      "entropy": 5.416661214828491,
      "epoch": 2.2017665740373165,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004517527786389349,
      "loss": 5.1422,
      "mean_token_accuracy": 0.1766665607690811,
      "num_tokens": 48592432.0,
      "step": 22185
    },
    {
      "entropy": 5.47410216331482,
      "epoch": 2.202262802699484,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045173085230728417,
      "loss": 5.1593,
      "mean_token_accuracy": 0.16797374784946442,
      "num_tokens": 48603251.0,
      "step": 22190
    },
    {
      "entropy": 5.560413408279419,
      "epoch": 2.2027590313616514,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004517089215930892,
      "loss": 5.2861,
      "mean_token_accuracy": 0.1626691609621048,
      "num_tokens": 48616118.0,
      "step": 22195
    },
    {
      "entropy": 5.571821641921997,
      "epoch": 2.203255260023819,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045168698649689375,
      "loss": 5.2406,
      "mean_token_accuracy": 0.16055337637662886,
      "num_tokens": 48626903.0,
      "step": 22200
    },
    {
      "entropy": 5.495360660552978,
      "epoch": 2.2037514886859864,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045166504701924176,
      "loss": 5.1094,
      "mean_token_accuracy": 0.16640628576278688,
      "num_tokens": 48638186.0,
      "step": 22205
    },
    {
      "entropy": 5.49380521774292,
      "epoch": 2.204247717348154,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045164310316067727,
      "loss": 5.2373,
      "mean_token_accuracy": 0.1602238193154335,
      "num_tokens": 48649948.0,
      "step": 22210
    },
    {
      "entropy": 5.4896351337432865,
      "epoch": 2.2047439460103218,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004516211549217444,
      "loss": 5.1376,
      "mean_token_accuracy": 0.16890038400888444,
      "num_tokens": 48660869.0,
      "step": 22215
    },
    {
      "entropy": 5.469353103637696,
      "epoch": 2.2052401746724892,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004515992023029875,
      "loss": 5.1821,
      "mean_token_accuracy": 0.16515360325574874,
      "num_tokens": 48673383.0,
      "step": 22220
    },
    {
      "entropy": 5.433482122421265,
      "epoch": 2.2057364033346567,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000451577245304951,
      "loss": 5.1215,
      "mean_token_accuracy": 0.16955526173114777,
      "num_tokens": 48684071.0,
      "step": 22225
    },
    {
      "entropy": 5.502228164672852,
      "epoch": 2.206232631996824,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004515552839281792,
      "loss": 5.1814,
      "mean_token_accuracy": 0.16138224303722382,
      "num_tokens": 48695432.0,
      "step": 22230
    },
    {
      "entropy": 5.48540358543396,
      "epoch": 2.2067288606589917,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004515333181732169,
      "loss": 5.1703,
      "mean_token_accuracy": 0.16490595489740373,
      "num_tokens": 48706102.0,
      "step": 22235
    },
    {
      "entropy": 5.475807523727417,
      "epoch": 2.207225089321159,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045151134804060863,
      "loss": 5.2185,
      "mean_token_accuracy": 0.15849417299032212,
      "num_tokens": 48717748.0,
      "step": 22240
    },
    {
      "entropy": 5.518850803375244,
      "epoch": 2.2077213179833266,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004514893735308993,
      "loss": 5.1246,
      "mean_token_accuracy": 0.16828972548246385,
      "num_tokens": 48728941.0,
      "step": 22245
    },
    {
      "entropy": 5.472284984588623,
      "epoch": 2.208217546645494,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004514673946446338,
      "loss": 5.1623,
      "mean_token_accuracy": 0.1675392895936966,
      "num_tokens": 48738565.0,
      "step": 22250
    },
    {
      "entropy": 5.478096437454224,
      "epoch": 2.208713775307662,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004514454113823572,
      "loss": 5.169,
      "mean_token_accuracy": 0.15991166979074478,
      "num_tokens": 48749537.0,
      "step": 22255
    },
    {
      "entropy": 5.464058494567871,
      "epoch": 2.2092100039698295,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004514234237446146,
      "loss": 5.1573,
      "mean_token_accuracy": 0.1621959075331688,
      "num_tokens": 48760388.0,
      "step": 22260
    },
    {
      "entropy": 5.463026905059815,
      "epoch": 2.209706232631997,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045140143173195117,
      "loss": 5.1831,
      "mean_token_accuracy": 0.16195523738861084,
      "num_tokens": 48770551.0,
      "step": 22265
    },
    {
      "entropy": 5.48559536933899,
      "epoch": 2.2102024612941644,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004513794353449124,
      "loss": 5.131,
      "mean_token_accuracy": 0.1652864173054695,
      "num_tokens": 48780739.0,
      "step": 22270
    },
    {
      "entropy": 5.520423650741577,
      "epoch": 2.210698689956332,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004513574345840437,
      "loss": 5.2547,
      "mean_token_accuracy": 0.16335922926664354,
      "num_tokens": 48791478.0,
      "step": 22275
    },
    {
      "entropy": 5.566363096237183,
      "epoch": 2.2111949186184994,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004513354294498907,
      "loss": 5.2348,
      "mean_token_accuracy": 0.16108437329530717,
      "num_tokens": 48802202.0,
      "step": 22280
    },
    {
      "entropy": 5.53155632019043,
      "epoch": 2.211691147280667,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045131341994299895,
      "loss": 5.1941,
      "mean_token_accuracy": 0.16902452558279038,
      "num_tokens": 48813619.0,
      "step": 22285
    },
    {
      "entropy": 5.524889850616455,
      "epoch": 2.2121873759428343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004512914060639143,
      "loss": 5.148,
      "mean_token_accuracy": 0.1713162824511528,
      "num_tokens": 48824951.0,
      "step": 22290
    },
    {
      "entropy": 5.47394027709961,
      "epoch": 2.212683604605002,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004512693878131827,
      "loss": 5.1737,
      "mean_token_accuracy": 0.16664145439863204,
      "num_tokens": 48834328.0,
      "step": 22295
    },
    {
      "entropy": 5.4979386806488035,
      "epoch": 2.2131798332671697,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045124736519135003,
      "loss": 5.207,
      "mean_token_accuracy": 0.15992942154407502,
      "num_tokens": 48843880.0,
      "step": 22300
    },
    {
      "entropy": 5.54250316619873,
      "epoch": 2.213676061929337,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045122533819896257,
      "loss": 5.2074,
      "mean_token_accuracy": 0.16574738174676895,
      "num_tokens": 48854893.0,
      "step": 22305
    },
    {
      "entropy": 5.589069175720215,
      "epoch": 2.2141722905915047,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004512033068365664,
      "loss": 5.2831,
      "mean_token_accuracy": 0.1591109812259674,
      "num_tokens": 48867308.0,
      "step": 22310
    },
    {
      "entropy": 5.559221458435059,
      "epoch": 2.214668519253672,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004511812711047079,
      "loss": 5.2689,
      "mean_token_accuracy": 0.1638999953866005,
      "num_tokens": 48880222.0,
      "step": 22315
    },
    {
      "entropy": 5.570576477050781,
      "epoch": 2.2151647479158396,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045115923100393356,
      "loss": 5.1915,
      "mean_token_accuracy": 0.1612447053194046,
      "num_tokens": 48890345.0,
      "step": 22320
    },
    {
      "entropy": 5.479608774185181,
      "epoch": 2.215660976578007,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045113718653478984,
      "loss": 5.1204,
      "mean_token_accuracy": 0.16430535167455673,
      "num_tokens": 48901267.0,
      "step": 22325
    },
    {
      "entropy": 5.530904817581177,
      "epoch": 2.2161572052401746,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004511151376978234,
      "loss": 5.1965,
      "mean_token_accuracy": 0.1709487482905388,
      "num_tokens": 48911739.0,
      "step": 22330
    },
    {
      "entropy": 5.554531621932983,
      "epoch": 2.216653433902342,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045109308449358114,
      "loss": 5.1673,
      "mean_token_accuracy": 0.16833251416683198,
      "num_tokens": 48922376.0,
      "step": 22335
    },
    {
      "entropy": 5.4719194889068605,
      "epoch": 2.21714966256451,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004510710269226098,
      "loss": 5.0841,
      "mean_token_accuracy": 0.17195679247379303,
      "num_tokens": 48934101.0,
      "step": 22340
    },
    {
      "entropy": 5.440015697479248,
      "epoch": 2.2176458912266774,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004510489649854564,
      "loss": 5.1682,
      "mean_token_accuracy": 0.16887156963348388,
      "num_tokens": 48944684.0,
      "step": 22345
    },
    {
      "entropy": 5.495899677276611,
      "epoch": 2.218142119888845,
      "grad_norm": 1.3125,
      "learning_rate": 0.000451026898682668,
      "loss": 5.1751,
      "mean_token_accuracy": 0.15959163159132003,
      "num_tokens": 48955255.0,
      "step": 22350
    },
    {
      "entropy": 5.538818264007569,
      "epoch": 2.2186383485510124,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004510048280147918,
      "loss": 5.1835,
      "mean_token_accuracy": 0.16904345154762268,
      "num_tokens": 48967034.0,
      "step": 22355
    },
    {
      "entropy": 5.519360780715942,
      "epoch": 2.21913457721318,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004509827529823751,
      "loss": 5.2953,
      "mean_token_accuracy": 0.16159544438123702,
      "num_tokens": 48977903.0,
      "step": 22360
    },
    {
      "entropy": 5.5115190029144285,
      "epoch": 2.2196308058753473,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045096067358596536,
      "loss": 5.172,
      "mean_token_accuracy": 0.16727620214223862,
      "num_tokens": 48988889.0,
      "step": 22365
    },
    {
      "entropy": 5.5954766273498535,
      "epoch": 2.220127034537515,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045093858982611006,
      "loss": 5.2625,
      "mean_token_accuracy": 0.1537916675209999,
      "num_tokens": 48998946.0,
      "step": 22370
    },
    {
      "entropy": 5.5374377250671385,
      "epoch": 2.2206232631996823,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004509165017033569,
      "loss": 5.1658,
      "mean_token_accuracy": 0.17188843935728074,
      "num_tokens": 49008141.0,
      "step": 22375
    },
    {
      "entropy": 5.451020431518555,
      "epoch": 2.2211194918618498,
      "grad_norm": 1.25,
      "learning_rate": 0.00045089440921825354,
      "loss": 5.0786,
      "mean_token_accuracy": 0.1752004399895668,
      "num_tokens": 49019387.0,
      "step": 22380
    },
    {
      "entropy": 5.457706069946289,
      "epoch": 2.2216157205240172,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045087231237134787,
      "loss": 5.1411,
      "mean_token_accuracy": 0.16081384271383287,
      "num_tokens": 49030280.0,
      "step": 22385
    },
    {
      "entropy": 5.534147596359253,
      "epoch": 2.222111949186185,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045085021116318783,
      "loss": 5.1524,
      "mean_token_accuracy": 0.16463924795389176,
      "num_tokens": 49041766.0,
      "step": 22390
    },
    {
      "entropy": 5.663443326950073,
      "epoch": 2.2226081778483526,
      "grad_norm": 1.5,
      "learning_rate": 0.0004508281055943214,
      "loss": 5.3666,
      "mean_token_accuracy": 0.1557164818048477,
      "num_tokens": 49052931.0,
      "step": 22395
    },
    {
      "entropy": 5.527263879776001,
      "epoch": 2.22310440651052,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004508059956652969,
      "loss": 5.1661,
      "mean_token_accuracy": 0.16395096629858016,
      "num_tokens": 49064153.0,
      "step": 22400
    },
    {
      "entropy": 5.505408763885498,
      "epoch": 2.2236006351726876,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004507838813766625,
      "loss": 5.1925,
      "mean_token_accuracy": 0.16409283727407456,
      "num_tokens": 49075857.0,
      "step": 22405
    },
    {
      "entropy": 5.581838321685791,
      "epoch": 2.224096863834855,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004507617627289666,
      "loss": 5.2309,
      "mean_token_accuracy": 0.16302660405635833,
      "num_tokens": 49086632.0,
      "step": 22410
    },
    {
      "entropy": 5.531138610839844,
      "epoch": 2.2245930924970225,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004507396397227577,
      "loss": 5.1739,
      "mean_token_accuracy": 0.16322098970413207,
      "num_tokens": 49098086.0,
      "step": 22415
    },
    {
      "entropy": 5.473271179199219,
      "epoch": 2.22508932115919,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045071751235858454,
      "loss": 5.1048,
      "mean_token_accuracy": 0.17038651406764985,
      "num_tokens": 49108857.0,
      "step": 22420
    },
    {
      "entropy": 5.542001628875733,
      "epoch": 2.2255855498213575,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004506953806369956,
      "loss": 5.1194,
      "mean_token_accuracy": 0.16567240059375762,
      "num_tokens": 49119609.0,
      "step": 22425
    },
    {
      "entropy": 5.5174476146698,
      "epoch": 2.2260817784835254,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004506732445585399,
      "loss": 5.2167,
      "mean_token_accuracy": 0.16634688526391983,
      "num_tokens": 49130047.0,
      "step": 22430
    },
    {
      "entropy": 5.441845369338989,
      "epoch": 2.226578007145693,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004506511041237662,
      "loss": 5.1442,
      "mean_token_accuracy": 0.16671421229839326,
      "num_tokens": 49140517.0,
      "step": 22435
    },
    {
      "entropy": 5.522845029830933,
      "epoch": 2.2270742358078603,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045062895933322366,
      "loss": 5.1947,
      "mean_token_accuracy": 0.1639205738902092,
      "num_tokens": 49151497.0,
      "step": 22440
    },
    {
      "entropy": 5.452581119537354,
      "epoch": 2.227570464470028,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004506068101874614,
      "loss": 5.0969,
      "mean_token_accuracy": 0.16718845218420028,
      "num_tokens": 49161568.0,
      "step": 22445
    },
    {
      "entropy": 5.523007202148437,
      "epoch": 2.2280666931321953,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045058465668702856,
      "loss": 5.2216,
      "mean_token_accuracy": 0.16356701403856277,
      "num_tokens": 49173554.0,
      "step": 22450
    },
    {
      "entropy": 5.552880525588989,
      "epoch": 2.2285629217943628,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045056249883247457,
      "loss": 5.2223,
      "mean_token_accuracy": 0.1611184611916542,
      "num_tokens": 49185399.0,
      "step": 22455
    },
    {
      "entropy": 5.541726541519165,
      "epoch": 2.2290591504565302,
      "grad_norm": 1.125,
      "learning_rate": 0.00045054033662434903,
      "loss": 5.1297,
      "mean_token_accuracy": 0.16780281960964202,
      "num_tokens": 49195489.0,
      "step": 22460
    },
    {
      "entropy": 5.4334760189056395,
      "epoch": 2.2295553791186977,
      "grad_norm": 1.84375,
      "learning_rate": 0.00045051817006320127,
      "loss": 5.1138,
      "mean_token_accuracy": 0.16616333425045013,
      "num_tokens": 49204652.0,
      "step": 22465
    },
    {
      "entropy": 5.471486568450928,
      "epoch": 2.230051607780865,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045049599914958124,
      "loss": 5.203,
      "mean_token_accuracy": 0.16174644976854324,
      "num_tokens": 49215420.0,
      "step": 22470
    },
    {
      "entropy": 5.542993259429932,
      "epoch": 2.230547836443033,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004504738238840385,
      "loss": 5.157,
      "mean_token_accuracy": 0.16914481073617935,
      "num_tokens": 49225375.0,
      "step": 22475
    },
    {
      "entropy": 5.537234449386597,
      "epoch": 2.2310440651052006,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004504516442671231,
      "loss": 5.1652,
      "mean_token_accuracy": 0.16388435661792755,
      "num_tokens": 49238399.0,
      "step": 22480
    },
    {
      "entropy": 5.525661420822144,
      "epoch": 2.231540293767368,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045042946029938495,
      "loss": 5.1619,
      "mean_token_accuracy": 0.16205092668533325,
      "num_tokens": 49251069.0,
      "step": 22485
    },
    {
      "entropy": 5.530865716934204,
      "epoch": 2.2320365224295355,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045040727198137416,
      "loss": 5.1844,
      "mean_token_accuracy": 0.1645346015691757,
      "num_tokens": 49263859.0,
      "step": 22490
    },
    {
      "entropy": 5.531777858734131,
      "epoch": 2.232532751091703,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045038507931364103,
      "loss": 5.1966,
      "mean_token_accuracy": 0.1670854791998863,
      "num_tokens": 49274846.0,
      "step": 22495
    },
    {
      "entropy": 5.472803831100464,
      "epoch": 2.2330289797538705,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045036288229673583,
      "loss": 5.1271,
      "mean_token_accuracy": 0.17054740488529205,
      "num_tokens": 49285197.0,
      "step": 22500
    },
    {
      "entropy": 5.429881572723389,
      "epoch": 2.233525208416038,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045034068093120907,
      "loss": 5.1127,
      "mean_token_accuracy": 0.16362819373607634,
      "num_tokens": 49296238.0,
      "step": 22505
    },
    {
      "entropy": 5.6028687953948975,
      "epoch": 2.2340214370782054,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004503184752176112,
      "loss": 5.3533,
      "mean_token_accuracy": 0.15771825462579728,
      "num_tokens": 49308262.0,
      "step": 22510
    },
    {
      "entropy": 5.562515592575073,
      "epoch": 2.2345176657403734,
      "grad_norm": 1.21875,
      "learning_rate": 0.000450296265156493,
      "loss": 5.1388,
      "mean_token_accuracy": 0.16799918115139006,
      "num_tokens": 49318881.0,
      "step": 22515
    },
    {
      "entropy": 5.598639631271363,
      "epoch": 2.235013894402541,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045027405074840506,
      "loss": 5.2783,
      "mean_token_accuracy": 0.1611021175980568,
      "num_tokens": 49329470.0,
      "step": 22520
    },
    {
      "entropy": 5.526965236663818,
      "epoch": 2.2355101230647083,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004502518319938984,
      "loss": 5.1834,
      "mean_token_accuracy": 0.16693605184555055,
      "num_tokens": 49339806.0,
      "step": 22525
    },
    {
      "entropy": 5.585491085052491,
      "epoch": 2.236006351726876,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045022960889352394,
      "loss": 5.2736,
      "mean_token_accuracy": 0.1620868518948555,
      "num_tokens": 49350295.0,
      "step": 22530
    },
    {
      "entropy": 5.5564580917358395,
      "epoch": 2.2365025803890433,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00045020738144783273,
      "loss": 5.1834,
      "mean_token_accuracy": 0.16766526252031327,
      "num_tokens": 49360735.0,
      "step": 22535
    },
    {
      "entropy": 5.510238313674927,
      "epoch": 2.2369988090512107,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004501851496573761,
      "loss": 5.1636,
      "mean_token_accuracy": 0.1614219442009926,
      "num_tokens": 49372278.0,
      "step": 22540
    },
    {
      "entropy": 5.512216281890869,
      "epoch": 2.237495037713378,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004501629135227051,
      "loss": 5.2265,
      "mean_token_accuracy": 0.1597373053431511,
      "num_tokens": 49382842.0,
      "step": 22545
    },
    {
      "entropy": 5.562554121017456,
      "epoch": 2.2379912663755457,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004501406730443714,
      "loss": 5.2529,
      "mean_token_accuracy": 0.15954573601484298,
      "num_tokens": 49394633.0,
      "step": 22550
    },
    {
      "entropy": 5.589963102340699,
      "epoch": 2.238487495037713,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045011842822292636,
      "loss": 5.213,
      "mean_token_accuracy": 0.16050321310758592,
      "num_tokens": 49404731.0,
      "step": 22555
    },
    {
      "entropy": 5.584553241729736,
      "epoch": 2.238983723699881,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004500961790589216,
      "loss": 5.2123,
      "mean_token_accuracy": 0.16071494072675704,
      "num_tokens": 49415918.0,
      "step": 22560
    },
    {
      "entropy": 5.53318190574646,
      "epoch": 2.2394799523620486,
      "grad_norm": 1.09375,
      "learning_rate": 0.000450073925552909,
      "loss": 5.1495,
      "mean_token_accuracy": 0.1670832961797714,
      "num_tokens": 49427458.0,
      "step": 22565
    },
    {
      "entropy": 5.532553768157959,
      "epoch": 2.239976181024216,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004500516677054403,
      "loss": 5.1649,
      "mean_token_accuracy": 0.1678555816411972,
      "num_tokens": 49437696.0,
      "step": 22570
    },
    {
      "entropy": 5.5632165431976315,
      "epoch": 2.2404724096863835,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045002940551706726,
      "loss": 5.375,
      "mean_token_accuracy": 0.15483309850096702,
      "num_tokens": 49449220.0,
      "step": 22575
    },
    {
      "entropy": 5.538780212402344,
      "epoch": 2.240968638348551,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045000713898834225,
      "loss": 5.1928,
      "mean_token_accuracy": 0.16183760464191438,
      "num_tokens": 49459624.0,
      "step": 22580
    },
    {
      "entropy": 5.533479642868042,
      "epoch": 2.2414648670107185,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004499848681198173,
      "loss": 5.1266,
      "mean_token_accuracy": 0.1651563048362732,
      "num_tokens": 49471813.0,
      "step": 22585
    },
    {
      "entropy": 5.5378954887390135,
      "epoch": 2.241961095672886,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004499625929120447,
      "loss": 5.1947,
      "mean_token_accuracy": 0.16771238595247268,
      "num_tokens": 49482624.0,
      "step": 22590
    },
    {
      "entropy": 5.560544967651367,
      "epoch": 2.2424573243350534,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004499403133655768,
      "loss": 5.2212,
      "mean_token_accuracy": 0.15997350215911865,
      "num_tokens": 49493791.0,
      "step": 22595
    },
    {
      "entropy": 5.588617277145386,
      "epoch": 2.2429535529972213,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004499180294809661,
      "loss": 5.254,
      "mean_token_accuracy": 0.16321105808019637,
      "num_tokens": 49504060.0,
      "step": 22600
    },
    {
      "entropy": 5.4740190505981445,
      "epoch": 2.243449781659389,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044989574125876516,
      "loss": 5.075,
      "mean_token_accuracy": 0.16971904337406157,
      "num_tokens": 49515962.0,
      "step": 22605
    },
    {
      "entropy": 5.543039560317993,
      "epoch": 2.2439460103215563,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004498734486995267,
      "loss": 5.208,
      "mean_token_accuracy": 0.15757187604904174,
      "num_tokens": 49526676.0,
      "step": 22610
    },
    {
      "entropy": 5.497791576385498,
      "epoch": 2.2444422389837237,
      "grad_norm": 1.25,
      "learning_rate": 0.00044985115180380355,
      "loss": 5.2029,
      "mean_token_accuracy": 0.16146352887153625,
      "num_tokens": 49537347.0,
      "step": 22615
    },
    {
      "entropy": 5.545706796646118,
      "epoch": 2.2449384676458912,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044982885057214857,
      "loss": 5.2095,
      "mean_token_accuracy": 0.1590315192937851,
      "num_tokens": 49547422.0,
      "step": 22620
    },
    {
      "entropy": 5.519092178344726,
      "epoch": 2.2454346963080587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044980654500511485,
      "loss": 5.0494,
      "mean_token_accuracy": 0.17380220741033553,
      "num_tokens": 49557250.0,
      "step": 22625
    },
    {
      "entropy": 5.578054523468017,
      "epoch": 2.245930924970226,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044978423510325544,
      "loss": 5.2713,
      "mean_token_accuracy": 0.15856112837791442,
      "num_tokens": 49567106.0,
      "step": 22630
    },
    {
      "entropy": 5.552417945861817,
      "epoch": 2.2464271536323936,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004497619208671237,
      "loss": 5.2064,
      "mean_token_accuracy": 0.16164312809705733,
      "num_tokens": 49578978.0,
      "step": 22635
    },
    {
      "entropy": 5.583727931976318,
      "epoch": 2.246923382294561,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004497396022972728,
      "loss": 5.2406,
      "mean_token_accuracy": 0.16321008801460266,
      "num_tokens": 49589730.0,
      "step": 22640
    },
    {
      "entropy": 5.505625438690186,
      "epoch": 2.247419610956729,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004497172793942564,
      "loss": 5.1508,
      "mean_token_accuracy": 0.16679488867521286,
      "num_tokens": 49600881.0,
      "step": 22645
    },
    {
      "entropy": 5.487559747695923,
      "epoch": 2.2479158396188965,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004496949521586279,
      "loss": 5.1931,
      "mean_token_accuracy": 0.17071129083633424,
      "num_tokens": 49611120.0,
      "step": 22650
    },
    {
      "entropy": 5.571318197250366,
      "epoch": 2.248412068281064,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044967262059094097,
      "loss": 5.2216,
      "mean_token_accuracy": 0.1588459461927414,
      "num_tokens": 49622978.0,
      "step": 22655
    },
    {
      "entropy": 5.5652402400970455,
      "epoch": 2.2489082969432315,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00044965028469174956,
      "loss": 5.2087,
      "mean_token_accuracy": 0.1645622357726097,
      "num_tokens": 49633678.0,
      "step": 22660
    },
    {
      "entropy": 5.520512104034424,
      "epoch": 2.249404525605399,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044962794446160736,
      "loss": 5.2028,
      "mean_token_accuracy": 0.1625180423259735,
      "num_tokens": 49644886.0,
      "step": 22665
    },
    {
      "entropy": 5.545183563232422,
      "epoch": 2.2499007542675664,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044960559990106844,
      "loss": 5.1792,
      "mean_token_accuracy": 0.16685843020677565,
      "num_tokens": 49656790.0,
      "step": 22670
    },
    {
      "entropy": 5.558004474639892,
      "epoch": 2.250396982929734,
      "grad_norm": 1.125,
      "learning_rate": 0.0004495832510106869,
      "loss": 5.1995,
      "mean_token_accuracy": 0.16222488433122634,
      "num_tokens": 49667943.0,
      "step": 22675
    },
    {
      "entropy": 5.49074535369873,
      "epoch": 2.2508932115919014,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004495608977910169,
      "loss": 5.1743,
      "mean_token_accuracy": 0.16041896790266036,
      "num_tokens": 49678652.0,
      "step": 22680
    },
    {
      "entropy": 5.499684906005859,
      "epoch": 2.2513894402540693,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004495385402426128,
      "loss": 5.1418,
      "mean_token_accuracy": 0.17031644582748412,
      "num_tokens": 49688951.0,
      "step": 22685
    },
    {
      "entropy": 5.520634651184082,
      "epoch": 2.2518856689162368,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000449516178366029,
      "loss": 5.1839,
      "mean_token_accuracy": 0.16520969718694686,
      "num_tokens": 49698750.0,
      "step": 22690
    },
    {
      "entropy": 5.550082874298096,
      "epoch": 2.2523818975784042,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044949381216182,
      "loss": 5.287,
      "mean_token_accuracy": 0.15742112398147584,
      "num_tokens": 49709442.0,
      "step": 22695
    },
    {
      "entropy": 5.504216527938842,
      "epoch": 2.2528781262405717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044947144163054036,
      "loss": 5.1804,
      "mean_token_accuracy": 0.16845146119594573,
      "num_tokens": 49720517.0,
      "step": 22700
    },
    {
      "entropy": 5.543765735626221,
      "epoch": 2.253374354902739,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004494490667727451,
      "loss": 5.1429,
      "mean_token_accuracy": 0.1664794147014618,
      "num_tokens": 49731413.0,
      "step": 22705
    },
    {
      "entropy": 5.6410962581634525,
      "epoch": 2.2538705835649067,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044942668758898873,
      "loss": 5.344,
      "mean_token_accuracy": 0.156699288636446,
      "num_tokens": 49742492.0,
      "step": 22710
    },
    {
      "entropy": 5.543501901626587,
      "epoch": 2.254366812227074,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004494043040798265,
      "loss": 5.1843,
      "mean_token_accuracy": 0.16017158478498458,
      "num_tokens": 49753981.0,
      "step": 22715
    },
    {
      "entropy": 5.544346666336059,
      "epoch": 2.2548630408892416,
      "grad_norm": 1.25,
      "learning_rate": 0.0004493819162458133,
      "loss": 5.213,
      "mean_token_accuracy": 0.1617165371775627,
      "num_tokens": 49764718.0,
      "step": 22720
    },
    {
      "entropy": 5.543286371231079,
      "epoch": 2.255359269551409,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004493595240875043,
      "loss": 5.1812,
      "mean_token_accuracy": 0.15777661949396132,
      "num_tokens": 49775361.0,
      "step": 22725
    },
    {
      "entropy": 5.494995403289795,
      "epoch": 2.255855498213577,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044933712760545486,
      "loss": 5.1553,
      "mean_token_accuracy": 0.16475388109683992,
      "num_tokens": 49786670.0,
      "step": 22730
    },
    {
      "entropy": 5.575387334823608,
      "epoch": 2.2563517268757445,
      "grad_norm": 1.25,
      "learning_rate": 0.0004493147268002203,
      "loss": 5.2788,
      "mean_token_accuracy": 0.15981075465679168,
      "num_tokens": 49796514.0,
      "step": 22735
    },
    {
      "entropy": 5.4886702537536625,
      "epoch": 2.256847955537912,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044929232167235613,
      "loss": 5.1197,
      "mean_token_accuracy": 0.17324003130197524,
      "num_tokens": 49807042.0,
      "step": 22740
    },
    {
      "entropy": 5.529919385910034,
      "epoch": 2.2573441842000794,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004492699122224179,
      "loss": 5.2092,
      "mean_token_accuracy": 0.16209136694669724,
      "num_tokens": 49818779.0,
      "step": 22745
    },
    {
      "entropy": 5.548088932037354,
      "epoch": 2.257840412862247,
      "grad_norm": 1.25,
      "learning_rate": 0.00044924749845096144,
      "loss": 5.2188,
      "mean_token_accuracy": 0.16055838614702225,
      "num_tokens": 49829901.0,
      "step": 22750
    },
    {
      "entropy": 5.5553676128387455,
      "epoch": 2.2583366415244144,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004492250803585425,
      "loss": 5.205,
      "mean_token_accuracy": 0.16677282750606537,
      "num_tokens": 49840837.0,
      "step": 22755
    },
    {
      "entropy": 5.531028890609742,
      "epoch": 2.258832870186582,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004492026579457169,
      "loss": 5.1797,
      "mean_token_accuracy": 0.1593444511294365,
      "num_tokens": 49853052.0,
      "step": 22760
    },
    {
      "entropy": 5.46493558883667,
      "epoch": 2.2593290988487493,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004491802312130408,
      "loss": 5.1786,
      "mean_token_accuracy": 0.1582983136177063,
      "num_tokens": 49864210.0,
      "step": 22765
    },
    {
      "entropy": 5.484230852127075,
      "epoch": 2.2598253275109172,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004491578001610703,
      "loss": 5.1502,
      "mean_token_accuracy": 0.1741468131542206,
      "num_tokens": 49874995.0,
      "step": 22770
    },
    {
      "entropy": 5.487445640563965,
      "epoch": 2.2603215561730847,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044913536479036163,
      "loss": 5.1597,
      "mean_token_accuracy": 0.16286660581827164,
      "num_tokens": 49886240.0,
      "step": 22775
    },
    {
      "entropy": 5.513260316848755,
      "epoch": 2.260817784835252,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044911292510147117,
      "loss": 5.1603,
      "mean_token_accuracy": 0.16467565894126893,
      "num_tokens": 49897580.0,
      "step": 22780
    },
    {
      "entropy": 5.501335620880127,
      "epoch": 2.2613140134974197,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004490904810949553,
      "loss": 5.1861,
      "mean_token_accuracy": 0.1686963677406311,
      "num_tokens": 49907280.0,
      "step": 22785
    },
    {
      "entropy": 5.540164566040039,
      "epoch": 2.261810242159587,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044906803277137064,
      "loss": 5.1981,
      "mean_token_accuracy": 0.16935697942972183,
      "num_tokens": 49917552.0,
      "step": 22790
    },
    {
      "entropy": 5.507814741134643,
      "epoch": 2.2623064708217546,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004490455801312738,
      "loss": 5.1893,
      "mean_token_accuracy": 0.16202062666416167,
      "num_tokens": 49930610.0,
      "step": 22795
    },
    {
      "entropy": 5.518650197982788,
      "epoch": 2.262802699483922,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004490231231752217,
      "loss": 5.2147,
      "mean_token_accuracy": 0.16054128855466843,
      "num_tokens": 49941054.0,
      "step": 22800
    },
    {
      "entropy": 5.4156805038452145,
      "epoch": 2.2632989281460896,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004490006619037711,
      "loss": 5.0674,
      "mean_token_accuracy": 0.17487505972385406,
      "num_tokens": 49950885.0,
      "step": 22805
    },
    {
      "entropy": 5.4728758335113525,
      "epoch": 2.263795156808257,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044897819631747903,
      "loss": 5.1814,
      "mean_token_accuracy": 0.16392190754413605,
      "num_tokens": 49961723.0,
      "step": 22810
    },
    {
      "entropy": 5.529411172866821,
      "epoch": 2.2642913854704245,
      "grad_norm": 1.59375,
      "learning_rate": 0.00044895572641690256,
      "loss": 5.1482,
      "mean_token_accuracy": 0.16382334530353546,
      "num_tokens": 49974124.0,
      "step": 22815
    },
    {
      "entropy": 5.551183843612671,
      "epoch": 2.2647876141325924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000448933252202599,
      "loss": 5.1814,
      "mean_token_accuracy": 0.16717819571495057,
      "num_tokens": 49985792.0,
      "step": 22820
    },
    {
      "entropy": 5.493423271179199,
      "epoch": 2.26528384279476,
      "grad_norm": 1.125,
      "learning_rate": 0.00044891077367512544,
      "loss": 5.1072,
      "mean_token_accuracy": 0.17037007659673692,
      "num_tokens": 49996935.0,
      "step": 22825
    },
    {
      "entropy": 5.583881092071533,
      "epoch": 2.2657800714569274,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004488882908350396,
      "loss": 5.2767,
      "mean_token_accuracy": 0.15852364003658295,
      "num_tokens": 50007545.0,
      "step": 22830
    },
    {
      "entropy": 5.572339582443237,
      "epoch": 2.266276300119095,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004488658036828987,
      "loss": 5.1938,
      "mean_token_accuracy": 0.1636246994137764,
      "num_tokens": 50018797.0,
      "step": 22835
    },
    {
      "entropy": 5.531377267837525,
      "epoch": 2.2667725287812623,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004488433122192606,
      "loss": 5.1781,
      "mean_token_accuracy": 0.16415890753269197,
      "num_tokens": 50030884.0,
      "step": 22840
    },
    {
      "entropy": 5.579358768463135,
      "epoch": 2.26726875744343,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044882081644468286,
      "loss": 5.2939,
      "mean_token_accuracy": 0.15983881950378417,
      "num_tokens": 50042327.0,
      "step": 22845
    },
    {
      "entropy": 5.572912549972534,
      "epoch": 2.2677649861055973,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044879831635972345,
      "loss": 5.2354,
      "mean_token_accuracy": 0.16568584889173507,
      "num_tokens": 50053733.0,
      "step": 22850
    },
    {
      "entropy": 5.508528375625611,
      "epoch": 2.268261214767765,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004487758119649403,
      "loss": 5.1589,
      "mean_token_accuracy": 0.16907001733779908,
      "num_tokens": 50064153.0,
      "step": 22855
    },
    {
      "entropy": 5.458028554916382,
      "epoch": 2.2687574434299327,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00044875330326089153,
      "loss": 5.111,
      "mean_token_accuracy": 0.175497604906559,
      "num_tokens": 50074397.0,
      "step": 22860
    },
    {
      "entropy": 5.467675113677979,
      "epoch": 2.2692536720921,
      "grad_norm": 1.125,
      "learning_rate": 0.0004487307902481352,
      "loss": 5.1117,
      "mean_token_accuracy": 0.16177179366350175,
      "num_tokens": 50084074.0,
      "step": 22865
    },
    {
      "entropy": 5.497901725769043,
      "epoch": 2.2697499007542676,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044870827292722967,
      "loss": 5.1294,
      "mean_token_accuracy": 0.16412718147039412,
      "num_tokens": 50094535.0,
      "step": 22870
    },
    {
      "entropy": 5.5345312595367435,
      "epoch": 2.270246129416435,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004486857512987333,
      "loss": 5.225,
      "mean_token_accuracy": 0.15932272523641586,
      "num_tokens": 50105907.0,
      "step": 22875
    },
    {
      "entropy": 5.494274616241455,
      "epoch": 2.2707423580786026,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044866322536320447,
      "loss": 5.1893,
      "mean_token_accuracy": 0.16565956324338912,
      "num_tokens": 50117361.0,
      "step": 22880
    },
    {
      "entropy": 5.554203176498413,
      "epoch": 2.27123858674077,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044864069512120185,
      "loss": 5.1982,
      "mean_token_accuracy": 0.16444935128092766,
      "num_tokens": 50128598.0,
      "step": 22885
    },
    {
      "entropy": 5.538686895370484,
      "epoch": 2.2717348154029375,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044861816057328424,
      "loss": 5.1797,
      "mean_token_accuracy": 0.16531308144330978,
      "num_tokens": 50140470.0,
      "step": 22890
    },
    {
      "entropy": 5.520463657379151,
      "epoch": 2.272231044065105,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044859562172001036,
      "loss": 5.2286,
      "mean_token_accuracy": 0.16670664250850678,
      "num_tokens": 50150950.0,
      "step": 22895
    },
    {
      "entropy": 5.503335046768188,
      "epoch": 2.2727272727272725,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004485730785619391,
      "loss": 5.1112,
      "mean_token_accuracy": 0.1676504358649254,
      "num_tokens": 50161173.0,
      "step": 22900
    },
    {
      "entropy": 5.55971999168396,
      "epoch": 2.2732235013894404,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004485505310996295,
      "loss": 5.3161,
      "mean_token_accuracy": 0.15900599658489228,
      "num_tokens": 50173874.0,
      "step": 22905
    },
    {
      "entropy": 5.626149415969849,
      "epoch": 2.273719730051608,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004485279793336406,
      "loss": 5.2605,
      "mean_token_accuracy": 0.1620275631546974,
      "num_tokens": 50185123.0,
      "step": 22910
    },
    {
      "entropy": 5.539774322509766,
      "epoch": 2.2742159587137754,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004485054232645319,
      "loss": 5.2187,
      "mean_token_accuracy": 0.1682631716132164,
      "num_tokens": 50195699.0,
      "step": 22915
    },
    {
      "entropy": 5.572580862045288,
      "epoch": 2.274712187375943,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004484828628928625,
      "loss": 5.2346,
      "mean_token_accuracy": 0.16529457420110702,
      "num_tokens": 50206623.0,
      "step": 22920
    },
    {
      "entropy": 5.5937744140625,
      "epoch": 2.2752084160381103,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044846029821919193,
      "loss": 5.2839,
      "mean_token_accuracy": 0.1573447048664093,
      "num_tokens": 50216086.0,
      "step": 22925
    },
    {
      "entropy": 5.494003582000732,
      "epoch": 2.275704644700278,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004484377292440797,
      "loss": 5.1838,
      "mean_token_accuracy": 0.16632306575775146,
      "num_tokens": 50228472.0,
      "step": 22930
    },
    {
      "entropy": 5.5336864471435545,
      "epoch": 2.2762008733624453,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044841515596808553,
      "loss": 5.2162,
      "mean_token_accuracy": 0.15755362063646317,
      "num_tokens": 50238607.0,
      "step": 22935
    },
    {
      "entropy": 5.523842430114746,
      "epoch": 2.276697102024613,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044839257839176914,
      "loss": 5.1718,
      "mean_token_accuracy": 0.16707295030355454,
      "num_tokens": 50249890.0,
      "step": 22940
    },
    {
      "entropy": 5.534036254882812,
      "epoch": 2.2771933306867806,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044836999651569055,
      "loss": 5.1507,
      "mean_token_accuracy": 0.1684509351849556,
      "num_tokens": 50260234.0,
      "step": 22945
    },
    {
      "entropy": 5.528905725479126,
      "epoch": 2.277689559348948,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004483474103404095,
      "loss": 5.2245,
      "mean_token_accuracy": 0.16490687876939775,
      "num_tokens": 50271605.0,
      "step": 22950
    },
    {
      "entropy": 5.537434148788452,
      "epoch": 2.2781857880111156,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044832481986648633,
      "loss": 5.0707,
      "mean_token_accuracy": 0.16849280297756195,
      "num_tokens": 50281837.0,
      "step": 22955
    },
    {
      "entropy": 5.505376529693604,
      "epoch": 2.278682016673283,
      "grad_norm": 1.265625,
      "learning_rate": 0.000448302225094481,
      "loss": 5.1441,
      "mean_token_accuracy": 0.16414084732532502,
      "num_tokens": 50292524.0,
      "step": 22960
    },
    {
      "entropy": 5.5564182758331295,
      "epoch": 2.2791782453354505,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044827962602495393,
      "loss": 5.1437,
      "mean_token_accuracy": 0.1656780645251274,
      "num_tokens": 50304009.0,
      "step": 22965
    },
    {
      "entropy": 5.509713745117187,
      "epoch": 2.279674473997618,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044825702265846557,
      "loss": 5.2309,
      "mean_token_accuracy": 0.1632654994726181,
      "num_tokens": 50316103.0,
      "step": 22970
    },
    {
      "entropy": 5.494571256637573,
      "epoch": 2.2801707026597855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004482344149955763,
      "loss": 5.2204,
      "mean_token_accuracy": 0.16556442379951478,
      "num_tokens": 50326127.0,
      "step": 22975
    },
    {
      "entropy": 5.467488718032837,
      "epoch": 2.280666931321953,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004482118030368469,
      "loss": 5.1129,
      "mean_token_accuracy": 0.16749798357486725,
      "num_tokens": 50337044.0,
      "step": 22980
    },
    {
      "entropy": 5.530239009857178,
      "epoch": 2.2811631599841204,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000448189186782838,
      "loss": 5.3192,
      "mean_token_accuracy": 0.15741738975048064,
      "num_tokens": 50347796.0,
      "step": 22985
    },
    {
      "entropy": 5.49828953742981,
      "epoch": 2.2816593886462884,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044816656623411044,
      "loss": 5.1958,
      "mean_token_accuracy": 0.16473031789064407,
      "num_tokens": 50358454.0,
      "step": 22990
    },
    {
      "entropy": 5.494074583053589,
      "epoch": 2.282155617308456,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004481439413912252,
      "loss": 5.1208,
      "mean_token_accuracy": 0.16513656973838806,
      "num_tokens": 50369415.0,
      "step": 22995
    },
    {
      "entropy": 5.510201644897461,
      "epoch": 2.2826518459706233,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044812131225474333,
      "loss": 5.2079,
      "mean_token_accuracy": 0.16132623702287674,
      "num_tokens": 50379977.0,
      "step": 23000
    },
    {
      "entropy": 5.551107740402221,
      "epoch": 2.283148074632791,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004480986788252259,
      "loss": 5.2275,
      "mean_token_accuracy": 0.16261208057403564,
      "num_tokens": 50390465.0,
      "step": 23005
    },
    {
      "entropy": 5.566496419906616,
      "epoch": 2.2836443032949583,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004480760411032343,
      "loss": 5.2437,
      "mean_token_accuracy": 0.16154441386461257,
      "num_tokens": 50402230.0,
      "step": 23010
    },
    {
      "entropy": 5.444364356994629,
      "epoch": 2.2841405319571257,
      "grad_norm": 1.125,
      "learning_rate": 0.0004480533990893298,
      "loss": 5.0952,
      "mean_token_accuracy": 0.1708941325545311,
      "num_tokens": 50413819.0,
      "step": 23015
    },
    {
      "entropy": 5.702847576141357,
      "epoch": 2.284636760619293,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004480307527840739,
      "loss": 5.3854,
      "mean_token_accuracy": 0.14839544892311096,
      "num_tokens": 50424977.0,
      "step": 23020
    },
    {
      "entropy": 5.591185188293457,
      "epoch": 2.285132989281461,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004480081021880282,
      "loss": 5.2018,
      "mean_token_accuracy": 0.16406745165586473,
      "num_tokens": 50438357.0,
      "step": 23025
    },
    {
      "entropy": 5.512991237640381,
      "epoch": 2.2856292179436286,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004479854473017543,
      "loss": 5.2008,
      "mean_token_accuracy": 0.16150484830141068,
      "num_tokens": 50449081.0,
      "step": 23030
    },
    {
      "entropy": 5.419689416885376,
      "epoch": 2.286125446605796,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044796278812581417,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1682195246219635,
      "num_tokens": 50460358.0,
      "step": 23035
    },
    {
      "entropy": 5.517882299423218,
      "epoch": 2.2866216752679636,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004479401246607695,
      "loss": 5.1935,
      "mean_token_accuracy": 0.16109024584293366,
      "num_tokens": 50470955.0,
      "step": 23040
    },
    {
      "entropy": 5.546634531021118,
      "epoch": 2.287117903930131,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004479174569071823,
      "loss": 5.1607,
      "mean_token_accuracy": 0.16720037758350373,
      "num_tokens": 50482250.0,
      "step": 23045
    },
    {
      "entropy": 5.527398872375488,
      "epoch": 2.2876141325922985,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004478947848656149,
      "loss": 5.1968,
      "mean_token_accuracy": 0.1602372258901596,
      "num_tokens": 50493868.0,
      "step": 23050
    },
    {
      "entropy": 5.543408441543579,
      "epoch": 2.288110361254466,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044787210853662943,
      "loss": 5.2208,
      "mean_token_accuracy": 0.16457109153270721,
      "num_tokens": 50504012.0,
      "step": 23055
    },
    {
      "entropy": 5.506428956985474,
      "epoch": 2.2886065899166335,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004478494279207881,
      "loss": 5.1456,
      "mean_token_accuracy": 0.16957450807094573,
      "num_tokens": 50515503.0,
      "step": 23060
    },
    {
      "entropy": 5.471718215942383,
      "epoch": 2.289102818578801,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044782674301865346,
      "loss": 5.1731,
      "mean_token_accuracy": 0.16475246995687484,
      "num_tokens": 50526180.0,
      "step": 23065
    },
    {
      "entropy": 5.55273208618164,
      "epoch": 2.2895990472409684,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004478040538307879,
      "loss": 5.1929,
      "mean_token_accuracy": 0.1603195384144783,
      "num_tokens": 50536711.0,
      "step": 23070
    },
    {
      "entropy": 5.587864112854004,
      "epoch": 2.2900952759031363,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004477813603577543,
      "loss": 5.2509,
      "mean_token_accuracy": 0.15663746744394302,
      "num_tokens": 50547736.0,
      "step": 23075
    },
    {
      "entropy": 5.511465740203858,
      "epoch": 2.290591504565304,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044775866260011516,
      "loss": 5.152,
      "mean_token_accuracy": 0.1706218957901001,
      "num_tokens": 50558173.0,
      "step": 23080
    },
    {
      "entropy": 5.537224340438843,
      "epoch": 2.2910877332274713,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004477359605584335,
      "loss": 5.1935,
      "mean_token_accuracy": 0.16558971256017685,
      "num_tokens": 50568421.0,
      "step": 23085
    },
    {
      "entropy": 5.489017009735107,
      "epoch": 2.2915839618896388,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044771325423327227,
      "loss": 5.146,
      "mean_token_accuracy": 0.16722068786621094,
      "num_tokens": 50579182.0,
      "step": 23090
    },
    {
      "entropy": 5.507821083068848,
      "epoch": 2.2920801905518062,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004476905436251945,
      "loss": 5.181,
      "mean_token_accuracy": 0.1685139700770378,
      "num_tokens": 50590916.0,
      "step": 23095
    },
    {
      "entropy": 5.3898697853088375,
      "epoch": 2.2925764192139737,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044766782873476333,
      "loss": 5.0644,
      "mean_token_accuracy": 0.1788877069950104,
      "num_tokens": 50600682.0,
      "step": 23100
    },
    {
      "entropy": 5.604827785491944,
      "epoch": 2.293072647876141,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000447645109562542,
      "loss": 5.3293,
      "mean_token_accuracy": 0.15433455482125283,
      "num_tokens": 50612103.0,
      "step": 23105
    },
    {
      "entropy": 5.625207424163818,
      "epoch": 2.293568876538309,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044762238610909414,
      "loss": 5.2746,
      "mean_token_accuracy": 0.1595138505101204,
      "num_tokens": 50623894.0,
      "step": 23110
    },
    {
      "entropy": 5.570926141738892,
      "epoch": 2.2940651052004766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004475996583749829,
      "loss": 5.2156,
      "mean_token_accuracy": 0.16281801760196685,
      "num_tokens": 50634495.0,
      "step": 23115
    },
    {
      "entropy": 5.588433694839478,
      "epoch": 2.294561333862644,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044757692636077215,
      "loss": 5.3577,
      "mean_token_accuracy": 0.15984580665826797,
      "num_tokens": 50645921.0,
      "step": 23120
    },
    {
      "entropy": 5.6192728042602536,
      "epoch": 2.2950575625248115,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044755419006702545,
      "loss": 5.2097,
      "mean_token_accuracy": 0.1578975647687912,
      "num_tokens": 50657047.0,
      "step": 23125
    },
    {
      "entropy": 5.575136756896972,
      "epoch": 2.295553791186979,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044753144949430667,
      "loss": 5.156,
      "mean_token_accuracy": 0.16231949925422667,
      "num_tokens": 50668271.0,
      "step": 23130
    },
    {
      "entropy": 5.465125226974488,
      "epoch": 2.2960500198491465,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004475087046431798,
      "loss": 5.1238,
      "mean_token_accuracy": 0.1692367047071457,
      "num_tokens": 50678870.0,
      "step": 23135
    },
    {
      "entropy": 5.475437116622925,
      "epoch": 2.296546248511314,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004474859555142087,
      "loss": 5.1765,
      "mean_token_accuracy": 0.1676391452550888,
      "num_tokens": 50690322.0,
      "step": 23140
    },
    {
      "entropy": 5.488492345809936,
      "epoch": 2.2970424771734814,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044746320210795753,
      "loss": 5.1201,
      "mean_token_accuracy": 0.16897793859243393,
      "num_tokens": 50701958.0,
      "step": 23145
    },
    {
      "entropy": 5.4656308650970455,
      "epoch": 2.297538705835649,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004474404444249906,
      "loss": 5.1865,
      "mean_token_accuracy": 0.16683382987976075,
      "num_tokens": 50712962.0,
      "step": 23150
    },
    {
      "entropy": 5.510072565078735,
      "epoch": 2.2980349344978164,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044741768246587226,
      "loss": 5.2188,
      "mean_token_accuracy": 0.1677071586251259,
      "num_tokens": 50723442.0,
      "step": 23155
    },
    {
      "entropy": 5.5159416675567625,
      "epoch": 2.2985311631599843,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004473949162311669,
      "loss": 5.084,
      "mean_token_accuracy": 0.1706620842218399,
      "num_tokens": 50733188.0,
      "step": 23160
    },
    {
      "entropy": 5.529516506195068,
      "epoch": 2.2990273918221518,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004473721457214391,
      "loss": 5.2433,
      "mean_token_accuracy": 0.15896586030721666,
      "num_tokens": 50744023.0,
      "step": 23165
    },
    {
      "entropy": 5.632857084274292,
      "epoch": 2.2995236204843192,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004473493709372535,
      "loss": 5.4267,
      "mean_token_accuracy": 0.15596348494291307,
      "num_tokens": 50756807.0,
      "step": 23170
    },
    {
      "entropy": 5.577497911453247,
      "epoch": 2.3000198491464867,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004473265918791749,
      "loss": 5.2879,
      "mean_token_accuracy": 0.15598273724317552,
      "num_tokens": 50768981.0,
      "step": 23175
    },
    {
      "entropy": 5.574667024612427,
      "epoch": 2.300516077808654,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004473038085477682,
      "loss": 5.2032,
      "mean_token_accuracy": 0.16779337972402572,
      "num_tokens": 50780111.0,
      "step": 23180
    },
    {
      "entropy": 5.548598289489746,
      "epoch": 2.3010123064708217,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044728102094359826,
      "loss": 5.2116,
      "mean_token_accuracy": 0.16476257145404816,
      "num_tokens": 50790318.0,
      "step": 23185
    },
    {
      "entropy": 5.517019748687744,
      "epoch": 2.301508535132989,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004472582290672303,
      "loss": 5.2536,
      "mean_token_accuracy": 0.15986888110637665,
      "num_tokens": 50800734.0,
      "step": 23190
    },
    {
      "entropy": 5.458087730407715,
      "epoch": 2.3020047637951566,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044723543291922936,
      "loss": 5.139,
      "mean_token_accuracy": 0.17335109710693358,
      "num_tokens": 50811236.0,
      "step": 23195
    },
    {
      "entropy": 5.479608583450317,
      "epoch": 2.3025009924573245,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044721263250016086,
      "loss": 5.1608,
      "mean_token_accuracy": 0.16454117000102997,
      "num_tokens": 50821776.0,
      "step": 23200
    },
    {
      "entropy": 5.593709135055542,
      "epoch": 2.302997221119492,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044718982781059017,
      "loss": 5.2943,
      "mean_token_accuracy": 0.15888661593198777,
      "num_tokens": 50834156.0,
      "step": 23205
    },
    {
      "entropy": 5.56753134727478,
      "epoch": 2.3034934497816595,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044716701885108275,
      "loss": 5.1737,
      "mean_token_accuracy": 0.16384712606668472,
      "num_tokens": 50845086.0,
      "step": 23210
    },
    {
      "entropy": 5.549552631378174,
      "epoch": 2.303989678443827,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044714420562220423,
      "loss": 5.2005,
      "mean_token_accuracy": 0.16435045301914214,
      "num_tokens": 50858637.0,
      "step": 23215
    },
    {
      "entropy": 5.545665121078491,
      "epoch": 2.3044859071059944,
      "grad_norm": 1.125,
      "learning_rate": 0.0004471213881245204,
      "loss": 5.3001,
      "mean_token_accuracy": 0.15856747180223466,
      "num_tokens": 50871159.0,
      "step": 23220
    },
    {
      "entropy": 5.517868852615356,
      "epoch": 2.304982135768162,
      "grad_norm": 1.3125,
      "learning_rate": 0.000447098566358597,
      "loss": 5.172,
      "mean_token_accuracy": 0.16628472208976747,
      "num_tokens": 50884072.0,
      "step": 23225
    },
    {
      "entropy": 5.483020639419555,
      "epoch": 2.3054783644303294,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044707574032500005,
      "loss": 5.1816,
      "mean_token_accuracy": 0.16826854795217513,
      "num_tokens": 50896627.0,
      "step": 23230
    },
    {
      "entropy": 5.458249759674072,
      "epoch": 2.305974593092497,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004470529100242955,
      "loss": 5.0946,
      "mean_token_accuracy": 0.17024115175008775,
      "num_tokens": 50907476.0,
      "step": 23235
    },
    {
      "entropy": 5.479052782058716,
      "epoch": 2.3064708217546643,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004470300754570495,
      "loss": 5.1934,
      "mean_token_accuracy": 0.16521177291870118,
      "num_tokens": 50918104.0,
      "step": 23240
    },
    {
      "entropy": 5.510082101821899,
      "epoch": 2.3069670504168323,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004470072366238284,
      "loss": 5.1561,
      "mean_token_accuracy": 0.16520793735980988,
      "num_tokens": 50930575.0,
      "step": 23245
    },
    {
      "entropy": 5.516611862182617,
      "epoch": 2.3074632790789997,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044698439352519836,
      "loss": 5.1178,
      "mean_token_accuracy": 0.165229956805706,
      "num_tokens": 50940592.0,
      "step": 23250
    },
    {
      "entropy": 5.454147720336914,
      "epoch": 2.307959507741167,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044696154616172604,
      "loss": 5.1828,
      "mean_token_accuracy": 0.16399523466825486,
      "num_tokens": 50952134.0,
      "step": 23255
    },
    {
      "entropy": 5.522246932983398,
      "epoch": 2.3084557364033347,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004469386945339779,
      "loss": 5.2332,
      "mean_token_accuracy": 0.16181970238685608,
      "num_tokens": 50962721.0,
      "step": 23260
    },
    {
      "entropy": 5.48744387626648,
      "epoch": 2.308951965065502,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004469158386425206,
      "loss": 5.1468,
      "mean_token_accuracy": 0.1635687083005905,
      "num_tokens": 50974893.0,
      "step": 23265
    },
    {
      "entropy": 5.497821283340454,
      "epoch": 2.3094481937276696,
      "grad_norm": 1.046875,
      "learning_rate": 0.000446892978487921,
      "loss": 5.1374,
      "mean_token_accuracy": 0.16366032361984253,
      "num_tokens": 50986225.0,
      "step": 23270
    },
    {
      "entropy": 5.51474027633667,
      "epoch": 2.309944422389837,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044687011407074583,
      "loss": 5.2305,
      "mean_token_accuracy": 0.1554247796535492,
      "num_tokens": 50997924.0,
      "step": 23275
    },
    {
      "entropy": 5.505477619171143,
      "epoch": 2.3104406510520046,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004468472453915623,
      "loss": 5.1046,
      "mean_token_accuracy": 0.16592035591602325,
      "num_tokens": 51008179.0,
      "step": 23280
    },
    {
      "entropy": 5.44039568901062,
      "epoch": 2.3109368797141725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004468243724509373,
      "loss": 5.1283,
      "mean_token_accuracy": 0.17302992045879365,
      "num_tokens": 51018462.0,
      "step": 23285
    },
    {
      "entropy": 5.546918869018555,
      "epoch": 2.31143310837634,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004468014952494382,
      "loss": 5.235,
      "mean_token_accuracy": 0.160286109149456,
      "num_tokens": 51029769.0,
      "step": 23290
    },
    {
      "entropy": 5.473975419998169,
      "epoch": 2.3119293370385074,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044677861378763205,
      "loss": 5.2178,
      "mean_token_accuracy": 0.1636495292186737,
      "num_tokens": 51041058.0,
      "step": 23295
    },
    {
      "entropy": 5.457054281234742,
      "epoch": 2.312425565700675,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044675572806608655,
      "loss": 5.14,
      "mean_token_accuracy": 0.16908894777297973,
      "num_tokens": 51050958.0,
      "step": 23300
    },
    {
      "entropy": 5.577691173553466,
      "epoch": 2.3129217943628424,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044673283808536915,
      "loss": 5.1521,
      "mean_token_accuracy": 0.17401891350746154,
      "num_tokens": 51062459.0,
      "step": 23305
    },
    {
      "entropy": 5.574757719039917,
      "epoch": 2.31341802302501,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004467099438460474,
      "loss": 5.3524,
      "mean_token_accuracy": 0.1588505432009697,
      "num_tokens": 51074596.0,
      "step": 23310
    },
    {
      "entropy": 5.569075632095337,
      "epoch": 2.3139142516871773,
      "grad_norm": 1.5,
      "learning_rate": 0.00044668704534868886,
      "loss": 5.2085,
      "mean_token_accuracy": 0.15863564610481262,
      "num_tokens": 51085265.0,
      "step": 23315
    },
    {
      "entropy": 5.546475791931153,
      "epoch": 2.314410480349345,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044666414259386173,
      "loss": 5.1559,
      "mean_token_accuracy": 0.1618427023291588,
      "num_tokens": 51095986.0,
      "step": 23320
    },
    {
      "entropy": 5.513752698898315,
      "epoch": 2.3149067090115123,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044664123558213367,
      "loss": 5.1757,
      "mean_token_accuracy": 0.16042264848947524,
      "num_tokens": 51107518.0,
      "step": 23325
    },
    {
      "entropy": 5.467346096038819,
      "epoch": 2.31540293767368,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004466183243140729,
      "loss": 5.117,
      "mean_token_accuracy": 0.17237696349620818,
      "num_tokens": 51119715.0,
      "step": 23330
    },
    {
      "entropy": 5.413383102416992,
      "epoch": 2.3158991663358477,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044659540879024746,
      "loss": 5.1432,
      "mean_token_accuracy": 0.1677184894680977,
      "num_tokens": 51130801.0,
      "step": 23335
    },
    {
      "entropy": 5.459241151809692,
      "epoch": 2.316395394998015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004465724890112256,
      "loss": 5.1655,
      "mean_token_accuracy": 0.16505540311336517,
      "num_tokens": 51142509.0,
      "step": 23340
    },
    {
      "entropy": 5.611122274398804,
      "epoch": 2.3168916236601826,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044654956497757576,
      "loss": 5.2158,
      "mean_token_accuracy": 0.1658434808254242,
      "num_tokens": 51154839.0,
      "step": 23345
    },
    {
      "entropy": 5.5215777397155765,
      "epoch": 2.31738785232235,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044652663668986634,
      "loss": 5.1505,
      "mean_token_accuracy": 0.1739455506205559,
      "num_tokens": 51165293.0,
      "step": 23350
    },
    {
      "entropy": 5.487966537475586,
      "epoch": 2.3178840809845176,
      "grad_norm": 1.171875,
      "learning_rate": 0.000446503704148666,
      "loss": 5.2279,
      "mean_token_accuracy": 0.16109009981155395,
      "num_tokens": 51176371.0,
      "step": 23355
    },
    {
      "entropy": 5.4919170379638675,
      "epoch": 2.318380309646685,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004464807673545433,
      "loss": 5.146,
      "mean_token_accuracy": 0.17582867592573165,
      "num_tokens": 51187240.0,
      "step": 23360
    },
    {
      "entropy": 5.576871013641357,
      "epoch": 2.3188765383088525,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004464578263080671,
      "loss": 5.2483,
      "mean_token_accuracy": 0.16261623427271843,
      "num_tokens": 51199084.0,
      "step": 23365
    },
    {
      "entropy": 5.484199190139771,
      "epoch": 2.3193727669710205,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044643488100980623,
      "loss": 5.1127,
      "mean_token_accuracy": 0.16749345362186432,
      "num_tokens": 51209885.0,
      "step": 23370
    },
    {
      "entropy": 5.516084480285644,
      "epoch": 2.319868995633188,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004464119314603298,
      "loss": 5.1687,
      "mean_token_accuracy": 0.1671769365668297,
      "num_tokens": 51220644.0,
      "step": 23375
    },
    {
      "entropy": 5.48397741317749,
      "epoch": 2.3203652242953554,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044638897766020673,
      "loss": 5.1428,
      "mean_token_accuracy": 0.16626554280519484,
      "num_tokens": 51231342.0,
      "step": 23380
    },
    {
      "entropy": 5.406710147857666,
      "epoch": 2.320861452957523,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004463660196100064,
      "loss": 5.0613,
      "mean_token_accuracy": 0.16973731517791749,
      "num_tokens": 51241368.0,
      "step": 23385
    },
    {
      "entropy": 5.516503477096558,
      "epoch": 2.3213576816196904,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044634305731029803,
      "loss": 5.2691,
      "mean_token_accuracy": 0.16407191455364228,
      "num_tokens": 51252784.0,
      "step": 23390
    },
    {
      "entropy": 5.547989130020142,
      "epoch": 2.321853910281858,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044632009076165096,
      "loss": 5.1391,
      "mean_token_accuracy": 0.16711972802877426,
      "num_tokens": 51264246.0,
      "step": 23395
    },
    {
      "entropy": 5.467084121704102,
      "epoch": 2.3223501389440253,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000446297119964635,
      "loss": 5.0942,
      "mean_token_accuracy": 0.1734147474169731,
      "num_tokens": 51274575.0,
      "step": 23400
    },
    {
      "entropy": 5.494536542892456,
      "epoch": 2.322846367606193,
      "grad_norm": 1.25,
      "learning_rate": 0.0004462741449198194,
      "loss": 5.1475,
      "mean_token_accuracy": 0.16127973347902297,
      "num_tokens": 51285384.0,
      "step": 23405
    },
    {
      "entropy": 5.542532539367675,
      "epoch": 2.3233425962683603,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044625116562777413,
      "loss": 5.22,
      "mean_token_accuracy": 0.1681498944759369,
      "num_tokens": 51297983.0,
      "step": 23410
    },
    {
      "entropy": 5.4409102439880375,
      "epoch": 2.323838824930528,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004462281820890689,
      "loss": 5.1231,
      "mean_token_accuracy": 0.17250384539365768,
      "num_tokens": 51308795.0,
      "step": 23415
    },
    {
      "entropy": 5.47532958984375,
      "epoch": 2.3243350535926957,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044620519430427376,
      "loss": 5.1317,
      "mean_token_accuracy": 0.17116790860891343,
      "num_tokens": 51318597.0,
      "step": 23420
    },
    {
      "entropy": 5.4592360496521,
      "epoch": 2.324831282254863,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004461822022739588,
      "loss": 5.0747,
      "mean_token_accuracy": 0.16997887641191484,
      "num_tokens": 51329238.0,
      "step": 23425
    },
    {
      "entropy": 5.524947166442871,
      "epoch": 2.3253275109170306,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000446159205998694,
      "loss": 5.1855,
      "mean_token_accuracy": 0.16574927121400834,
      "num_tokens": 51340295.0,
      "step": 23430
    },
    {
      "entropy": 5.483934307098389,
      "epoch": 2.325823739579198,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004461362054790497,
      "loss": 5.1646,
      "mean_token_accuracy": 0.1659070372581482,
      "num_tokens": 51350702.0,
      "step": 23435
    },
    {
      "entropy": 5.542941904067993,
      "epoch": 2.3263199682413656,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044611320071559637,
      "loss": 5.1604,
      "mean_token_accuracy": 0.16908841878175734,
      "num_tokens": 51359794.0,
      "step": 23440
    },
    {
      "entropy": 5.520611047744751,
      "epoch": 2.326816196903533,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044609019170890424,
      "loss": 5.1342,
      "mean_token_accuracy": 0.16705866008996964,
      "num_tokens": 51368932.0,
      "step": 23445
    },
    {
      "entropy": 5.527321672439575,
      "epoch": 2.3273124255657005,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044606717845954415,
      "loss": 5.2214,
      "mean_token_accuracy": 0.16576973497867584,
      "num_tokens": 51381111.0,
      "step": 23450
    },
    {
      "entropy": 5.482883548736572,
      "epoch": 2.3278086542278684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044604416096808656,
      "loss": 5.1406,
      "mean_token_accuracy": 0.17759720981121063,
      "num_tokens": 51392976.0,
      "step": 23455
    },
    {
      "entropy": 5.509830284118652,
      "epoch": 2.328304882890036,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044602113923510235,
      "loss": 5.1149,
      "mean_token_accuracy": 0.16722202599048613,
      "num_tokens": 51403586.0,
      "step": 23460
    },
    {
      "entropy": 5.496737241744995,
      "epoch": 2.3288011115522034,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044599811326116246,
      "loss": 5.2184,
      "mean_token_accuracy": 0.161575610935688,
      "num_tokens": 51414035.0,
      "step": 23465
    },
    {
      "entropy": 5.502129602432251,
      "epoch": 2.329297340214371,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044597508304683784,
      "loss": 5.1987,
      "mean_token_accuracy": 0.1647410809993744,
      "num_tokens": 51424111.0,
      "step": 23470
    },
    {
      "entropy": 5.4829357147216795,
      "epoch": 2.3297935688765383,
      "grad_norm": 1.125,
      "learning_rate": 0.0004459520485926995,
      "loss": 5.1258,
      "mean_token_accuracy": 0.17445645332336426,
      "num_tokens": 51434566.0,
      "step": 23475
    },
    {
      "entropy": 5.513744449615478,
      "epoch": 2.330289797538706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004459290098993188,
      "loss": 5.1855,
      "mean_token_accuracy": 0.1661531463265419,
      "num_tokens": 51445568.0,
      "step": 23480
    },
    {
      "entropy": 5.509911108016968,
      "epoch": 2.3307860262008733,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044590596696726686,
      "loss": 5.1603,
      "mean_token_accuracy": 0.1736524611711502,
      "num_tokens": 51456482.0,
      "step": 23485
    },
    {
      "entropy": 5.509155988693237,
      "epoch": 2.3312822548630407,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004458829197971153,
      "loss": 5.2396,
      "mean_token_accuracy": 0.16105173975229264,
      "num_tokens": 51467809.0,
      "step": 23490
    },
    {
      "entropy": 5.542853832244873,
      "epoch": 2.3317784835252082,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044585986838943556,
      "loss": 5.176,
      "mean_token_accuracy": 0.16123056784272194,
      "num_tokens": 51477927.0,
      "step": 23495
    },
    {
      "entropy": 5.5057494163513185,
      "epoch": 2.3322747121873757,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044583681274479927,
      "loss": 5.1963,
      "mean_token_accuracy": 0.16560157239437104,
      "num_tokens": 51488432.0,
      "step": 23500
    },
    {
      "entropy": 5.527063941955566,
      "epoch": 2.3327709408495436,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004458137528637781,
      "loss": 5.1716,
      "mean_token_accuracy": 0.16758323460817337,
      "num_tokens": 51499149.0,
      "step": 23505
    },
    {
      "entropy": 5.5628923892974855,
      "epoch": 2.333267169511711,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044579068874694386,
      "loss": 5.2178,
      "mean_token_accuracy": 0.15666577368974685,
      "num_tokens": 51510672.0,
      "step": 23510
    },
    {
      "entropy": 5.514298725128174,
      "epoch": 2.3337633981738786,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004457676203948687,
      "loss": 5.2005,
      "mean_token_accuracy": 0.16708512753248214,
      "num_tokens": 51521855.0,
      "step": 23515
    },
    {
      "entropy": 5.549034500122071,
      "epoch": 2.334259626836046,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004457445478081244,
      "loss": 5.1673,
      "mean_token_accuracy": 0.16862370520830156,
      "num_tokens": 51531970.0,
      "step": 23520
    },
    {
      "entropy": 5.520298147201538,
      "epoch": 2.3347558554982135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044572147098728335,
      "loss": 5.2399,
      "mean_token_accuracy": 0.16543546169996262,
      "num_tokens": 51543110.0,
      "step": 23525
    },
    {
      "entropy": 5.488470125198364,
      "epoch": 2.335252084160381,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044569838993291757,
      "loss": 5.1905,
      "mean_token_accuracy": 0.16190362274646758,
      "num_tokens": 51552991.0,
      "step": 23530
    },
    {
      "entropy": 5.533665513992309,
      "epoch": 2.3357483128225485,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044567530464559963,
      "loss": 5.2266,
      "mean_token_accuracy": 0.1675412654876709,
      "num_tokens": 51564472.0,
      "step": 23535
    },
    {
      "entropy": 5.490886878967285,
      "epoch": 2.3362445414847164,
      "grad_norm": 1.125,
      "learning_rate": 0.0004456522151259019,
      "loss": 5.212,
      "mean_token_accuracy": 0.1662023261189461,
      "num_tokens": 51576051.0,
      "step": 23540
    },
    {
      "entropy": 5.585345840454101,
      "epoch": 2.336740770146884,
      "grad_norm": 1.1875,
      "learning_rate": 0.000445629121374397,
      "loss": 5.2049,
      "mean_token_accuracy": 0.15972771346569062,
      "num_tokens": 51586658.0,
      "step": 23545
    },
    {
      "entropy": 5.479167079925537,
      "epoch": 2.3372369988090513,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044560602339165746,
      "loss": 5.1662,
      "mean_token_accuracy": 0.16474380046129228,
      "num_tokens": 51596614.0,
      "step": 23550
    },
    {
      "entropy": 5.515455293655395,
      "epoch": 2.337733227471219,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004455829211782562,
      "loss": 5.196,
      "mean_token_accuracy": 0.16765576899051665,
      "num_tokens": 51607102.0,
      "step": 23555
    },
    {
      "entropy": 5.609760189056397,
      "epoch": 2.3382294561333863,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004455598147347661,
      "loss": 5.2585,
      "mean_token_accuracy": 0.16184215545654296,
      "num_tokens": 51619079.0,
      "step": 23560
    },
    {
      "entropy": 5.5110859870910645,
      "epoch": 2.3387256847955538,
      "grad_norm": 1.125,
      "learning_rate": 0.0004455367040617601,
      "loss": 5.1031,
      "mean_token_accuracy": 0.16969509422779083,
      "num_tokens": 51630273.0,
      "step": 23565
    },
    {
      "entropy": 5.483106899261474,
      "epoch": 2.3392219134577212,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044551358915981135,
      "loss": 5.1625,
      "mean_token_accuracy": 0.16495077908039094,
      "num_tokens": 51640480.0,
      "step": 23570
    },
    {
      "entropy": 5.557659482955932,
      "epoch": 2.3397181421198887,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004454904700294929,
      "loss": 5.2173,
      "mean_token_accuracy": 0.1671823412179947,
      "num_tokens": 51651831.0,
      "step": 23575
    },
    {
      "entropy": 5.529864645004272,
      "epoch": 2.340214370782056,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004454673466713783,
      "loss": 5.2152,
      "mean_token_accuracy": 0.1621822088956833,
      "num_tokens": 51664061.0,
      "step": 23580
    },
    {
      "entropy": 5.500938129425049,
      "epoch": 2.3407105994442237,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044544421908604073,
      "loss": 5.1721,
      "mean_token_accuracy": 0.17073223292827605,
      "num_tokens": 51675081.0,
      "step": 23585
    },
    {
      "entropy": 5.476324605941772,
      "epoch": 2.3412068281063916,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044542108727405386,
      "loss": 5.0647,
      "mean_token_accuracy": 0.17367087453603744,
      "num_tokens": 51684477.0,
      "step": 23590
    },
    {
      "entropy": 5.568332147598267,
      "epoch": 2.341703056768559,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044539795123599117,
      "loss": 5.2182,
      "mean_token_accuracy": 0.16266769766807557,
      "num_tokens": 51694420.0,
      "step": 23595
    },
    {
      "entropy": 5.366077470779419,
      "epoch": 2.3421992854307265,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044537481097242654,
      "loss": 5.0681,
      "mean_token_accuracy": 0.17389643043279648,
      "num_tokens": 51704635.0,
      "step": 23600
    },
    {
      "entropy": 5.554231405258179,
      "epoch": 2.342695514092894,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00044535166648393377,
      "loss": 5.2874,
      "mean_token_accuracy": 0.15890335142612458,
      "num_tokens": 51716149.0,
      "step": 23605
    },
    {
      "entropy": 5.486980724334717,
      "epoch": 2.3431917427550615,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044532851777108663,
      "loss": 5.1237,
      "mean_token_accuracy": 0.17132244855165482,
      "num_tokens": 51726861.0,
      "step": 23610
    },
    {
      "entropy": 5.536518144607544,
      "epoch": 2.343687971417229,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004453053648344593,
      "loss": 5.3609,
      "mean_token_accuracy": 0.1551014557480812,
      "num_tokens": 51738111.0,
      "step": 23615
    },
    {
      "entropy": 5.468129730224609,
      "epoch": 2.3441842000793964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044528220767462595,
      "loss": 5.1868,
      "mean_token_accuracy": 0.16392059922218322,
      "num_tokens": 51749004.0,
      "step": 23620
    },
    {
      "entropy": 5.503826427459717,
      "epoch": 2.3446804287415643,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004452590462921607,
      "loss": 5.158,
      "mean_token_accuracy": 0.1619345486164093,
      "num_tokens": 51759557.0,
      "step": 23625
    },
    {
      "entropy": 5.553966522216797,
      "epoch": 2.345176657403732,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044523588068763804,
      "loss": 5.2313,
      "mean_token_accuracy": 0.1616302788257599,
      "num_tokens": 51769921.0,
      "step": 23630
    },
    {
      "entropy": 5.4331374168396,
      "epoch": 2.3456728860658993,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004452127108616324,
      "loss": 5.084,
      "mean_token_accuracy": 0.16962784081697463,
      "num_tokens": 51780158.0,
      "step": 23635
    },
    {
      "entropy": 5.551636743545532,
      "epoch": 2.3461691147280668,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044518953681471824,
      "loss": 5.1886,
      "mean_token_accuracy": 0.16603257060050963,
      "num_tokens": 51790502.0,
      "step": 23640
    },
    {
      "entropy": 5.530608654022217,
      "epoch": 2.3466653433902342,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004451663585474703,
      "loss": 5.2052,
      "mean_token_accuracy": 0.1594868317246437,
      "num_tokens": 51801420.0,
      "step": 23645
    },
    {
      "entropy": 5.525176620483398,
      "epoch": 2.3471615720524017,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044514317606046336,
      "loss": 5.1997,
      "mean_token_accuracy": 0.17168948650360108,
      "num_tokens": 51812399.0,
      "step": 23650
    },
    {
      "entropy": 5.574924325942993,
      "epoch": 2.347657800714569,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004451199893542723,
      "loss": 5.2694,
      "mean_token_accuracy": 0.15996618270874025,
      "num_tokens": 51824268.0,
      "step": 23655
    },
    {
      "entropy": 5.481704187393189,
      "epoch": 2.3481540293767367,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004450967984294721,
      "loss": 5.1931,
      "mean_token_accuracy": 0.16419116407632828,
      "num_tokens": 51835842.0,
      "step": 23660
    },
    {
      "entropy": 5.508743667602539,
      "epoch": 2.348650258038904,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004450736032866377,
      "loss": 5.13,
      "mean_token_accuracy": 0.16764180064201356,
      "num_tokens": 51847501.0,
      "step": 23665
    },
    {
      "entropy": 5.459465265274048,
      "epoch": 2.3491464867010716,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044505040392634447,
      "loss": 5.1785,
      "mean_token_accuracy": 0.17465372681617736,
      "num_tokens": 51858056.0,
      "step": 23670
    },
    {
      "entropy": 5.542801189422607,
      "epoch": 2.3496427153632395,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004450272003491677,
      "loss": 5.1638,
      "mean_token_accuracy": 0.16871362775564194,
      "num_tokens": 51869248.0,
      "step": 23675
    },
    {
      "entropy": 5.497925758361816,
      "epoch": 2.350138944025407,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044500399255568274,
      "loss": 5.2265,
      "mean_token_accuracy": 0.16237985342741013,
      "num_tokens": 51881496.0,
      "step": 23680
    },
    {
      "entropy": 5.497312593460083,
      "epoch": 2.3506351726875745,
      "grad_norm": 1.421875,
      "learning_rate": 0.00044498078054646507,
      "loss": 5.1686,
      "mean_token_accuracy": 0.16664096117019653,
      "num_tokens": 51891851.0,
      "step": 23685
    },
    {
      "entropy": 5.526036548614502,
      "epoch": 2.351131401349742,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044495756432209025,
      "loss": 5.2541,
      "mean_token_accuracy": 0.164555361866951,
      "num_tokens": 51903581.0,
      "step": 23690
    },
    {
      "entropy": 5.457759857177734,
      "epoch": 2.3516276300119094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044493434388313423,
      "loss": 5.0941,
      "mean_token_accuracy": 0.171679125726223,
      "num_tokens": 51913650.0,
      "step": 23695
    },
    {
      "entropy": 5.453234434127808,
      "epoch": 2.352123858674077,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044491111923017254,
      "loss": 5.1102,
      "mean_token_accuracy": 0.17264010459184648,
      "num_tokens": 51924704.0,
      "step": 23700
    },
    {
      "entropy": 5.575533676147461,
      "epoch": 2.3526200873362444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044488789036378117,
      "loss": 5.2643,
      "mean_token_accuracy": 0.16296306252479553,
      "num_tokens": 51937576.0,
      "step": 23705
    },
    {
      "entropy": 5.508750343322754,
      "epoch": 2.3531163159984123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004448646572845363,
      "loss": 5.0891,
      "mean_token_accuracy": 0.16791151314973832,
      "num_tokens": 51947359.0,
      "step": 23710
    },
    {
      "entropy": 5.512045574188233,
      "epoch": 2.35361254466058,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004448414199930139,
      "loss": 5.2271,
      "mean_token_accuracy": 0.15906078219413758,
      "num_tokens": 51958016.0,
      "step": 23715
    },
    {
      "entropy": 5.573837184906006,
      "epoch": 2.3541087733227473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044481817848979026,
      "loss": 5.2211,
      "mean_token_accuracy": 0.161355459690094,
      "num_tokens": 51968770.0,
      "step": 23720
    },
    {
      "entropy": 5.474053287506104,
      "epoch": 2.3546050019849147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004447949327754418,
      "loss": 5.0978,
      "mean_token_accuracy": 0.17149334400892258,
      "num_tokens": 51980052.0,
      "step": 23725
    },
    {
      "entropy": 5.540418100357056,
      "epoch": 2.355101230647082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004447716828505447,
      "loss": 5.2709,
      "mean_token_accuracy": 0.16292626410722733,
      "num_tokens": 51992080.0,
      "step": 23730
    },
    {
      "entropy": 5.508413982391358,
      "epoch": 2.3555974593092497,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004447484287156758,
      "loss": 5.2047,
      "mean_token_accuracy": 0.17347456365823746,
      "num_tokens": 52004313.0,
      "step": 23735
    },
    {
      "entropy": 5.577845668792724,
      "epoch": 2.356093687971417,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004447251703714117,
      "loss": 5.1747,
      "mean_token_accuracy": 0.16130775362253189,
      "num_tokens": 52015207.0,
      "step": 23740
    },
    {
      "entropy": 5.51321063041687,
      "epoch": 2.3565899166335846,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000444701907818329,
      "loss": 5.1738,
      "mean_token_accuracy": 0.15786434561014176,
      "num_tokens": 52026565.0,
      "step": 23745
    },
    {
      "entropy": 5.522815942764282,
      "epoch": 2.357086145295752,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004446786410570048,
      "loss": 5.2076,
      "mean_token_accuracy": 0.16164450645446776,
      "num_tokens": 52037654.0,
      "step": 23750
    },
    {
      "entropy": 5.483359718322754,
      "epoch": 2.3575823739579196,
      "grad_norm": 1.421875,
      "learning_rate": 0.00044465537008801575,
      "loss": 5.1224,
      "mean_token_accuracy": 0.16779845058918,
      "num_tokens": 52049772.0,
      "step": 23755
    },
    {
      "entropy": 5.4951197624206545,
      "epoch": 2.3580786026200875,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004446320949119392,
      "loss": 5.233,
      "mean_token_accuracy": 0.16528684198856353,
      "num_tokens": 52061363.0,
      "step": 23760
    },
    {
      "entropy": 5.511103105545044,
      "epoch": 2.358574831282255,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044460881552935215,
      "loss": 5.1195,
      "mean_token_accuracy": 0.16895556598901748,
      "num_tokens": 52073153.0,
      "step": 23765
    },
    {
      "entropy": 5.443757152557373,
      "epoch": 2.3590710599444225,
      "grad_norm": 1.65625,
      "learning_rate": 0.00044458553194083204,
      "loss": 5.1456,
      "mean_token_accuracy": 0.16925849616527558,
      "num_tokens": 52083369.0,
      "step": 23770
    },
    {
      "entropy": 5.4413618564605715,
      "epoch": 2.35956728860659,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004445622441469562,
      "loss": 5.0826,
      "mean_token_accuracy": 0.17364209443330764,
      "num_tokens": 52093701.0,
      "step": 23775
    },
    {
      "entropy": 5.4845386981964115,
      "epoch": 2.3600635172687574,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044453895214830195,
      "loss": 5.1572,
      "mean_token_accuracy": 0.16576993763446807,
      "num_tokens": 52105863.0,
      "step": 23780
    },
    {
      "entropy": 5.434888696670532,
      "epoch": 2.360559745930925,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004445156559454471,
      "loss": 5.1324,
      "mean_token_accuracy": 0.1680961474776268,
      "num_tokens": 52116476.0,
      "step": 23785
    },
    {
      "entropy": 5.588758659362793,
      "epoch": 2.3610559745930924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044449235553896914,
      "loss": 5.2635,
      "mean_token_accuracy": 0.15957557708024978,
      "num_tokens": 52127262.0,
      "step": 23790
    },
    {
      "entropy": 5.552884149551391,
      "epoch": 2.36155220325526,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004444690509294461,
      "loss": 5.2019,
      "mean_token_accuracy": 0.16344035565853118,
      "num_tokens": 52138575.0,
      "step": 23795
    },
    {
      "entropy": 5.486243391036988,
      "epoch": 2.3620484319174277,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004444457421174558,
      "loss": 5.1702,
      "mean_token_accuracy": 0.1687913492321968,
      "num_tokens": 52149620.0,
      "step": 23800
    },
    {
      "entropy": 5.560820817947388,
      "epoch": 2.3625446605795952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004444224291035761,
      "loss": 5.2119,
      "mean_token_accuracy": 0.16276225447654724,
      "num_tokens": 52161929.0,
      "step": 23805
    },
    {
      "entropy": 5.455647706985474,
      "epoch": 2.3630408892417627,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044439911188838533,
      "loss": 5.1078,
      "mean_token_accuracy": 0.17356567829847336,
      "num_tokens": 52172283.0,
      "step": 23810
    },
    {
      "entropy": 5.489011669158936,
      "epoch": 2.36353711790393,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004443757904724615,
      "loss": 5.2414,
      "mean_token_accuracy": 0.16601398438215256,
      "num_tokens": 52184130.0,
      "step": 23815
    },
    {
      "entropy": 5.497793912887573,
      "epoch": 2.3640333465660976,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044435246485638317,
      "loss": 5.2066,
      "mean_token_accuracy": 0.16756702214479446,
      "num_tokens": 52195358.0,
      "step": 23820
    },
    {
      "entropy": 5.440282106399536,
      "epoch": 2.364529575228265,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004443291350407285,
      "loss": 5.1038,
      "mean_token_accuracy": 0.1661811113357544,
      "num_tokens": 52207070.0,
      "step": 23825
    },
    {
      "entropy": 5.491967439651489,
      "epoch": 2.3650258038904326,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004443058010260762,
      "loss": 5.235,
      "mean_token_accuracy": 0.16397435814142228,
      "num_tokens": 52217658.0,
      "step": 23830
    },
    {
      "entropy": 5.481918859481811,
      "epoch": 2.3655220325526,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044428246281300493,
      "loss": 5.144,
      "mean_token_accuracy": 0.16817955523729325,
      "num_tokens": 52229350.0,
      "step": 23835
    },
    {
      "entropy": 5.598840188980103,
      "epoch": 2.3660182612147675,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044425912040209326,
      "loss": 5.2919,
      "mean_token_accuracy": 0.16082797348499298,
      "num_tokens": 52241260.0,
      "step": 23840
    },
    {
      "entropy": 5.599735832214355,
      "epoch": 2.3665144898769355,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044423577379392023,
      "loss": 5.1999,
      "mean_token_accuracy": 0.162227064371109,
      "num_tokens": 52253049.0,
      "step": 23845
    },
    {
      "entropy": 5.47116494178772,
      "epoch": 2.367010718539103,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004442124229890646,
      "loss": 5.1683,
      "mean_token_accuracy": 0.16543960124254226,
      "num_tokens": 52263405.0,
      "step": 23850
    },
    {
      "entropy": 5.493868446350097,
      "epoch": 2.3675069472012704,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044418906798810554,
      "loss": 5.1879,
      "mean_token_accuracy": 0.16330794990062714,
      "num_tokens": 52273075.0,
      "step": 23855
    },
    {
      "entropy": 5.621155071258545,
      "epoch": 2.368003175863438,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00044416570879162213,
      "loss": 5.4135,
      "mean_token_accuracy": 0.15600521266460418,
      "num_tokens": 52284275.0,
      "step": 23860
    },
    {
      "entropy": 5.555344820022583,
      "epoch": 2.3684994045256054,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004441423454001936,
      "loss": 5.207,
      "mean_token_accuracy": 0.16029175817966462,
      "num_tokens": 52295084.0,
      "step": 23865
    },
    {
      "entropy": 5.482954216003418,
      "epoch": 2.368995633187773,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044411897781439946,
      "loss": 5.1459,
      "mean_token_accuracy": 0.1684451475739479,
      "num_tokens": 52304172.0,
      "step": 23870
    },
    {
      "entropy": 5.425190496444702,
      "epoch": 2.3694918618499403,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004440956060348191,
      "loss": 5.1012,
      "mean_token_accuracy": 0.17213122248649598,
      "num_tokens": 52315370.0,
      "step": 23875
    },
    {
      "entropy": 5.50019645690918,
      "epoch": 2.369988090512108,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004440722300620321,
      "loss": 5.1542,
      "mean_token_accuracy": 0.16806354820728303,
      "num_tokens": 52326517.0,
      "step": 23880
    },
    {
      "entropy": 5.537506437301635,
      "epoch": 2.3704843191742757,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044404884989661804,
      "loss": 5.2347,
      "mean_token_accuracy": 0.16935124546289443,
      "num_tokens": 52337110.0,
      "step": 23885
    },
    {
      "entropy": 5.490321445465088,
      "epoch": 2.370980547836443,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004440254655391568,
      "loss": 5.1269,
      "mean_token_accuracy": 0.1731445610523224,
      "num_tokens": 52347509.0,
      "step": 23890
    },
    {
      "entropy": 5.610657596588135,
      "epoch": 2.3714767764986107,
      "grad_norm": 1.125,
      "learning_rate": 0.0004440020769902282,
      "loss": 5.2622,
      "mean_token_accuracy": 0.1642740249633789,
      "num_tokens": 52357920.0,
      "step": 23895
    },
    {
      "entropy": 5.521284198760986,
      "epoch": 2.371973005160778,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044397868425041234,
      "loss": 5.2275,
      "mean_token_accuracy": 0.16086383908987045,
      "num_tokens": 52368138.0,
      "step": 23900
    },
    {
      "entropy": 5.5524882793426515,
      "epoch": 2.3724692338229456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004439552873202891,
      "loss": 5.1888,
      "mean_token_accuracy": 0.16865319460630418,
      "num_tokens": 52380503.0,
      "step": 23905
    },
    {
      "entropy": 5.494733095169067,
      "epoch": 2.372965462485113,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044393188620043893,
      "loss": 5.1778,
      "mean_token_accuracy": 0.16667150110006332,
      "num_tokens": 52390850.0,
      "step": 23910
    },
    {
      "entropy": 5.430724000930786,
      "epoch": 2.3734616911472806,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004439084808914419,
      "loss": 5.1075,
      "mean_token_accuracy": 0.17282207310199738,
      "num_tokens": 52400829.0,
      "step": 23915
    },
    {
      "entropy": 5.551803302764893,
      "epoch": 2.373957919809448,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044388507139387864,
      "loss": 5.2246,
      "mean_token_accuracy": 0.16473632603883742,
      "num_tokens": 52412892.0,
      "step": 23920
    },
    {
      "entropy": 5.498248291015625,
      "epoch": 2.3744541484716155,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004438616577083293,
      "loss": 5.151,
      "mean_token_accuracy": 0.16887165009975433,
      "num_tokens": 52423662.0,
      "step": 23925
    },
    {
      "entropy": 5.576331758499146,
      "epoch": 2.3749503771337834,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000443838239835375,
      "loss": 5.177,
      "mean_token_accuracy": 0.1680815562605858,
      "num_tokens": 52433541.0,
      "step": 23930
    },
    {
      "entropy": 5.563626050949097,
      "epoch": 2.375446605795951,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004438148177755959,
      "loss": 5.237,
      "mean_token_accuracy": 0.16271129995584488,
      "num_tokens": 52445504.0,
      "step": 23935
    },
    {
      "entropy": 5.527077102661133,
      "epoch": 2.3759428344581184,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044379139152957327,
      "loss": 5.1986,
      "mean_token_accuracy": 0.16126647889614104,
      "num_tokens": 52456116.0,
      "step": 23940
    },
    {
      "entropy": 5.470192718505859,
      "epoch": 2.376439063120286,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044376796109788777,
      "loss": 5.067,
      "mean_token_accuracy": 0.17693418711423875,
      "num_tokens": 52465462.0,
      "step": 23945
    },
    {
      "entropy": 5.480688762664795,
      "epoch": 2.3769352917824533,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044374452648112046,
      "loss": 5.2381,
      "mean_token_accuracy": 0.16175909340381622,
      "num_tokens": 52477213.0,
      "step": 23950
    },
    {
      "entropy": 5.573980283737183,
      "epoch": 2.377431520444621,
      "grad_norm": 1.125,
      "learning_rate": 0.00044372108767985253,
      "loss": 5.2186,
      "mean_token_accuracy": 0.15870444923639299,
      "num_tokens": 52488095.0,
      "step": 23955
    },
    {
      "entropy": 5.580857944488526,
      "epoch": 2.3779277491067883,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004436976446946652,
      "loss": 5.2944,
      "mean_token_accuracy": 0.15948594361543655,
      "num_tokens": 52499078.0,
      "step": 23960
    },
    {
      "entropy": 5.514412975311279,
      "epoch": 2.3784239777689558,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044367419752613983,
      "loss": 5.1573,
      "mean_token_accuracy": 0.1655002638697624,
      "num_tokens": 52510829.0,
      "step": 23965
    },
    {
      "entropy": 5.561578273773193,
      "epoch": 2.3789202064311237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004436507461748577,
      "loss": 5.1775,
      "mean_token_accuracy": 0.16813046485185623,
      "num_tokens": 52521466.0,
      "step": 23970
    },
    {
      "entropy": 5.538183164596558,
      "epoch": 2.379416435093291,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004436272906414007,
      "loss": 5.0945,
      "mean_token_accuracy": 0.17464903891086578,
      "num_tokens": 52532617.0,
      "step": 23975
    },
    {
      "entropy": 5.386767292022705,
      "epoch": 2.3799126637554586,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004436038309263501,
      "loss": 5.1511,
      "mean_token_accuracy": 0.1625063106417656,
      "num_tokens": 52543826.0,
      "step": 23980
    },
    {
      "entropy": 5.459353971481323,
      "epoch": 2.380408892417626,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044358036703028773,
      "loss": 5.1989,
      "mean_token_accuracy": 0.16158126294612885,
      "num_tokens": 52556752.0,
      "step": 23985
    },
    {
      "entropy": 5.570642614364624,
      "epoch": 2.3809051210797936,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044355689895379563,
      "loss": 5.2508,
      "mean_token_accuracy": 0.1575981080532074,
      "num_tokens": 52567268.0,
      "step": 23990
    },
    {
      "entropy": 5.55045804977417,
      "epoch": 2.381401349741961,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044353342669745566,
      "loss": 5.2078,
      "mean_token_accuracy": 0.1649333879351616,
      "num_tokens": 52576894.0,
      "step": 23995
    },
    {
      "entropy": 5.4725549697875975,
      "epoch": 2.3818975784041285,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044350995026184976,
      "loss": 5.1717,
      "mean_token_accuracy": 0.16418764740228653,
      "num_tokens": 52587434.0,
      "step": 24000
    },
    {
      "epoch": 2.3818975784041285,
      "eval_entropy": 5.358027359742001,
      "eval_loss": 5.317162036895752,
      "eval_mean_token_accuracy": 0.16507602981775185,
      "eval_num_tokens": 52587434.0,
      "eval_runtime": 26.6141,
      "eval_samples_per_second": 1265.081,
      "eval_steps_per_second": 158.149,
      "step": 24000
    },
    {
      "entropy": 5.557210206985474,
      "epoch": 2.382393807066296,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044348646964756036,
      "loss": 5.2544,
      "mean_token_accuracy": 0.16501716375350953,
      "num_tokens": 52599455.0,
      "step": 24005
    },
    {
      "entropy": 5.54779019355774,
      "epoch": 2.3828900357284635,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004434629848551694,
      "loss": 5.2259,
      "mean_token_accuracy": 0.1592721700668335,
      "num_tokens": 52611020.0,
      "step": 24010
    },
    {
      "entropy": 5.491063499450684,
      "epoch": 2.3833862643906314,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004434394958852595,
      "loss": 5.1601,
      "mean_token_accuracy": 0.1677895411849022,
      "num_tokens": 52621743.0,
      "step": 24015
    },
    {
      "entropy": 5.542436838150024,
      "epoch": 2.383882493052799,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044341600273841304,
      "loss": 5.212,
      "mean_token_accuracy": 0.16672587245702744,
      "num_tokens": 52633437.0,
      "step": 24020
    },
    {
      "entropy": 5.44253454208374,
      "epoch": 2.3843787217149663,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044339250541521264,
      "loss": 5.112,
      "mean_token_accuracy": 0.17198125571012496,
      "num_tokens": 52643849.0,
      "step": 24025
    },
    {
      "entropy": 5.601398277282715,
      "epoch": 2.384874950377134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044336900391624096,
      "loss": 5.317,
      "mean_token_accuracy": 0.15744950622320175,
      "num_tokens": 52656417.0,
      "step": 24030
    },
    {
      "entropy": 5.488827848434449,
      "epoch": 2.3853711790393013,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004433454982420808,
      "loss": 5.0959,
      "mean_token_accuracy": 0.17074553221464156,
      "num_tokens": 52668491.0,
      "step": 24035
    },
    {
      "entropy": 5.502847909927368,
      "epoch": 2.3858674077014688,
      "grad_norm": 1.203125,
      "learning_rate": 0.000443321988393315,
      "loss": 5.2266,
      "mean_token_accuracy": 0.16128681600093842,
      "num_tokens": 52679532.0,
      "step": 24040
    },
    {
      "entropy": 5.486696720123291,
      "epoch": 2.3863636363636362,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004432984743705265,
      "loss": 5.1015,
      "mean_token_accuracy": 0.1663249909877777,
      "num_tokens": 52689850.0,
      "step": 24045
    },
    {
      "entropy": 5.430696630477906,
      "epoch": 2.3868598650258037,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044327495617429866,
      "loss": 5.0974,
      "mean_token_accuracy": 0.16932584941387177,
      "num_tokens": 52699529.0,
      "step": 24050
    },
    {
      "entropy": 5.406080102920532,
      "epoch": 2.3873560936879716,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004432514338052144,
      "loss": 5.1001,
      "mean_token_accuracy": 0.1726502537727356,
      "num_tokens": 52711449.0,
      "step": 24055
    },
    {
      "entropy": 5.532567691802979,
      "epoch": 2.387852322350139,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004432279072638571,
      "loss": 5.2266,
      "mean_token_accuracy": 0.1562381625175476,
      "num_tokens": 52721528.0,
      "step": 24060
    },
    {
      "entropy": 5.521614503860474,
      "epoch": 2.3883485510123066,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00044320437655081033,
      "loss": 5.1786,
      "mean_token_accuracy": 0.1643952786922455,
      "num_tokens": 52731367.0,
      "step": 24065
    },
    {
      "entropy": 5.463799524307251,
      "epoch": 2.388844779674474,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044318084166665737,
      "loss": 5.0543,
      "mean_token_accuracy": 0.17602844685316085,
      "num_tokens": 52740822.0,
      "step": 24070
    },
    {
      "entropy": 5.425349187850952,
      "epoch": 2.3893410083366415,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000443157302611982,
      "loss": 5.1458,
      "mean_token_accuracy": 0.16143161356449126,
      "num_tokens": 52751609.0,
      "step": 24075
    },
    {
      "entropy": 5.502972650527954,
      "epoch": 2.389837236998809,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004431337593873678,
      "loss": 5.2124,
      "mean_token_accuracy": 0.1611618384718895,
      "num_tokens": 52762914.0,
      "step": 24080
    },
    {
      "entropy": 5.600886821746826,
      "epoch": 2.3903334656609765,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004431102119933987,
      "loss": 5.2575,
      "mean_token_accuracy": 0.1636226400732994,
      "num_tokens": 52773894.0,
      "step": 24085
    },
    {
      "entropy": 5.469949579238891,
      "epoch": 2.390829694323144,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004430866604306586,
      "loss": 5.1839,
      "mean_token_accuracy": 0.1684652179479599,
      "num_tokens": 52784606.0,
      "step": 24090
    },
    {
      "entropy": 5.474429655075073,
      "epoch": 2.3913259229853114,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044306310469973147,
      "loss": 5.138,
      "mean_token_accuracy": 0.16385134309530258,
      "num_tokens": 52795275.0,
      "step": 24095
    },
    {
      "entropy": 5.56477632522583,
      "epoch": 2.391822151647479,
      "grad_norm": 2.03125,
      "learning_rate": 0.00044303954480120155,
      "loss": 5.2237,
      "mean_token_accuracy": 0.16556355655193328,
      "num_tokens": 52805058.0,
      "step": 24100
    },
    {
      "entropy": 5.538788795471191,
      "epoch": 2.392318380309647,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044301598073565295,
      "loss": 5.1501,
      "mean_token_accuracy": 0.1644430860877037,
      "num_tokens": 52815486.0,
      "step": 24105
    },
    {
      "entropy": 5.477749347686768,
      "epoch": 2.3928146089718143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004429924125036702,
      "loss": 5.1702,
      "mean_token_accuracy": 0.16726619005203247,
      "num_tokens": 52827039.0,
      "step": 24110
    },
    {
      "entropy": 5.492331314086914,
      "epoch": 2.393310837633982,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004429688401058375,
      "loss": 5.1715,
      "mean_token_accuracy": 0.16768986880779266,
      "num_tokens": 52837712.0,
      "step": 24115
    },
    {
      "entropy": 5.458554983139038,
      "epoch": 2.3938070662961493,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004429452635427394,
      "loss": 5.0481,
      "mean_token_accuracy": 0.17227339148521423,
      "num_tokens": 52846321.0,
      "step": 24120
    },
    {
      "entropy": 5.496169757843018,
      "epoch": 2.3943032949583167,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044292168281496086,
      "loss": 5.1175,
      "mean_token_accuracy": 0.16970191150903702,
      "num_tokens": 52856342.0,
      "step": 24125
    },
    {
      "entropy": 5.539505815505981,
      "epoch": 2.394799523620484,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004428980979230863,
      "loss": 5.2646,
      "mean_token_accuracy": 0.16487545818090438,
      "num_tokens": 52866602.0,
      "step": 24130
    },
    {
      "entropy": 5.459436416625977,
      "epoch": 2.3952957522826517,
      "grad_norm": 1.984375,
      "learning_rate": 0.00044287450886770075,
      "loss": 5.0859,
      "mean_token_accuracy": 0.16918046325445174,
      "num_tokens": 52877066.0,
      "step": 24135
    },
    {
      "entropy": 5.484340953826904,
      "epoch": 2.3957919809448196,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004428509156493891,
      "loss": 5.1395,
      "mean_token_accuracy": 0.1715347483754158,
      "num_tokens": 52887907.0,
      "step": 24140
    },
    {
      "entropy": 5.577832412719727,
      "epoch": 2.396288209606987,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004428273182687364,
      "loss": 5.2705,
      "mean_token_accuracy": 0.15924368500709535,
      "num_tokens": 52899308.0,
      "step": 24145
    },
    {
      "entropy": 5.459581470489502,
      "epoch": 2.3967844382691545,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004428037167263279,
      "loss": 5.0802,
      "mean_token_accuracy": 0.1740162342786789,
      "num_tokens": 52909127.0,
      "step": 24150
    },
    {
      "entropy": 5.442922782897949,
      "epoch": 2.397280666931322,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044278011102274887,
      "loss": 5.1026,
      "mean_token_accuracy": 0.17385817617177962,
      "num_tokens": 52920503.0,
      "step": 24155
    },
    {
      "entropy": 5.486176156997681,
      "epoch": 2.3977768955934895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044275650115858445,
      "loss": 5.1392,
      "mean_token_accuracy": 0.16671485900878907,
      "num_tokens": 52931904.0,
      "step": 24160
    },
    {
      "entropy": 5.531282091140747,
      "epoch": 2.398273124255657,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004427328871344204,
      "loss": 5.2865,
      "mean_token_accuracy": 0.15889964997768402,
      "num_tokens": 52945101.0,
      "step": 24165
    },
    {
      "entropy": 5.504766178131104,
      "epoch": 2.3987693529178244,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004427092689508422,
      "loss": 5.1952,
      "mean_token_accuracy": 0.16370865255594252,
      "num_tokens": 52955999.0,
      "step": 24170
    },
    {
      "entropy": 5.511669874191284,
      "epoch": 2.399265581579992,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004426856466084354,
      "loss": 5.1894,
      "mean_token_accuracy": 0.16192275285720825,
      "num_tokens": 52967273.0,
      "step": 24175
    },
    {
      "entropy": 5.547943449020385,
      "epoch": 2.3997618102421594,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000442662020107786,
      "loss": 5.2432,
      "mean_token_accuracy": 0.1661027655005455,
      "num_tokens": 52977415.0,
      "step": 24180
    },
    {
      "entropy": 5.460249185562134,
      "epoch": 2.400258038904327,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004426383894494797,
      "loss": 5.142,
      "mean_token_accuracy": 0.16300302296876906,
      "num_tokens": 52990071.0,
      "step": 24185
    },
    {
      "entropy": 5.435583639144897,
      "epoch": 2.400754267566495,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004426147546341027,
      "loss": 5.1167,
      "mean_token_accuracy": 0.16600433737039566,
      "num_tokens": 52999874.0,
      "step": 24190
    },
    {
      "entropy": 5.544477462768555,
      "epoch": 2.4012504962286623,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044259111566224084,
      "loss": 5.1418,
      "mean_token_accuracy": 0.16603493988513945,
      "num_tokens": 53011349.0,
      "step": 24195
    },
    {
      "entropy": 5.441782093048095,
      "epoch": 2.4017467248908297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044256747253448045,
      "loss": 5.1209,
      "mean_token_accuracy": 0.17614084333181382,
      "num_tokens": 53024256.0,
      "step": 24200
    },
    {
      "entropy": 5.4532874584197994,
      "epoch": 2.402242953552997,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044254382525140784,
      "loss": 5.1028,
      "mean_token_accuracy": 0.1795370563864708,
      "num_tokens": 53035617.0,
      "step": 24205
    },
    {
      "entropy": 5.459782266616822,
      "epoch": 2.4027391822151647,
      "grad_norm": 1.125,
      "learning_rate": 0.00044252017381360935,
      "loss": 5.1084,
      "mean_token_accuracy": 0.17614209651947021,
      "num_tokens": 53046219.0,
      "step": 24210
    },
    {
      "entropy": 5.547263669967651,
      "epoch": 2.403235410877332,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004424965182216716,
      "loss": 5.2577,
      "mean_token_accuracy": 0.15765416920185088,
      "num_tokens": 53057114.0,
      "step": 24215
    },
    {
      "entropy": 5.564920663833618,
      "epoch": 2.4037316395394996,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044247285847618105,
      "loss": 5.1671,
      "mean_token_accuracy": 0.16860369890928267,
      "num_tokens": 53067481.0,
      "step": 24220
    },
    {
      "entropy": 5.587635183334351,
      "epoch": 2.4042278682016676,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004424491945777245,
      "loss": 5.2754,
      "mean_token_accuracy": 0.16719166785478592,
      "num_tokens": 53078129.0,
      "step": 24225
    },
    {
      "entropy": 5.461627674102783,
      "epoch": 2.404724096863835,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044242552652688885,
      "loss": 5.1032,
      "mean_token_accuracy": 0.17059106677770614,
      "num_tokens": 53088919.0,
      "step": 24230
    },
    {
      "entropy": 5.461589622497558,
      "epoch": 2.4052203255260025,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004424018543242608,
      "loss": 5.1874,
      "mean_token_accuracy": 0.1715749219059944,
      "num_tokens": 53099716.0,
      "step": 24235
    },
    {
      "entropy": 5.445217752456665,
      "epoch": 2.40571655418817,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004423781779704275,
      "loss": 5.0783,
      "mean_token_accuracy": 0.17372402995824815,
      "num_tokens": 53108798.0,
      "step": 24240
    },
    {
      "entropy": 5.525392675399781,
      "epoch": 2.4062127828503375,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004423544974659761,
      "loss": 5.2087,
      "mean_token_accuracy": 0.16193539500236512,
      "num_tokens": 53120655.0,
      "step": 24245
    },
    {
      "entropy": 5.430534696578979,
      "epoch": 2.406709011512505,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044233081281149373,
      "loss": 5.0577,
      "mean_token_accuracy": 0.17135241776704788,
      "num_tokens": 53131760.0,
      "step": 24250
    },
    {
      "entropy": 5.489073038101196,
      "epoch": 2.4072052401746724,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044230712400756774,
      "loss": 5.1214,
      "mean_token_accuracy": 0.17090380787849427,
      "num_tokens": 53142347.0,
      "step": 24255
    },
    {
      "entropy": 5.4653363704681395,
      "epoch": 2.40770146883684,
      "grad_norm": 1.125,
      "learning_rate": 0.00044228343105478564,
      "loss": 5.1701,
      "mean_token_accuracy": 0.16465016901493074,
      "num_tokens": 53153678.0,
      "step": 24260
    },
    {
      "entropy": 5.474195718765259,
      "epoch": 2.4081976974990074,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004422597339537349,
      "loss": 5.2148,
      "mean_token_accuracy": 0.1659219667315483,
      "num_tokens": 53165089.0,
      "step": 24265
    },
    {
      "entropy": 5.5227361679077145,
      "epoch": 2.408693926161175,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004422360327050032,
      "loss": 5.1362,
      "mean_token_accuracy": 0.16279854923486708,
      "num_tokens": 53175614.0,
      "step": 24270
    },
    {
      "entropy": 5.52642970085144,
      "epoch": 2.4091901548233428,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044221232730917823,
      "loss": 5.1124,
      "mean_token_accuracy": 0.1673891916871071,
      "num_tokens": 53186306.0,
      "step": 24275
    },
    {
      "entropy": 5.498962163925171,
      "epoch": 2.4096863834855102,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044218861776684776,
      "loss": 5.183,
      "mean_token_accuracy": 0.1664038732647896,
      "num_tokens": 53197730.0,
      "step": 24280
    },
    {
      "entropy": 5.441062927246094,
      "epoch": 2.4101826121476777,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044216490407859995,
      "loss": 5.0979,
      "mean_token_accuracy": 0.17222774773836136,
      "num_tokens": 53208486.0,
      "step": 24285
    },
    {
      "entropy": 5.546136379241943,
      "epoch": 2.410678840809845,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00044214118624502267,
      "loss": 5.2028,
      "mean_token_accuracy": 0.16365479677915573,
      "num_tokens": 53220045.0,
      "step": 24290
    },
    {
      "entropy": 5.521191263198853,
      "epoch": 2.4111750694720127,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004421174642667042,
      "loss": 5.259,
      "mean_token_accuracy": 0.1582031100988388,
      "num_tokens": 53231385.0,
      "step": 24295
    },
    {
      "entropy": 5.559382963180542,
      "epoch": 2.41167129813418,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044209373814423265,
      "loss": 5.24,
      "mean_token_accuracy": 0.16141867637634277,
      "num_tokens": 53243038.0,
      "step": 24300
    },
    {
      "entropy": 5.5269218444824215,
      "epoch": 2.4121675267963476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004420700078781965,
      "loss": 5.1992,
      "mean_token_accuracy": 0.1642425388097763,
      "num_tokens": 53252959.0,
      "step": 24305
    },
    {
      "entropy": 5.397058486938477,
      "epoch": 2.4126637554585155,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044204627346918407,
      "loss": 5.1164,
      "mean_token_accuracy": 0.17574351131916047,
      "num_tokens": 53263542.0,
      "step": 24310
    },
    {
      "entropy": 5.489332675933838,
      "epoch": 2.413159984120683,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044202253491778404,
      "loss": 5.1256,
      "mean_token_accuracy": 0.17372291386127472,
      "num_tokens": 53273511.0,
      "step": 24315
    },
    {
      "entropy": 5.543849802017212,
      "epoch": 2.4136562127828505,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004419987922245852,
      "loss": 5.2595,
      "mean_token_accuracy": 0.163771453499794,
      "num_tokens": 53286173.0,
      "step": 24320
    },
    {
      "entropy": 5.4946287155151365,
      "epoch": 2.414152441445018,
      "grad_norm": 1.234375,
      "learning_rate": 0.000441975045390176,
      "loss": 5.1314,
      "mean_token_accuracy": 0.16695608645677568,
      "num_tokens": 53296784.0,
      "step": 24325
    },
    {
      "entropy": 5.523386764526367,
      "epoch": 2.4146486701071854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044195129441514547,
      "loss": 5.203,
      "mean_token_accuracy": 0.1700603649020195,
      "num_tokens": 53308044.0,
      "step": 24330
    },
    {
      "entropy": 5.442051267623901,
      "epoch": 2.415144898769353,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044192753930008267,
      "loss": 5.1071,
      "mean_token_accuracy": 0.17101516276597978,
      "num_tokens": 53319161.0,
      "step": 24335
    },
    {
      "entropy": 5.415552759170533,
      "epoch": 2.4156411274315204,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004419037800455765,
      "loss": 5.0808,
      "mean_token_accuracy": 0.17500595897436141,
      "num_tokens": 53329766.0,
      "step": 24340
    },
    {
      "entropy": 5.519049501419067,
      "epoch": 2.416137356093688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004418800166522163,
      "loss": 5.178,
      "mean_token_accuracy": 0.16706898659467698,
      "num_tokens": 53340329.0,
      "step": 24345
    },
    {
      "entropy": 5.552421903610229,
      "epoch": 2.4166335847558553,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044185624912059125,
      "loss": 5.2029,
      "mean_token_accuracy": 0.16667893528938293,
      "num_tokens": 53352592.0,
      "step": 24350
    },
    {
      "entropy": 5.562212371826172,
      "epoch": 2.417129813418023,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004418324774512908,
      "loss": 5.1672,
      "mean_token_accuracy": 0.16416443437337874,
      "num_tokens": 53362877.0,
      "step": 24355
    },
    {
      "entropy": 5.500103616714478,
      "epoch": 2.4176260420801907,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004418087016449043,
      "loss": 5.1751,
      "mean_token_accuracy": 0.16681524217128754,
      "num_tokens": 53374612.0,
      "step": 24360
    },
    {
      "entropy": 5.516773366928101,
      "epoch": 2.418122270742358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044178492170202153,
      "loss": 5.188,
      "mean_token_accuracy": 0.16329614520072938,
      "num_tokens": 53385129.0,
      "step": 24365
    },
    {
      "entropy": 5.505854701995849,
      "epoch": 2.4186184994045257,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044176113762323213,
      "loss": 5.25,
      "mean_token_accuracy": 0.1657223030924797,
      "num_tokens": 53395403.0,
      "step": 24370
    },
    {
      "entropy": 5.528942823410034,
      "epoch": 2.419114728066693,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044173734940912586,
      "loss": 5.221,
      "mean_token_accuracy": 0.15854802429676057,
      "num_tokens": 53406684.0,
      "step": 24375
    },
    {
      "entropy": 5.460414266586303,
      "epoch": 2.4196109567288606,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044171355706029247,
      "loss": 5.1031,
      "mean_token_accuracy": 0.16930858045816422,
      "num_tokens": 53417256.0,
      "step": 24380
    },
    {
      "entropy": 5.414714241027832,
      "epoch": 2.420107185391028,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044168976057732225,
      "loss": 5.0848,
      "mean_token_accuracy": 0.16225146651268005,
      "num_tokens": 53428305.0,
      "step": 24385
    },
    {
      "entropy": 5.426676273345947,
      "epoch": 2.4206034140531956,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004416659599608051,
      "loss": 5.1521,
      "mean_token_accuracy": 0.16316914111375808,
      "num_tokens": 53439301.0,
      "step": 24390
    },
    {
      "entropy": 5.543714618682861,
      "epoch": 2.4210996427153635,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004416421552113312,
      "loss": 5.2634,
      "mean_token_accuracy": 0.154072168469429,
      "num_tokens": 53450478.0,
      "step": 24395
    },
    {
      "entropy": 5.528044176101685,
      "epoch": 2.421595871377531,
      "grad_norm": 1.1875,
      "learning_rate": 0.000441618346329491,
      "loss": 5.1667,
      "mean_token_accuracy": 0.16749908030033112,
      "num_tokens": 53461293.0,
      "step": 24400
    },
    {
      "entropy": 5.465088844299316,
      "epoch": 2.4220921000396984,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044159453331587493,
      "loss": 5.1534,
      "mean_token_accuracy": 0.17203670293092727,
      "num_tokens": 53472014.0,
      "step": 24405
    },
    {
      "entropy": 5.505947303771973,
      "epoch": 2.422588328701866,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044157071617107325,
      "loss": 5.2161,
      "mean_token_accuracy": 0.16954043954610826,
      "num_tokens": 53483010.0,
      "step": 24410
    },
    {
      "entropy": 5.528653907775879,
      "epoch": 2.4230845573640334,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044154689489567684,
      "loss": 5.2059,
      "mean_token_accuracy": 0.16374699473381044,
      "num_tokens": 53494151.0,
      "step": 24415
    },
    {
      "entropy": 5.489218521118164,
      "epoch": 2.423580786026201,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044152306949027625,
      "loss": 5.1873,
      "mean_token_accuracy": 0.1639280542731285,
      "num_tokens": 53504462.0,
      "step": 24420
    },
    {
      "entropy": 5.480090093612671,
      "epoch": 2.4240770146883683,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044149923995546244,
      "loss": 5.1589,
      "mean_token_accuracy": 0.16686140596866608,
      "num_tokens": 53514896.0,
      "step": 24425
    },
    {
      "entropy": 5.493260478973388,
      "epoch": 2.424573243350536,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004414754062918262,
      "loss": 5.1171,
      "mean_token_accuracy": 0.1708350121974945,
      "num_tokens": 53524656.0,
      "step": 24430
    },
    {
      "entropy": 5.5432881832122805,
      "epoch": 2.4250694720127033,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004414515684999587,
      "loss": 5.1547,
      "mean_token_accuracy": 0.1726398929953575,
      "num_tokens": 53534896.0,
      "step": 24435
    },
    {
      "entropy": 5.420786762237549,
      "epoch": 2.4255657006748708,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004414277265804508,
      "loss": 5.1156,
      "mean_token_accuracy": 0.16907624900341034,
      "num_tokens": 53546108.0,
      "step": 24440
    },
    {
      "entropy": 5.432462978363037,
      "epoch": 2.4260619293370387,
      "grad_norm": 1.0625,
      "learning_rate": 0.000441403880533894,
      "loss": 5.1373,
      "mean_token_accuracy": 0.16295867711305617,
      "num_tokens": 53557325.0,
      "step": 24445
    },
    {
      "entropy": 5.40432825088501,
      "epoch": 2.426558157999206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044138003036087956,
      "loss": 5.0576,
      "mean_token_accuracy": 0.168168406188488,
      "num_tokens": 53568737.0,
      "step": 24450
    },
    {
      "entropy": 5.503941059112549,
      "epoch": 2.4270543866613736,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004413561760619988,
      "loss": 5.1814,
      "mean_token_accuracy": 0.16551800668239594,
      "num_tokens": 53579312.0,
      "step": 24455
    },
    {
      "entropy": 5.451930665969849,
      "epoch": 2.427550615323541,
      "grad_norm": 1.125,
      "learning_rate": 0.0004413323176378434,
      "loss": 5.0794,
      "mean_token_accuracy": 0.17426641285419464,
      "num_tokens": 53590476.0,
      "step": 24460
    },
    {
      "entropy": 5.468898820877075,
      "epoch": 2.4280468439857086,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004413084550890049,
      "loss": 5.2052,
      "mean_token_accuracy": 0.16292957216501236,
      "num_tokens": 53601284.0,
      "step": 24465
    },
    {
      "entropy": 5.466270637512207,
      "epoch": 2.428543072647876,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004412845884160751,
      "loss": 5.1805,
      "mean_token_accuracy": 0.16526614129543304,
      "num_tokens": 53612590.0,
      "step": 24470
    },
    {
      "entropy": 5.50200514793396,
      "epoch": 2.4290393013100435,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004412607176196458,
      "loss": 5.1904,
      "mean_token_accuracy": 0.16213615238666534,
      "num_tokens": 53623049.0,
      "step": 24475
    },
    {
      "entropy": 5.493684101104736,
      "epoch": 2.429535529972211,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044123684270030907,
      "loss": 5.1351,
      "mean_token_accuracy": 0.1697455197572708,
      "num_tokens": 53633961.0,
      "step": 24480
    },
    {
      "entropy": 5.459229850769043,
      "epoch": 2.430031758634379,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044121296365865674,
      "loss": 5.1913,
      "mean_token_accuracy": 0.16524432748556137,
      "num_tokens": 53644634.0,
      "step": 24485
    },
    {
      "entropy": 5.464613533020019,
      "epoch": 2.4305279872965464,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044118908049528105,
      "loss": 5.1243,
      "mean_token_accuracy": 0.16983065605163575,
      "num_tokens": 53655982.0,
      "step": 24490
    },
    {
      "entropy": 5.575466585159302,
      "epoch": 2.431024215958714,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044116519321077435,
      "loss": 5.2089,
      "mean_token_accuracy": 0.16255413144826888,
      "num_tokens": 53667367.0,
      "step": 24495
    },
    {
      "entropy": 5.474284219741821,
      "epoch": 2.4315204446208813,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044114130180572883,
      "loss": 5.1003,
      "mean_token_accuracy": 0.16943545937538146,
      "num_tokens": 53677191.0,
      "step": 24500
    },
    {
      "entropy": 5.461921977996826,
      "epoch": 2.432016673283049,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004411174062807371,
      "loss": 5.1842,
      "mean_token_accuracy": 0.1662721648812294,
      "num_tokens": 53688636.0,
      "step": 24505
    },
    {
      "entropy": 5.509063816070556,
      "epoch": 2.4325129019452163,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004410935066363916,
      "loss": 5.1356,
      "mean_token_accuracy": 0.16394818127155303,
      "num_tokens": 53699588.0,
      "step": 24510
    },
    {
      "entropy": 5.512472105026245,
      "epoch": 2.4330091306073838,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044106960287328503,
      "loss": 5.2373,
      "mean_token_accuracy": 0.15902873128652573,
      "num_tokens": 53710427.0,
      "step": 24515
    },
    {
      "entropy": 5.531307935714722,
      "epoch": 2.4335053592695512,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004410456949920102,
      "loss": 5.2188,
      "mean_token_accuracy": 0.16043241322040558,
      "num_tokens": 53722319.0,
      "step": 24520
    },
    {
      "entropy": 5.518188190460205,
      "epoch": 2.4340015879317187,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044102178299315994,
      "loss": 5.2202,
      "mean_token_accuracy": 0.16112845093011857,
      "num_tokens": 53734059.0,
      "step": 24525
    },
    {
      "entropy": 5.506659078598022,
      "epoch": 2.4344978165938866,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044099786687732713,
      "loss": 5.188,
      "mean_token_accuracy": 0.1640896812081337,
      "num_tokens": 53745795.0,
      "step": 24530
    },
    {
      "entropy": 5.502665233612061,
      "epoch": 2.434994045256054,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004409739466451049,
      "loss": 5.1909,
      "mean_token_accuracy": 0.1639878585934639,
      "num_tokens": 53756781.0,
      "step": 24535
    },
    {
      "entropy": 5.586558485031128,
      "epoch": 2.4354902739182216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044095002229708655,
      "loss": 5.2748,
      "mean_token_accuracy": 0.16091121584177018,
      "num_tokens": 53769725.0,
      "step": 24540
    },
    {
      "entropy": 5.47840576171875,
      "epoch": 2.435986502580389,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044092609383386517,
      "loss": 5.142,
      "mean_token_accuracy": 0.17084171772003173,
      "num_tokens": 53781217.0,
      "step": 24545
    },
    {
      "entropy": 5.533466625213623,
      "epoch": 2.4364827312425565,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044090216125603426,
      "loss": 5.215,
      "mean_token_accuracy": 0.1717365264892578,
      "num_tokens": 53791478.0,
      "step": 24550
    },
    {
      "entropy": 5.4765993595123295,
      "epoch": 2.436978959904724,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004408782245641872,
      "loss": 5.1246,
      "mean_token_accuracy": 0.17112470269203187,
      "num_tokens": 53802844.0,
      "step": 24555
    },
    {
      "entropy": 5.5017781257629395,
      "epoch": 2.4374751885668915,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004408542837589176,
      "loss": 5.1847,
      "mean_token_accuracy": 0.16816852390766143,
      "num_tokens": 53812938.0,
      "step": 24560
    },
    {
      "entropy": 5.523629665374756,
      "epoch": 2.437971417229059,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044083033884081914,
      "loss": 5.1333,
      "mean_token_accuracy": 0.16593381762504578,
      "num_tokens": 53824105.0,
      "step": 24565
    },
    {
      "entropy": 5.546679067611694,
      "epoch": 2.438467645891227,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004408063898104856,
      "loss": 5.2783,
      "mean_token_accuracy": 0.1611944243311882,
      "num_tokens": 53834956.0,
      "step": 24570
    },
    {
      "entropy": 5.488417530059815,
      "epoch": 2.4389638745533944,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044078243666851094,
      "loss": 5.1739,
      "mean_token_accuracy": 0.16576400846242906,
      "num_tokens": 53846257.0,
      "step": 24575
    },
    {
      "entropy": 5.493237495422363,
      "epoch": 2.439460103215562,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044075847941548905,
      "loss": 5.0951,
      "mean_token_accuracy": 0.1689578264951706,
      "num_tokens": 53857107.0,
      "step": 24580
    },
    {
      "entropy": 5.528479385375976,
      "epoch": 2.4399563318777293,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044073451805201396,
      "loss": 5.1538,
      "mean_token_accuracy": 0.16554897427558898,
      "num_tokens": 53867779.0,
      "step": 24585
    },
    {
      "entropy": 5.518164157867432,
      "epoch": 2.440452560539897,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044071055257868,
      "loss": 5.1509,
      "mean_token_accuracy": 0.1671651378273964,
      "num_tokens": 53878661.0,
      "step": 24590
    },
    {
      "entropy": 5.5774003028869625,
      "epoch": 2.4409487892020643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044068658299608145,
      "loss": 5.246,
      "mean_token_accuracy": 0.16056869328022003,
      "num_tokens": 53891300.0,
      "step": 24595
    },
    {
      "entropy": 5.529367971420288,
      "epoch": 2.4414450178642317,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044066260930481264,
      "loss": 5.1075,
      "mean_token_accuracy": 0.16488881260156632,
      "num_tokens": 53901086.0,
      "step": 24600
    },
    {
      "entropy": 5.4881847381591795,
      "epoch": 2.441941246526399,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004406386315054681,
      "loss": 5.1942,
      "mean_token_accuracy": 0.15644340515136718,
      "num_tokens": 53912599.0,
      "step": 24605
    },
    {
      "entropy": 5.396753978729248,
      "epoch": 2.4424374751885667,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044061464959864243,
      "loss": 5.1727,
      "mean_token_accuracy": 0.16605943888425828,
      "num_tokens": 53924293.0,
      "step": 24610
    },
    {
      "entropy": 5.5342436790466305,
      "epoch": 2.4429337038507346,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004405906635849302,
      "loss": 5.2311,
      "mean_token_accuracy": 0.1612158566713333,
      "num_tokens": 53935074.0,
      "step": 24615
    },
    {
      "entropy": 5.538517761230469,
      "epoch": 2.443429932512902,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004405666734649265,
      "loss": 5.1713,
      "mean_token_accuracy": 0.1640710562467575,
      "num_tokens": 53946770.0,
      "step": 24620
    },
    {
      "entropy": 5.495964193344117,
      "epoch": 2.4439261611750696,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044054267923922594,
      "loss": 5.1558,
      "mean_token_accuracy": 0.16345238238573073,
      "num_tokens": 53957077.0,
      "step": 24625
    },
    {
      "entropy": 5.479445075988769,
      "epoch": 2.444422389837237,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044051868090842376,
      "loss": 5.166,
      "mean_token_accuracy": 0.17293217927217483,
      "num_tokens": 53968455.0,
      "step": 24630
    },
    {
      "entropy": 5.41504487991333,
      "epoch": 2.4449186184994045,
      "grad_norm": 1.125,
      "learning_rate": 0.00044049467847311486,
      "loss": 5.1252,
      "mean_token_accuracy": 0.1676115483045578,
      "num_tokens": 53980429.0,
      "step": 24635
    },
    {
      "entropy": 5.445856618881225,
      "epoch": 2.445414847161572,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004404706719338946,
      "loss": 5.1218,
      "mean_token_accuracy": 0.17168766111135483,
      "num_tokens": 53990937.0,
      "step": 24640
    },
    {
      "entropy": 5.509689140319824,
      "epoch": 2.4459110758237395,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044044666129135825,
      "loss": 5.1504,
      "mean_token_accuracy": 0.17052045464515686,
      "num_tokens": 54003526.0,
      "step": 24645
    },
    {
      "entropy": 5.482084560394287,
      "epoch": 2.446407304485907,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044042264654610115,
      "loss": 5.1061,
      "mean_token_accuracy": 0.17539103776216508,
      "num_tokens": 54013875.0,
      "step": 24650
    },
    {
      "entropy": 5.513458299636841,
      "epoch": 2.446903533148075,
      "grad_norm": 1.328125,
      "learning_rate": 0.00044039862769871884,
      "loss": 5.2784,
      "mean_token_accuracy": 0.1586374193429947,
      "num_tokens": 54024103.0,
      "step": 24655
    },
    {
      "entropy": 5.496229600906372,
      "epoch": 2.4473997618102423,
      "grad_norm": 1.0,
      "learning_rate": 0.00044037460474980703,
      "loss": 5.2342,
      "mean_token_accuracy": 0.16266815662384032,
      "num_tokens": 54035417.0,
      "step": 24660
    },
    {
      "entropy": 5.598473978042603,
      "epoch": 2.44789599047241,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004403505776999614,
      "loss": 5.2176,
      "mean_token_accuracy": 0.1628843516111374,
      "num_tokens": 54045180.0,
      "step": 24665
    },
    {
      "entropy": 5.554920482635498,
      "epoch": 2.4483922191345773,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044032654654977766,
      "loss": 5.1475,
      "mean_token_accuracy": 0.17041084319353103,
      "num_tokens": 54056493.0,
      "step": 24670
    },
    {
      "entropy": 5.35691876411438,
      "epoch": 2.4488884477967447,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004403025112998519,
      "loss": 5.1355,
      "mean_token_accuracy": 0.16864904314279555,
      "num_tokens": 54067445.0,
      "step": 24675
    },
    {
      "entropy": 5.477802753448486,
      "epoch": 2.4493846764589122,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004402784719507799,
      "loss": 5.1393,
      "mean_token_accuracy": 0.16945022195577622,
      "num_tokens": 54076937.0,
      "step": 24680
    },
    {
      "entropy": 5.604517889022827,
      "epoch": 2.4498809051210797,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004402544285031581,
      "loss": 5.1952,
      "mean_token_accuracy": 0.16965131163597108,
      "num_tokens": 54088014.0,
      "step": 24685
    },
    {
      "entropy": 5.526897430419922,
      "epoch": 2.450377133783247,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004402303809575825,
      "loss": 5.1385,
      "mean_token_accuracy": 0.1672704890370369,
      "num_tokens": 54098984.0,
      "step": 24690
    },
    {
      "entropy": 5.584397268295288,
      "epoch": 2.4508733624454146,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004402063293146495,
      "loss": 5.285,
      "mean_token_accuracy": 0.1610693007707596,
      "num_tokens": 54110490.0,
      "step": 24695
    },
    {
      "entropy": 5.505985593795776,
      "epoch": 2.451369591107582,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004401822735749555,
      "loss": 5.1737,
      "mean_token_accuracy": 0.16464611738920212,
      "num_tokens": 54120785.0,
      "step": 24700
    },
    {
      "entropy": 5.45318751335144,
      "epoch": 2.45186581976975,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00044015821373909707,
      "loss": 5.1056,
      "mean_token_accuracy": 0.1713878333568573,
      "num_tokens": 54131339.0,
      "step": 24705
    },
    {
      "entropy": 5.491992664337158,
      "epoch": 2.4523620484319175,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044013414980767084,
      "loss": 5.2049,
      "mean_token_accuracy": 0.16227352023124694,
      "num_tokens": 54142270.0,
      "step": 24710
    },
    {
      "entropy": 5.496163654327392,
      "epoch": 2.452858277094085,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004401100817812735,
      "loss": 5.1587,
      "mean_token_accuracy": 0.16668582260608672,
      "num_tokens": 54153045.0,
      "step": 24715
    },
    {
      "entropy": 5.452143573760987,
      "epoch": 2.4533545057562525,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044008600966050196,
      "loss": 5.1192,
      "mean_token_accuracy": 0.16840971112251282,
      "num_tokens": 54164069.0,
      "step": 24720
    },
    {
      "entropy": 5.527698516845703,
      "epoch": 2.45385073441842,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004400619334459531,
      "loss": 5.1935,
      "mean_token_accuracy": 0.16327081620693207,
      "num_tokens": 54176902.0,
      "step": 24725
    },
    {
      "entropy": 5.514942598342896,
      "epoch": 2.4543469630805874,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044003785313822386,
      "loss": 5.1595,
      "mean_token_accuracy": 0.17139191776514054,
      "num_tokens": 54188515.0,
      "step": 24730
    },
    {
      "entropy": 5.547676372528076,
      "epoch": 2.454843191742755,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044001376873791166,
      "loss": 5.1922,
      "mean_token_accuracy": 0.16792000383138656,
      "num_tokens": 54199077.0,
      "step": 24735
    },
    {
      "entropy": 5.56328592300415,
      "epoch": 2.455339420404923,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043998968024561337,
      "loss": 5.2424,
      "mean_token_accuracy": 0.16490699648857116,
      "num_tokens": 54210216.0,
      "step": 24740
    },
    {
      "entropy": 5.482396268844605,
      "epoch": 2.4558356490670903,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004399655876619267,
      "loss": 5.1551,
      "mean_token_accuracy": 0.1708293303847313,
      "num_tokens": 54221143.0,
      "step": 24745
    },
    {
      "entropy": 5.5501587867736815,
      "epoch": 2.4563318777292578,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043994149098744884,
      "loss": 5.2433,
      "mean_token_accuracy": 0.16570744663476944,
      "num_tokens": 54233070.0,
      "step": 24750
    },
    {
      "entropy": 5.507483625411988,
      "epoch": 2.4568281063914252,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004399173902227775,
      "loss": 5.1295,
      "mean_token_accuracy": 0.17055393904447555,
      "num_tokens": 54243793.0,
      "step": 24755
    },
    {
      "entropy": 5.442591857910156,
      "epoch": 2.4573243350535927,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043989328536851014,
      "loss": 5.1565,
      "mean_token_accuracy": 0.16612577587366104,
      "num_tokens": 54256983.0,
      "step": 24760
    },
    {
      "entropy": 5.401678419113159,
      "epoch": 2.45782056371576,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00043986917642524465,
      "loss": 5.1252,
      "mean_token_accuracy": 0.1720290005207062,
      "num_tokens": 54268789.0,
      "step": 24765
    },
    {
      "entropy": 5.579776573181152,
      "epoch": 2.4583167923779277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043984506339357887,
      "loss": 5.1709,
      "mean_token_accuracy": 0.16174198240041732,
      "num_tokens": 54279517.0,
      "step": 24770
    },
    {
      "entropy": 5.5332982540130615,
      "epoch": 2.458813021040095,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004398209462741107,
      "loss": 5.1595,
      "mean_token_accuracy": 0.16578841656446458,
      "num_tokens": 54290865.0,
      "step": 24775
    },
    {
      "entropy": 5.577926301956177,
      "epoch": 2.4593092497022626,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004397968250674383,
      "loss": 5.246,
      "mean_token_accuracy": 0.16433537751436234,
      "num_tokens": 54302658.0,
      "step": 24780
    },
    {
      "entropy": 5.500668239593506,
      "epoch": 2.45980547836443,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004397726997741597,
      "loss": 5.1993,
      "mean_token_accuracy": 0.16879079043865203,
      "num_tokens": 54314000.0,
      "step": 24785
    },
    {
      "entropy": 5.448645162582397,
      "epoch": 2.460301707026598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004397485703948732,
      "loss": 5.1639,
      "mean_token_accuracy": 0.16857107132673263,
      "num_tokens": 54325952.0,
      "step": 24790
    },
    {
      "entropy": 5.554739046096802,
      "epoch": 2.4607979356887655,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004397244369301771,
      "loss": 5.2446,
      "mean_token_accuracy": 0.15953220501542092,
      "num_tokens": 54338582.0,
      "step": 24795
    },
    {
      "entropy": 5.524844789505005,
      "epoch": 2.461294164350933,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043970029938066995,
      "loss": 5.2179,
      "mean_token_accuracy": 0.1613105446100235,
      "num_tokens": 54349513.0,
      "step": 24800
    },
    {
      "entropy": 5.522817897796631,
      "epoch": 2.4617903930131004,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004396761577469503,
      "loss": 5.1454,
      "mean_token_accuracy": 0.165677048265934,
      "num_tokens": 54360264.0,
      "step": 24805
    },
    {
      "entropy": 5.522361803054809,
      "epoch": 2.462286621675268,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043965201202961673,
      "loss": 5.2007,
      "mean_token_accuracy": 0.1603436842560768,
      "num_tokens": 54371379.0,
      "step": 24810
    },
    {
      "entropy": 5.506715393066406,
      "epoch": 2.4627828503374354,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043962786222926807,
      "loss": 5.1929,
      "mean_token_accuracy": 0.1676512196660042,
      "num_tokens": 54382844.0,
      "step": 24815
    },
    {
      "entropy": 5.484408187866211,
      "epoch": 2.463279078999603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043960370834650325,
      "loss": 5.1336,
      "mean_token_accuracy": 0.17228222042322158,
      "num_tokens": 54392807.0,
      "step": 24820
    },
    {
      "entropy": 5.457158994674683,
      "epoch": 2.4637753076617708,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004395795503819211,
      "loss": 5.1765,
      "mean_token_accuracy": 0.16779960244894027,
      "num_tokens": 54404060.0,
      "step": 24825
    },
    {
      "entropy": 5.483636522293091,
      "epoch": 2.4642715363239382,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004395553883361207,
      "loss": 5.0729,
      "mean_token_accuracy": 0.17015119791030883,
      "num_tokens": 54414636.0,
      "step": 24830
    },
    {
      "entropy": 5.453122663497925,
      "epoch": 2.4647677649861057,
      "grad_norm": 1.203125,
      "learning_rate": 0.00043953122220970127,
      "loss": 5.1334,
      "mean_token_accuracy": 0.17414361387491226,
      "num_tokens": 54426185.0,
      "step": 24835
    },
    {
      "entropy": 5.47751817703247,
      "epoch": 2.465263993648273,
      "grad_norm": 1.125,
      "learning_rate": 0.00043950705200326205,
      "loss": 5.1268,
      "mean_token_accuracy": 0.1645190715789795,
      "num_tokens": 54436468.0,
      "step": 24840
    },
    {
      "entropy": 5.520951128005981,
      "epoch": 2.4657602223104407,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004394828777174025,
      "loss": 5.1459,
      "mean_token_accuracy": 0.17281889617443086,
      "num_tokens": 54446228.0,
      "step": 24845
    },
    {
      "entropy": 5.452070331573486,
      "epoch": 2.466256450972608,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043945869935272194,
      "loss": 5.207,
      "mean_token_accuracy": 0.16236743479967117,
      "num_tokens": 54456704.0,
      "step": 24850
    },
    {
      "entropy": 5.511898994445801,
      "epoch": 2.4667526796347756,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043943451690982006,
      "loss": 5.1653,
      "mean_token_accuracy": 0.1704462304711342,
      "num_tokens": 54467780.0,
      "step": 24855
    },
    {
      "entropy": 5.614805793762207,
      "epoch": 2.467248908296943,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004394103303892964,
      "loss": 5.2791,
      "mean_token_accuracy": 0.16117028295993804,
      "num_tokens": 54478428.0,
      "step": 24860
    },
    {
      "entropy": 5.440987396240234,
      "epoch": 2.4677451369591106,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043938613979175087,
      "loss": 5.0496,
      "mean_token_accuracy": 0.17179275006055833,
      "num_tokens": 54489544.0,
      "step": 24865
    },
    {
      "entropy": 5.471953916549682,
      "epoch": 2.468241365621278,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004393619451177834,
      "loss": 5.1433,
      "mean_token_accuracy": 0.1736326366662979,
      "num_tokens": 54501438.0,
      "step": 24870
    },
    {
      "entropy": 5.4409678936004635,
      "epoch": 2.468737594283446,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043933774636799365,
      "loss": 5.0845,
      "mean_token_accuracy": 0.17101412415504455,
      "num_tokens": 54511746.0,
      "step": 24875
    },
    {
      "entropy": 5.50528736114502,
      "epoch": 2.4692338229456134,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043931354354298204,
      "loss": 5.1995,
      "mean_token_accuracy": 0.1656542420387268,
      "num_tokens": 54523141.0,
      "step": 24880
    },
    {
      "entropy": 5.643340587615967,
      "epoch": 2.469730051607781,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004392893366433486,
      "loss": 5.2931,
      "mean_token_accuracy": 0.16204717606306077,
      "num_tokens": 54533275.0,
      "step": 24885
    },
    {
      "entropy": 5.567493438720703,
      "epoch": 2.4702262802699484,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043926512566969366,
      "loss": 5.1467,
      "mean_token_accuracy": 0.16648277193307875,
      "num_tokens": 54545309.0,
      "step": 24890
    },
    {
      "entropy": 5.446080303192138,
      "epoch": 2.470722508932116,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043924091062261743,
      "loss": 5.1229,
      "mean_token_accuracy": 0.17546064853668214,
      "num_tokens": 54554905.0,
      "step": 24895
    },
    {
      "entropy": 5.473501920700073,
      "epoch": 2.4712187375942833,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004392166915027206,
      "loss": 5.1665,
      "mean_token_accuracy": 0.1626105025410652,
      "num_tokens": 54566050.0,
      "step": 24900
    },
    {
      "entropy": 5.555691146850586,
      "epoch": 2.471714966256451,
      "grad_norm": 1.125,
      "learning_rate": 0.0004391924683106037,
      "loss": 5.2128,
      "mean_token_accuracy": 0.1634070709347725,
      "num_tokens": 54577791.0,
      "step": 24905
    },
    {
      "entropy": 5.544863653182984,
      "epoch": 2.4722111949186187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004391682410468674,
      "loss": 5.1669,
      "mean_token_accuracy": 0.16895221024751664,
      "num_tokens": 54589183.0,
      "step": 24910
    },
    {
      "entropy": 5.449429273605347,
      "epoch": 2.472707423580786,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043914400971211255,
      "loss": 5.1659,
      "mean_token_accuracy": 0.16386642158031464,
      "num_tokens": 54600776.0,
      "step": 24915
    },
    {
      "entropy": 5.500791311264038,
      "epoch": 2.4732036522429537,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004391197743069398,
      "loss": 5.1056,
      "mean_token_accuracy": 0.17613673061132432,
      "num_tokens": 54611167.0,
      "step": 24920
    },
    {
      "entropy": 5.53645429611206,
      "epoch": 2.473699880905121,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004390955348319503,
      "loss": 5.1555,
      "mean_token_accuracy": 0.17050084471702576,
      "num_tokens": 54622811.0,
      "step": 24925
    },
    {
      "entropy": 5.50278377532959,
      "epoch": 2.4741961095672886,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043907129128774525,
      "loss": 5.1756,
      "mean_token_accuracy": 0.16641454249620438,
      "num_tokens": 54635109.0,
      "step": 24930
    },
    {
      "entropy": 5.55668158531189,
      "epoch": 2.474692338229456,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043904704367492563,
      "loss": 5.2054,
      "mean_token_accuracy": 0.1676779642701149,
      "num_tokens": 54644260.0,
      "step": 24935
    },
    {
      "entropy": 5.485967111587525,
      "epoch": 2.4751885668916236,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004390227919940928,
      "loss": 5.0779,
      "mean_token_accuracy": 0.16932486444711686,
      "num_tokens": 54654433.0,
      "step": 24940
    },
    {
      "entropy": 5.502370882034302,
      "epoch": 2.475684795553791,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043899853624584825,
      "loss": 5.2212,
      "mean_token_accuracy": 0.16272015422582625,
      "num_tokens": 54666582.0,
      "step": 24945
    },
    {
      "entropy": 5.478611612319947,
      "epoch": 2.4761810242159585,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043897427643079336,
      "loss": 5.1452,
      "mean_token_accuracy": 0.16869761943817138,
      "num_tokens": 54677247.0,
      "step": 24950
    },
    {
      "entropy": 5.514150667190552,
      "epoch": 2.476677252878126,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043895001254952963,
      "loss": 5.1792,
      "mean_token_accuracy": 0.16699732840061188,
      "num_tokens": 54688423.0,
      "step": 24955
    },
    {
      "entropy": 5.488408613204956,
      "epoch": 2.477173481540294,
      "grad_norm": 1.09375,
      "learning_rate": 0.000438925744602659,
      "loss": 5.1896,
      "mean_token_accuracy": 0.16956758946180345,
      "num_tokens": 54698517.0,
      "step": 24960
    },
    {
      "entropy": 5.549522066116333,
      "epoch": 2.4776697102024614,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043890147259078314,
      "loss": 5.2259,
      "mean_token_accuracy": 0.16108118891716003,
      "num_tokens": 54711635.0,
      "step": 24965
    },
    {
      "entropy": 5.524739074707031,
      "epoch": 2.478165938864629,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043887719651450384,
      "loss": 5.1748,
      "mean_token_accuracy": 0.17237301766872407,
      "num_tokens": 54723051.0,
      "step": 24970
    },
    {
      "entropy": 5.5924982070922855,
      "epoch": 2.4786621675267964,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004388529163744233,
      "loss": 5.1715,
      "mean_token_accuracy": 0.1659539222717285,
      "num_tokens": 54734191.0,
      "step": 24975
    },
    {
      "entropy": 5.496977710723877,
      "epoch": 2.479158396188964,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004388286321711435,
      "loss": 5.1169,
      "mean_token_accuracy": 0.17067847698926925,
      "num_tokens": 54745632.0,
      "step": 24980
    },
    {
      "entropy": 5.482267951965332,
      "epoch": 2.4796546248511313,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004388043439052665,
      "loss": 5.1836,
      "mean_token_accuracy": 0.15986712276935577,
      "num_tokens": 54755871.0,
      "step": 24985
    },
    {
      "entropy": 5.450384998321534,
      "epoch": 2.480150853513299,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043878005157739486,
      "loss": 5.1073,
      "mean_token_accuracy": 0.17065593004226684,
      "num_tokens": 54768036.0,
      "step": 24990
    },
    {
      "entropy": 5.501453876495361,
      "epoch": 2.4806470821754667,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004387557551881308,
      "loss": 5.1369,
      "mean_token_accuracy": 0.1677062526345253,
      "num_tokens": 54778518.0,
      "step": 24995
    },
    {
      "entropy": 5.472608041763306,
      "epoch": 2.481143310837634,
      "grad_norm": 1.140625,
      "learning_rate": 0.000438731454738077,
      "loss": 5.119,
      "mean_token_accuracy": 0.17052611857652664,
      "num_tokens": 54790786.0,
      "step": 25000
    },
    {
      "entropy": 5.442429780960083,
      "epoch": 2.4816395394998016,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043870715022783575,
      "loss": 5.1163,
      "mean_token_accuracy": 0.16827951818704606,
      "num_tokens": 54800063.0,
      "step": 25005
    },
    {
      "entropy": 5.5290265560150145,
      "epoch": 2.482135768161969,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004386828416580101,
      "loss": 5.1987,
      "mean_token_accuracy": 0.15889407843351364,
      "num_tokens": 54810605.0,
      "step": 25010
    },
    {
      "entropy": 5.530377960205078,
      "epoch": 2.4826319968241366,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004386585290292026,
      "loss": 5.1072,
      "mean_token_accuracy": 0.1703396275639534,
      "num_tokens": 54820626.0,
      "step": 25015
    },
    {
      "entropy": 5.444969034194946,
      "epoch": 2.483128225486304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004386342123420161,
      "loss": 5.0952,
      "mean_token_accuracy": 0.1714339464902878,
      "num_tokens": 54831395.0,
      "step": 25020
    },
    {
      "entropy": 5.427635335922242,
      "epoch": 2.4836244541484715,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043860989159705396,
      "loss": 5.1038,
      "mean_token_accuracy": 0.17104551047086716,
      "num_tokens": 54841238.0,
      "step": 25025
    },
    {
      "entropy": 5.525115823745727,
      "epoch": 2.484120682810639,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043858556679491896,
      "loss": 5.2129,
      "mean_token_accuracy": 0.16488487273454666,
      "num_tokens": 54851301.0,
      "step": 25030
    },
    {
      "entropy": 5.523067951202393,
      "epoch": 2.4846169114728065,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043856123793621437,
      "loss": 5.1317,
      "mean_token_accuracy": 0.16791995167732238,
      "num_tokens": 54863135.0,
      "step": 25035
    },
    {
      "entropy": 5.509669399261474,
      "epoch": 2.485113140134974,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043853690502154355,
      "loss": 5.1977,
      "mean_token_accuracy": 0.16255323439836503,
      "num_tokens": 54874780.0,
      "step": 25040
    },
    {
      "entropy": 5.5517970561981205,
      "epoch": 2.485609368797142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004385125680515099,
      "loss": 5.1845,
      "mean_token_accuracy": 0.17167921513319015,
      "num_tokens": 54885237.0,
      "step": 25045
    },
    {
      "entropy": 5.518408966064453,
      "epoch": 2.4861055974593094,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004384882270267168,
      "loss": 5.1854,
      "mean_token_accuracy": 0.16769324243068695,
      "num_tokens": 54896801.0,
      "step": 25050
    },
    {
      "entropy": 5.475362730026245,
      "epoch": 2.486601826121477,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004384638819477681,
      "loss": 5.1418,
      "mean_token_accuracy": 0.16666772961616516,
      "num_tokens": 54908427.0,
      "step": 25055
    },
    {
      "entropy": 5.386861515045166,
      "epoch": 2.4870980547836443,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00043843953281526727,
      "loss": 5.0932,
      "mean_token_accuracy": 0.1721251904964447,
      "num_tokens": 54918394.0,
      "step": 25060
    },
    {
      "entropy": 5.5137471675872805,
      "epoch": 2.487594283445812,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043841517962981823,
      "loss": 5.1307,
      "mean_token_accuracy": 0.17007603049278258,
      "num_tokens": 54928613.0,
      "step": 25065
    },
    {
      "entropy": 5.527058410644531,
      "epoch": 2.4880905121079793,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043839082239202496,
      "loss": 5.0776,
      "mean_token_accuracy": 0.17119349241256715,
      "num_tokens": 54938938.0,
      "step": 25070
    },
    {
      "entropy": 5.548755502700805,
      "epoch": 2.4885867407701467,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004383664611024912,
      "loss": 5.2286,
      "mean_token_accuracy": 0.16891767531633378,
      "num_tokens": 54950152.0,
      "step": 25075
    },
    {
      "entropy": 5.515759134292603,
      "epoch": 2.489082969432314,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004383420957618213,
      "loss": 5.2434,
      "mean_token_accuracy": 0.15976694971323013,
      "num_tokens": 54962191.0,
      "step": 25080
    },
    {
      "entropy": 5.460237216949463,
      "epoch": 2.489579198094482,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043831772637061944,
      "loss": 5.0574,
      "mean_token_accuracy": 0.17311082631349564,
      "num_tokens": 54972532.0,
      "step": 25085
    },
    {
      "entropy": 5.558468675613403,
      "epoch": 2.4900754267566496,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043829335292948986,
      "loss": 5.1982,
      "mean_token_accuracy": 0.16510104089975358,
      "num_tokens": 54983212.0,
      "step": 25090
    },
    {
      "entropy": 5.447900104522705,
      "epoch": 2.490571655418817,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043826897543903696,
      "loss": 5.0968,
      "mean_token_accuracy": 0.17600079476833344,
      "num_tokens": 54994094.0,
      "step": 25095
    },
    {
      "entropy": 5.484686470031738,
      "epoch": 2.4910678840809846,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043824459389986534,
      "loss": 5.1903,
      "mean_token_accuracy": 0.16439221799373627,
      "num_tokens": 55004098.0,
      "step": 25100
    },
    {
      "entropy": 5.430561685562134,
      "epoch": 2.491564112743152,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004382202083125795,
      "loss": 5.08,
      "mean_token_accuracy": 0.17327603250741958,
      "num_tokens": 55013680.0,
      "step": 25105
    },
    {
      "entropy": 5.436655139923095,
      "epoch": 2.4920603414053195,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004381958186777842,
      "loss": 5.0515,
      "mean_token_accuracy": 0.17473168075084686,
      "num_tokens": 55024311.0,
      "step": 25110
    },
    {
      "entropy": 5.498648738861084,
      "epoch": 2.492556570067487,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004381714249960843,
      "loss": 5.3183,
      "mean_token_accuracy": 0.15805618911981584,
      "num_tokens": 55036215.0,
      "step": 25115
    },
    {
      "entropy": 5.573705816268921,
      "epoch": 2.4930527987296545,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004381470272680846,
      "loss": 5.1538,
      "mean_token_accuracy": 0.16326837688684465,
      "num_tokens": 55046402.0,
      "step": 25120
    },
    {
      "entropy": 5.498614025115967,
      "epoch": 2.493549027391822,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004381226254943902,
      "loss": 5.0956,
      "mean_token_accuracy": 0.16598724722862243,
      "num_tokens": 55056307.0,
      "step": 25125
    },
    {
      "entropy": 5.468946743011474,
      "epoch": 2.49404525605399,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004380982196756062,
      "loss": 5.1212,
      "mean_token_accuracy": 0.1680994972586632,
      "num_tokens": 55066918.0,
      "step": 25130
    },
    {
      "entropy": 5.5065278053283695,
      "epoch": 2.4945414847161573,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043807380981233777,
      "loss": 5.159,
      "mean_token_accuracy": 0.16354202330112458,
      "num_tokens": 55078166.0,
      "step": 25135
    },
    {
      "entropy": 5.4950339794158936,
      "epoch": 2.495037713378325,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004380493959051903,
      "loss": 5.1815,
      "mean_token_accuracy": 0.16922224909067154,
      "num_tokens": 55090919.0,
      "step": 25140
    },
    {
      "entropy": 5.5446343421936035,
      "epoch": 2.4955339420404923,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043802497795476916,
      "loss": 5.2029,
      "mean_token_accuracy": 0.16707129925489425,
      "num_tokens": 55100512.0,
      "step": 25145
    },
    {
      "entropy": 5.4878379821777346,
      "epoch": 2.4960301707026598,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004380005559616798,
      "loss": 5.1379,
      "mean_token_accuracy": 0.17039562612771988,
      "num_tokens": 55111845.0,
      "step": 25150
    },
    {
      "entropy": 5.589270162582397,
      "epoch": 2.4965263993648272,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004379761299265279,
      "loss": 5.2829,
      "mean_token_accuracy": 0.15797600001096726,
      "num_tokens": 55123744.0,
      "step": 25155
    },
    {
      "entropy": 5.540172910690307,
      "epoch": 2.4970226280269947,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004379516998499192,
      "loss": 5.1702,
      "mean_token_accuracy": 0.16649853736162185,
      "num_tokens": 55133789.0,
      "step": 25160
    },
    {
      "entropy": 5.5170642852783205,
      "epoch": 2.497518856689162,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004379272657324594,
      "loss": 5.1667,
      "mean_token_accuracy": 0.16867182552814483,
      "num_tokens": 55145325.0,
      "step": 25165
    },
    {
      "entropy": 5.502835607528686,
      "epoch": 2.49801508535133,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004379028275747545,
      "loss": 5.1536,
      "mean_token_accuracy": 0.1671243667602539,
      "num_tokens": 55157639.0,
      "step": 25170
    },
    {
      "entropy": 5.481846570968628,
      "epoch": 2.4985113140134976,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043787838537741054,
      "loss": 5.0779,
      "mean_token_accuracy": 0.17261483818292617,
      "num_tokens": 55167796.0,
      "step": 25175
    },
    {
      "entropy": 5.461461114883423,
      "epoch": 2.499007542675665,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004378539391410336,
      "loss": 5.1881,
      "mean_token_accuracy": 0.16382477879524232,
      "num_tokens": 55178384.0,
      "step": 25180
    },
    {
      "entropy": 5.543841123580933,
      "epoch": 2.4995037713378325,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043782948886622987,
      "loss": 5.2257,
      "mean_token_accuracy": 0.1644720107316971,
      "num_tokens": 55188363.0,
      "step": 25185
    },
    {
      "entropy": 5.5675922393798825,
      "epoch": 2.5,
      "grad_norm": 1.125,
      "learning_rate": 0.0004378050345536057,
      "loss": 5.1888,
      "mean_token_accuracy": 0.16643865406513214,
      "num_tokens": 55198293.0,
      "step": 25190
    },
    {
      "entropy": 5.494694662094116,
      "epoch": 2.5004962286621675,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043778057620376745,
      "loss": 5.1543,
      "mean_token_accuracy": 0.1682966470718384,
      "num_tokens": 55208368.0,
      "step": 25195
    },
    {
      "entropy": 5.494013452529908,
      "epoch": 2.500992457324335,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043775611381732164,
      "loss": 5.218,
      "mean_token_accuracy": 0.16450946033000946,
      "num_tokens": 55220075.0,
      "step": 25200
    },
    {
      "entropy": 5.487249898910522,
      "epoch": 2.5014886859865024,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000437731647394875,
      "loss": 5.1527,
      "mean_token_accuracy": 0.16918678134679793,
      "num_tokens": 55231060.0,
      "step": 25205
    },
    {
      "entropy": 5.545643043518067,
      "epoch": 2.50198491464867,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004377071769370341,
      "loss": 5.1661,
      "mean_token_accuracy": 0.16470575630664824,
      "num_tokens": 55242443.0,
      "step": 25210
    },
    {
      "entropy": 5.402154922485352,
      "epoch": 2.5024811433108374,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043768270244440587,
      "loss": 5.1167,
      "mean_token_accuracy": 0.18023592829704285,
      "num_tokens": 55253285.0,
      "step": 25215
    },
    {
      "entropy": 5.440981245040893,
      "epoch": 2.5029773719730053,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043765822391759714,
      "loss": 5.1151,
      "mean_token_accuracy": 0.16870759576559066,
      "num_tokens": 55264694.0,
      "step": 25220
    },
    {
      "entropy": 5.524605417251587,
      "epoch": 2.5034736006351728,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004376337413572149,
      "loss": 5.1456,
      "mean_token_accuracy": 0.16900616139173508,
      "num_tokens": 55275429.0,
      "step": 25225
    },
    {
      "entropy": 5.522401475906372,
      "epoch": 2.5039698292973402,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043760925476386635,
      "loss": 5.2085,
      "mean_token_accuracy": 0.16366634368896485,
      "num_tokens": 55286764.0,
      "step": 25230
    },
    {
      "entropy": 5.3956835746765135,
      "epoch": 2.5044660579595077,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004375847641381587,
      "loss": 5.0536,
      "mean_token_accuracy": 0.17640433162450792,
      "num_tokens": 55296284.0,
      "step": 25235
    },
    {
      "entropy": 5.488988876342773,
      "epoch": 2.504962286621675,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004375602694806993,
      "loss": 5.1437,
      "mean_token_accuracy": 0.17335957884788514,
      "num_tokens": 55307684.0,
      "step": 25240
    },
    {
      "entropy": 5.47346863746643,
      "epoch": 2.5054585152838427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043753577079209537,
      "loss": 5.0936,
      "mean_token_accuracy": 0.17125789672136307,
      "num_tokens": 55318202.0,
      "step": 25245
    },
    {
      "entropy": 5.510216903686524,
      "epoch": 2.5059547439460106,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043751126807295464,
      "loss": 5.1961,
      "mean_token_accuracy": 0.16518656760454178,
      "num_tokens": 55328827.0,
      "step": 25250
    },
    {
      "entropy": 5.491761064529419,
      "epoch": 2.506450972608178,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043748676132388464,
      "loss": 5.1893,
      "mean_token_accuracy": 0.16039085239171982,
      "num_tokens": 55338171.0,
      "step": 25255
    },
    {
      "entropy": 5.431877994537354,
      "epoch": 2.5069472012703455,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043746225054549316,
      "loss": 5.1382,
      "mean_token_accuracy": 0.16868709176778793,
      "num_tokens": 55349081.0,
      "step": 25260
    },
    {
      "entropy": 5.485008192062378,
      "epoch": 2.507443429932513,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043743773573838785,
      "loss": 5.1321,
      "mean_token_accuracy": 0.17158937752246856,
      "num_tokens": 55359344.0,
      "step": 25265
    },
    {
      "entropy": 5.476859903335571,
      "epoch": 2.5079396585946805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043741321690317673,
      "loss": 5.0561,
      "mean_token_accuracy": 0.1742933988571167,
      "num_tokens": 55369643.0,
      "step": 25270
    },
    {
      "entropy": 5.419447135925293,
      "epoch": 2.508435887256848,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004373886940404677,
      "loss": 5.0646,
      "mean_token_accuracy": 0.17575088441371917,
      "num_tokens": 55380812.0,
      "step": 25275
    },
    {
      "entropy": 5.390502691268921,
      "epoch": 2.5089321159190154,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004373641671508691,
      "loss": 5.1303,
      "mean_token_accuracy": 0.17561345547437668,
      "num_tokens": 55393136.0,
      "step": 25280
    },
    {
      "entropy": 5.455101537704468,
      "epoch": 2.509428344581183,
      "grad_norm": 1.046875,
      "learning_rate": 0.000437339636234989,
      "loss": 5.0494,
      "mean_token_accuracy": 0.16825947910547256,
      "num_tokens": 55404317.0,
      "step": 25285
    },
    {
      "entropy": 5.510698699951172,
      "epoch": 2.5099245732433504,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043731510129343574,
      "loss": 5.2124,
      "mean_token_accuracy": 0.16514717787504196,
      "num_tokens": 55415812.0,
      "step": 25290
    },
    {
      "entropy": 5.528195142745972,
      "epoch": 2.510420801905518,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043729056232681766,
      "loss": 5.1723,
      "mean_token_accuracy": 0.16717670410871505,
      "num_tokens": 55426402.0,
      "step": 25295
    },
    {
      "entropy": 5.498236560821534,
      "epoch": 2.5109170305676853,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004372660193357434,
      "loss": 5.2004,
      "mean_token_accuracy": 0.16897969990968703,
      "num_tokens": 55437474.0,
      "step": 25300
    },
    {
      "entropy": 5.5077849388122555,
      "epoch": 2.5114132592298533,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004372414723208215,
      "loss": 5.1265,
      "mean_token_accuracy": 0.16855165660381316,
      "num_tokens": 55447587.0,
      "step": 25305
    },
    {
      "entropy": 5.612525510787964,
      "epoch": 2.5119094878920207,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004372169212826606,
      "loss": 5.3134,
      "mean_token_accuracy": 0.1594523772597313,
      "num_tokens": 55460587.0,
      "step": 25310
    },
    {
      "entropy": 5.519278860092163,
      "epoch": 2.512405716554188,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043719236622186976,
      "loss": 5.1378,
      "mean_token_accuracy": 0.17141944468021392,
      "num_tokens": 55471150.0,
      "step": 25315
    },
    {
      "entropy": 5.442029666900635,
      "epoch": 2.5129019452163557,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004371678071390576,
      "loss": 5.1288,
      "mean_token_accuracy": 0.17171914875507355,
      "num_tokens": 55480368.0,
      "step": 25320
    },
    {
      "entropy": 5.520255756378174,
      "epoch": 2.513398173878523,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043714324403483333,
      "loss": 5.1946,
      "mean_token_accuracy": 0.1646094188094139,
      "num_tokens": 55491954.0,
      "step": 25325
    },
    {
      "entropy": 5.513877725601196,
      "epoch": 2.5138944025406906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043711867690980605,
      "loss": 5.1557,
      "mean_token_accuracy": 0.16543475687503814,
      "num_tokens": 55502560.0,
      "step": 25330
    },
    {
      "entropy": 5.478540706634521,
      "epoch": 2.5143906312028586,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043709410576458477,
      "loss": 5.1464,
      "mean_token_accuracy": 0.16586503237485886,
      "num_tokens": 55514571.0,
      "step": 25335
    },
    {
      "entropy": 5.528464365005493,
      "epoch": 2.514886859865026,
      "grad_norm": 1.125,
      "learning_rate": 0.00043706953059977907,
      "loss": 5.1795,
      "mean_token_accuracy": 0.17013894468545915,
      "num_tokens": 55525206.0,
      "step": 25340
    },
    {
      "entropy": 5.486236524581909,
      "epoch": 2.5153830885271935,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043704495141599816,
      "loss": 5.1904,
      "mean_token_accuracy": 0.16144656240940095,
      "num_tokens": 55536551.0,
      "step": 25345
    },
    {
      "entropy": 5.489563655853272,
      "epoch": 2.515879317189361,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004370203682138517,
      "loss": 5.1389,
      "mean_token_accuracy": 0.16506134271621703,
      "num_tokens": 55547842.0,
      "step": 25350
    },
    {
      "entropy": 5.586037540435791,
      "epoch": 2.5163755458515285,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004369957809939492,
      "loss": 5.1047,
      "mean_token_accuracy": 0.1770031288266182,
      "num_tokens": 55558248.0,
      "step": 25355
    },
    {
      "entropy": 5.450740766525269,
      "epoch": 2.516871774513696,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004369711897569004,
      "loss": 5.0924,
      "mean_token_accuracy": 0.16723080724477768,
      "num_tokens": 55569195.0,
      "step": 25360
    },
    {
      "entropy": 5.470790433883667,
      "epoch": 2.5173680031758634,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004369465945033151,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17224919497966767,
      "num_tokens": 55579649.0,
      "step": 25365
    },
    {
      "entropy": 5.405758380889893,
      "epoch": 2.517864231838031,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043692199523380325,
      "loss": 4.9911,
      "mean_token_accuracy": 0.17933694422245025,
      "num_tokens": 55589550.0,
      "step": 25370
    },
    {
      "entropy": 5.439561557769776,
      "epoch": 2.5183604605001984,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004368973919489748,
      "loss": 5.1436,
      "mean_token_accuracy": 0.16628728210926055,
      "num_tokens": 55600673.0,
      "step": 25375
    },
    {
      "entropy": 5.346926307678222,
      "epoch": 2.518856689162366,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043687278464943986,
      "loss": 5.1005,
      "mean_token_accuracy": 0.16975285112857819,
      "num_tokens": 55612045.0,
      "step": 25380
    },
    {
      "entropy": 5.460721206665039,
      "epoch": 2.5193529178245333,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004368481733358088,
      "loss": 5.1737,
      "mean_token_accuracy": 0.16497254222631455,
      "num_tokens": 55623666.0,
      "step": 25385
    },
    {
      "entropy": 5.470662450790405,
      "epoch": 2.519849146486701,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004368235580086917,
      "loss": 5.1404,
      "mean_token_accuracy": 0.17090693414211272,
      "num_tokens": 55634236.0,
      "step": 25390
    },
    {
      "entropy": 5.4250541687011715,
      "epoch": 2.5203453751488687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043679893866869904,
      "loss": 5.1019,
      "mean_token_accuracy": 0.16903540194034578,
      "num_tokens": 55645412.0,
      "step": 25395
    },
    {
      "entropy": 5.443095636367798,
      "epoch": 2.520841603811036,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004367743153164414,
      "loss": 5.0785,
      "mean_token_accuracy": 0.16979025900363923,
      "num_tokens": 55655861.0,
      "step": 25400
    },
    {
      "entropy": 5.4572735786437985,
      "epoch": 2.5213378324732036,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043674968795252926,
      "loss": 5.1002,
      "mean_token_accuracy": 0.16740308254957198,
      "num_tokens": 55665227.0,
      "step": 25405
    },
    {
      "entropy": 5.4311285495758055,
      "epoch": 2.521834061135371,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004367250565775734,
      "loss": 5.1111,
      "mean_token_accuracy": 0.17267275899648665,
      "num_tokens": 55676104.0,
      "step": 25410
    },
    {
      "entropy": 5.462083864212036,
      "epoch": 2.5223302897975386,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004367004211921847,
      "loss": 5.2075,
      "mean_token_accuracy": 0.1643756166100502,
      "num_tokens": 55688222.0,
      "step": 25415
    },
    {
      "entropy": 5.508143758773803,
      "epoch": 2.522826518459706,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043667578179697387,
      "loss": 5.1574,
      "mean_token_accuracy": 0.16883934289216995,
      "num_tokens": 55700138.0,
      "step": 25420
    },
    {
      "entropy": 5.502370405197143,
      "epoch": 2.523322747121874,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004366511383925521,
      "loss": 5.0885,
      "mean_token_accuracy": 0.17379290759563445,
      "num_tokens": 55709444.0,
      "step": 25425
    },
    {
      "entropy": 5.5519139766693115,
      "epoch": 2.5238189757840415,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004366264909795305,
      "loss": 5.2516,
      "mean_token_accuracy": 0.1669851213693619,
      "num_tokens": 55720327.0,
      "step": 25430
    },
    {
      "entropy": 5.412032890319824,
      "epoch": 2.524315204446209,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004366018395585201,
      "loss": 5.0152,
      "mean_token_accuracy": 0.17469536662101745,
      "num_tokens": 55730727.0,
      "step": 25435
    },
    {
      "entropy": 5.504203224182129,
      "epoch": 2.5248114331083764,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004365771841301323,
      "loss": 5.1724,
      "mean_token_accuracy": 0.1738314986228943,
      "num_tokens": 55741568.0,
      "step": 25440
    },
    {
      "entropy": 5.431506204605102,
      "epoch": 2.525307661770544,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004365525246949786,
      "loss": 5.1165,
      "mean_token_accuracy": 0.17008257806301116,
      "num_tokens": 55752511.0,
      "step": 25445
    },
    {
      "entropy": 5.4787712574005125,
      "epoch": 2.5258038904327114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043652786125367035,
      "loss": 5.1296,
      "mean_token_accuracy": 0.16872530430555344,
      "num_tokens": 55763004.0,
      "step": 25450
    },
    {
      "entropy": 5.527592611312866,
      "epoch": 2.526300119094879,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043650319380681915,
      "loss": 5.1779,
      "mean_token_accuracy": 0.1681112602353096,
      "num_tokens": 55773276.0,
      "step": 25455
    },
    {
      "entropy": 5.511899614334107,
      "epoch": 2.5267963477570463,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004364785223550368,
      "loss": 5.1638,
      "mean_token_accuracy": 0.1663696900010109,
      "num_tokens": 55784607.0,
      "step": 25460
    },
    {
      "entropy": 5.544256114959717,
      "epoch": 2.527292576419214,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043645384689893504,
      "loss": 5.2016,
      "mean_token_accuracy": 0.15452762097120284,
      "num_tokens": 55796730.0,
      "step": 25465
    },
    {
      "entropy": 5.55146427154541,
      "epoch": 2.5277888050813813,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004364291674391258,
      "loss": 5.1857,
      "mean_token_accuracy": 0.1601922482252121,
      "num_tokens": 55806940.0,
      "step": 25470
    },
    {
      "entropy": 5.472479009628296,
      "epoch": 2.528285033743549,
      "grad_norm": 1.0,
      "learning_rate": 0.000436404483976221,
      "loss": 5.1182,
      "mean_token_accuracy": 0.1701662227511406,
      "num_tokens": 55818229.0,
      "step": 25475
    },
    {
      "entropy": 5.4733827114105225,
      "epoch": 2.5287812624057167,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043637979651083285,
      "loss": 5.2091,
      "mean_token_accuracy": 0.16509607881307603,
      "num_tokens": 55831100.0,
      "step": 25480
    },
    {
      "entropy": 5.499732160568238,
      "epoch": 2.529277491067884,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043635510504357343,
      "loss": 5.141,
      "mean_token_accuracy": 0.16303637772798538,
      "num_tokens": 55841713.0,
      "step": 25485
    },
    {
      "entropy": 5.488565731048584,
      "epoch": 2.5297737197300516,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043633040957505515,
      "loss": 5.1213,
      "mean_token_accuracy": 0.16245962530374528,
      "num_tokens": 55852737.0,
      "step": 25490
    },
    {
      "entropy": 5.540199422836304,
      "epoch": 2.530269948392219,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004363057101058903,
      "loss": 5.2153,
      "mean_token_accuracy": 0.1669032320380211,
      "num_tokens": 55863471.0,
      "step": 25495
    },
    {
      "entropy": 5.461113214492798,
      "epoch": 2.5307661770543866,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004362810066366915,
      "loss": 5.0911,
      "mean_token_accuracy": 0.16559586226940154,
      "num_tokens": 55874305.0,
      "step": 25500
    },
    {
      "entropy": 5.52692403793335,
      "epoch": 2.531262405716554,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043625629916807117,
      "loss": 5.2279,
      "mean_token_accuracy": 0.16551934331655502,
      "num_tokens": 55887400.0,
      "step": 25505
    },
    {
      "entropy": 5.455056285858154,
      "epoch": 2.531758634378722,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043623158770064214,
      "loss": 5.1673,
      "mean_token_accuracy": 0.16810615062713624,
      "num_tokens": 55900052.0,
      "step": 25510
    },
    {
      "entropy": 5.671394777297974,
      "epoch": 2.5322548630408894,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004362068722350172,
      "loss": 5.3209,
      "mean_token_accuracy": 0.15895611122250558,
      "num_tokens": 55911609.0,
      "step": 25515
    },
    {
      "entropy": 5.510286521911621,
      "epoch": 2.532751091703057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043618215277180917,
      "loss": 5.1007,
      "mean_token_accuracy": 0.17337660491466522,
      "num_tokens": 55922873.0,
      "step": 25520
    },
    {
      "entropy": 5.472643852233887,
      "epoch": 2.5332473203652244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043615742931163106,
      "loss": 5.1461,
      "mean_token_accuracy": 0.1649425819516182,
      "num_tokens": 55932161.0,
      "step": 25525
    },
    {
      "entropy": 5.5265278816223145,
      "epoch": 2.533743549027392,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000436132701855096,
      "loss": 5.27,
      "mean_token_accuracy": 0.15683186799287796,
      "num_tokens": 55943601.0,
      "step": 25530
    },
    {
      "entropy": 5.41484751701355,
      "epoch": 2.5342397776895593,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043610797040281705,
      "loss": 5.1317,
      "mean_token_accuracy": 0.16963061839342117,
      "num_tokens": 55954981.0,
      "step": 25535
    },
    {
      "entropy": 5.495623970031739,
      "epoch": 2.534736006351727,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004360832349554077,
      "loss": 5.1894,
      "mean_token_accuracy": 0.1631482109427452,
      "num_tokens": 55964407.0,
      "step": 25540
    },
    {
      "entropy": 5.4779150009155275,
      "epoch": 2.5352322350138943,
      "grad_norm": 1.125,
      "learning_rate": 0.0004360584955134811,
      "loss": 5.0844,
      "mean_token_accuracy": 0.16757139861583709,
      "num_tokens": 55974634.0,
      "step": 25545
    },
    {
      "entropy": 5.495425796508789,
      "epoch": 2.5357284636760618,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043603375207765103,
      "loss": 5.1722,
      "mean_token_accuracy": 0.1582540452480316,
      "num_tokens": 55986753.0,
      "step": 25550
    },
    {
      "entropy": 5.466502141952515,
      "epoch": 2.5362246923382292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004360090046485308,
      "loss": 5.124,
      "mean_token_accuracy": 0.1684194803237915,
      "num_tokens": 55997839.0,
      "step": 25555
    },
    {
      "entropy": 5.534509658813477,
      "epoch": 2.536720921000397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004359842532267342,
      "loss": 5.1513,
      "mean_token_accuracy": 0.17103424668312073,
      "num_tokens": 56008765.0,
      "step": 25560
    },
    {
      "entropy": 5.479816961288452,
      "epoch": 2.5372171496625646,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043595949781287505,
      "loss": 5.1821,
      "mean_token_accuracy": 0.1679423436522484,
      "num_tokens": 56020077.0,
      "step": 25565
    },
    {
      "entropy": 5.4479585647583,
      "epoch": 2.537713378324732,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00043593473840756715,
      "loss": 5.1349,
      "mean_token_accuracy": 0.1742580845952034,
      "num_tokens": 56031298.0,
      "step": 25570
    },
    {
      "entropy": 5.5696861267089846,
      "epoch": 2.5382096069868996,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004359099750114245,
      "loss": 5.1922,
      "mean_token_accuracy": 0.16902057975530624,
      "num_tokens": 56043955.0,
      "step": 25575
    },
    {
      "entropy": 5.489414548873901,
      "epoch": 2.538705835649067,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004358852076250613,
      "loss": 5.1382,
      "mean_token_accuracy": 0.17031347006559372,
      "num_tokens": 56055734.0,
      "step": 25580
    },
    {
      "entropy": 5.5368349075317385,
      "epoch": 2.5392020643112345,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043586043624909153,
      "loss": 5.1527,
      "mean_token_accuracy": 0.1700321301817894,
      "num_tokens": 56065472.0,
      "step": 25585
    },
    {
      "entropy": 5.559461641311645,
      "epoch": 2.539698292973402,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004358356608841296,
      "loss": 5.1681,
      "mean_token_accuracy": 0.169459967315197,
      "num_tokens": 56075691.0,
      "step": 25590
    },
    {
      "entropy": 5.47702989578247,
      "epoch": 2.54019452163557,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00043581088153078984,
      "loss": 5.1561,
      "mean_token_accuracy": 0.17165803164243698,
      "num_tokens": 56085900.0,
      "step": 25595
    },
    {
      "entropy": 5.547142744064331,
      "epoch": 2.5406907502977374,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043578609818968676,
      "loss": 5.1648,
      "mean_token_accuracy": 0.16800468862056733,
      "num_tokens": 56097675.0,
      "step": 25600
    },
    {
      "entropy": 5.577536773681641,
      "epoch": 2.541186978959905,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043576131086143496,
      "loss": 5.218,
      "mean_token_accuracy": 0.15844055861234665,
      "num_tokens": 56108102.0,
      "step": 25605
    },
    {
      "entropy": 5.422367334365845,
      "epoch": 2.5416832076220723,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043573651954664897,
      "loss": 5.0624,
      "mean_token_accuracy": 0.16978320032358168,
      "num_tokens": 56118489.0,
      "step": 25610
    },
    {
      "entropy": 5.430506324768066,
      "epoch": 2.54217943628424,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043571172424594363,
      "loss": 5.0473,
      "mean_token_accuracy": 0.17855793684720994,
      "num_tokens": 56127949.0,
      "step": 25615
    },
    {
      "entropy": 5.419778633117676,
      "epoch": 2.5426756649464073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004356869249599339,
      "loss": 5.0973,
      "mean_token_accuracy": 0.16855050921440123,
      "num_tokens": 56138209.0,
      "step": 25620
    },
    {
      "entropy": 5.481716108322144,
      "epoch": 2.5431718936085748,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004356621216892347,
      "loss": 5.175,
      "mean_token_accuracy": 0.16933394968509674,
      "num_tokens": 56149005.0,
      "step": 25625
    },
    {
      "entropy": 5.507562255859375,
      "epoch": 2.5436681222707422,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043563731443446095,
      "loss": 5.0969,
      "mean_token_accuracy": 0.16568993777036667,
      "num_tokens": 56159406.0,
      "step": 25630
    },
    {
      "entropy": 5.461356592178345,
      "epoch": 2.5441643509329097,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043561250319622806,
      "loss": 5.1457,
      "mean_token_accuracy": 0.16840128004550933,
      "num_tokens": 56169998.0,
      "step": 25635
    },
    {
      "entropy": 5.548934650421143,
      "epoch": 2.544660579595077,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043558768797515116,
      "loss": 5.2377,
      "mean_token_accuracy": 0.16502212285995482,
      "num_tokens": 56181096.0,
      "step": 25640
    },
    {
      "entropy": 5.518092918395996,
      "epoch": 2.5451568082572447,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043556286877184566,
      "loss": 5.1535,
      "mean_token_accuracy": 0.16440659761428833,
      "num_tokens": 56191909.0,
      "step": 25645
    },
    {
      "entropy": 5.520347166061401,
      "epoch": 2.5456530369194126,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004355380455869269,
      "loss": 5.1699,
      "mean_token_accuracy": 0.16150893419981002,
      "num_tokens": 56202725.0,
      "step": 25650
    },
    {
      "entropy": 5.493874216079712,
      "epoch": 2.54614926558158,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004355132184210106,
      "loss": 5.1317,
      "mean_token_accuracy": 0.17275637984275818,
      "num_tokens": 56213596.0,
      "step": 25655
    },
    {
      "entropy": 5.429909610748291,
      "epoch": 2.5466454942437475,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004354883872747124,
      "loss": 5.131,
      "mean_token_accuracy": 0.1738610252737999,
      "num_tokens": 56224416.0,
      "step": 25660
    },
    {
      "entropy": 5.4895953178405765,
      "epoch": 2.547141722905915,
      "grad_norm": 1.125,
      "learning_rate": 0.00043546355214864796,
      "loss": 5.145,
      "mean_token_accuracy": 0.17014395594596862,
      "num_tokens": 56234325.0,
      "step": 25665
    },
    {
      "entropy": 5.550550842285157,
      "epoch": 2.5476379515680825,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004354387130434332,
      "loss": 5.1354,
      "mean_token_accuracy": 0.16444703042507172,
      "num_tokens": 56244178.0,
      "step": 25670
    },
    {
      "entropy": 5.479727506637573,
      "epoch": 2.54813418023025,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004354138699596841,
      "loss": 5.1798,
      "mean_token_accuracy": 0.1663906008005142,
      "num_tokens": 56256036.0,
      "step": 25675
    },
    {
      "entropy": 5.4672808170318605,
      "epoch": 2.548630408892418,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004353890228980167,
      "loss": 5.138,
      "mean_token_accuracy": 0.16710673570632933,
      "num_tokens": 56266203.0,
      "step": 25680
    },
    {
      "entropy": 5.4830952167510985,
      "epoch": 2.5491266375545854,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004353641718590471,
      "loss": 5.1611,
      "mean_token_accuracy": 0.1695963129401207,
      "num_tokens": 56276958.0,
      "step": 25685
    },
    {
      "entropy": 5.5398876667022705,
      "epoch": 2.549622866216753,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043533931684339156,
      "loss": 5.1829,
      "mean_token_accuracy": 0.17099445760250093,
      "num_tokens": 56287941.0,
      "step": 25690
    },
    {
      "entropy": 5.4712896823883055,
      "epoch": 2.5501190948789203,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004353144578516665,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17015835791826248,
      "num_tokens": 56299390.0,
      "step": 25695
    },
    {
      "entropy": 5.462546634674072,
      "epoch": 2.5506153235410878,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043528959488448835,
      "loss": 5.1105,
      "mean_token_accuracy": 0.16514968276023864,
      "num_tokens": 56310762.0,
      "step": 25700
    },
    {
      "entropy": 5.5297768115997314,
      "epoch": 2.5511115522032553,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043526472794247365,
      "loss": 5.1363,
      "mean_token_accuracy": 0.17023834586143494,
      "num_tokens": 56322522.0,
      "step": 25705
    },
    {
      "entropy": 5.499858856201172,
      "epoch": 2.5516077808654227,
      "grad_norm": 1.0,
      "learning_rate": 0.00043523985702623896,
      "loss": 5.1876,
      "mean_token_accuracy": 0.163614721596241,
      "num_tokens": 56334124.0,
      "step": 25710
    },
    {
      "entropy": 5.455598592758179,
      "epoch": 2.55210400952759,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043521498213640113,
      "loss": 5.1153,
      "mean_token_accuracy": 0.1698628082871437,
      "num_tokens": 56344761.0,
      "step": 25715
    },
    {
      "entropy": 5.424559116363525,
      "epoch": 2.5526002381897577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043519010327357704,
      "loss": 5.0842,
      "mean_token_accuracy": 0.16810876429080962,
      "num_tokens": 56355518.0,
      "step": 25720
    },
    {
      "entropy": 5.490734577178955,
      "epoch": 2.553096466851925,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004351652204383836,
      "loss": 5.1779,
      "mean_token_accuracy": 0.17197754234075546,
      "num_tokens": 56366307.0,
      "step": 25725
    },
    {
      "entropy": 5.531389999389648,
      "epoch": 2.5535926955140926,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004351403336314377,
      "loss": 5.0948,
      "mean_token_accuracy": 0.17290879040956497,
      "num_tokens": 56377065.0,
      "step": 25730
    },
    {
      "entropy": 5.503006315231323,
      "epoch": 2.5540889241762605,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004351154428533566,
      "loss": 5.1539,
      "mean_token_accuracy": 0.1705860286951065,
      "num_tokens": 56387741.0,
      "step": 25735
    },
    {
      "entropy": 5.473850250244141,
      "epoch": 2.554585152838428,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004350905481047576,
      "loss": 5.3118,
      "mean_token_accuracy": 0.16133136302232742,
      "num_tokens": 56398927.0,
      "step": 25740
    },
    {
      "entropy": 5.584835004806519,
      "epoch": 2.5550813815005955,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004350656493862579,
      "loss": 5.2704,
      "mean_token_accuracy": 0.16070390939712526,
      "num_tokens": 56408791.0,
      "step": 25745
    },
    {
      "entropy": 5.520975589752197,
      "epoch": 2.555577610162763,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043504074669847517,
      "loss": 5.0882,
      "mean_token_accuracy": 0.1767783507704735,
      "num_tokens": 56419206.0,
      "step": 25750
    },
    {
      "entropy": 5.596791791915893,
      "epoch": 2.5560738388249304,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004350158400420266,
      "loss": 5.2805,
      "mean_token_accuracy": 0.16050629764795304,
      "num_tokens": 56431774.0,
      "step": 25755
    },
    {
      "entropy": 5.520423316955567,
      "epoch": 2.556570067487098,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043499092941753003,
      "loss": 5.1725,
      "mean_token_accuracy": 0.16570473909378053,
      "num_tokens": 56442048.0,
      "step": 25760
    },
    {
      "entropy": 5.4691935062408445,
      "epoch": 2.557066296149266,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043496601482560327,
      "loss": 5.1492,
      "mean_token_accuracy": 0.1624267280101776,
      "num_tokens": 56453347.0,
      "step": 25765
    },
    {
      "entropy": 5.485492134094239,
      "epoch": 2.5575625248114333,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043494109626686394,
      "loss": 5.1099,
      "mean_token_accuracy": 0.17653509676456453,
      "num_tokens": 56465323.0,
      "step": 25770
    },
    {
      "entropy": 5.488577032089234,
      "epoch": 2.558058753473601,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043491617374193017,
      "loss": 5.1547,
      "mean_token_accuracy": 0.1666026756167412,
      "num_tokens": 56474704.0,
      "step": 25775
    },
    {
      "entropy": 5.443729829788208,
      "epoch": 2.5585549821357683,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004348912472514197,
      "loss": 5.1874,
      "mean_token_accuracy": 0.16869424134492875,
      "num_tokens": 56485028.0,
      "step": 25780
    },
    {
      "entropy": 5.504279088973999,
      "epoch": 2.5590512107979357,
      "grad_norm": 1.46875,
      "learning_rate": 0.00043486631679595097,
      "loss": 5.0752,
      "mean_token_accuracy": 0.17121723294258118,
      "num_tokens": 56496169.0,
      "step": 25785
    },
    {
      "entropy": 5.449408864974975,
      "epoch": 2.559547439460103,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004348413823761421,
      "loss": 5.0661,
      "mean_token_accuracy": 0.17120427936315535,
      "num_tokens": 56506324.0,
      "step": 25790
    },
    {
      "entropy": 5.4176185607910154,
      "epoch": 2.5600436681222707,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004348164439926112,
      "loss": 5.0545,
      "mean_token_accuracy": 0.16928955912590027,
      "num_tokens": 56516901.0,
      "step": 25795
    },
    {
      "entropy": 5.41863317489624,
      "epoch": 2.560539896784438,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043479150164597695,
      "loss": 5.0462,
      "mean_token_accuracy": 0.17429985404014586,
      "num_tokens": 56527241.0,
      "step": 25800
    },
    {
      "entropy": 5.543785858154297,
      "epoch": 2.5610361254466056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043476655533685783,
      "loss": 5.2624,
      "mean_token_accuracy": 0.16226056814193726,
      "num_tokens": 56539078.0,
      "step": 25805
    },
    {
      "entropy": 5.469307327270508,
      "epoch": 2.561532354108773,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043474160506587234,
      "loss": 5.0939,
      "mean_token_accuracy": 0.16935459822416304,
      "num_tokens": 56551141.0,
      "step": 25810
    },
    {
      "entropy": 5.5096698760986325,
      "epoch": 2.5620285827709406,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004347166508336393,
      "loss": 5.2064,
      "mean_token_accuracy": 0.16620081961154937,
      "num_tokens": 56561836.0,
      "step": 25815
    },
    {
      "entropy": 5.504616022109985,
      "epoch": 2.5625248114331085,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043469169264077754,
      "loss": 5.1238,
      "mean_token_accuracy": 0.17154245376586913,
      "num_tokens": 56572699.0,
      "step": 25820
    },
    {
      "entropy": 5.5093255043029785,
      "epoch": 2.563021040095276,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004346667304879058,
      "loss": 5.2038,
      "mean_token_accuracy": 0.16640128195285797,
      "num_tokens": 56582296.0,
      "step": 25825
    },
    {
      "entropy": 5.466437244415284,
      "epoch": 2.5635172687574435,
      "grad_norm": 1.0,
      "learning_rate": 0.00043464176437564317,
      "loss": 5.1296,
      "mean_token_accuracy": 0.16634306460618972,
      "num_tokens": 56592907.0,
      "step": 25830
    },
    {
      "entropy": 5.466111373901367,
      "epoch": 2.564013497419611,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004346167943046089,
      "loss": 5.1271,
      "mean_token_accuracy": 0.1777269273996353,
      "num_tokens": 56604995.0,
      "step": 25835
    },
    {
      "entropy": 5.4422682762146,
      "epoch": 2.5645097260817784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000434591820275422,
      "loss": 5.0741,
      "mean_token_accuracy": 0.17216482907533645,
      "num_tokens": 56614809.0,
      "step": 25840
    },
    {
      "entropy": 5.516414594650269,
      "epoch": 2.565005954743946,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004345668422887018,
      "loss": 5.1685,
      "mean_token_accuracy": 0.1628707766532898,
      "num_tokens": 56624409.0,
      "step": 25845
    },
    {
      "entropy": 5.491698598861694,
      "epoch": 2.565502183406114,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004345418603450678,
      "loss": 5.1545,
      "mean_token_accuracy": 0.16442273408174515,
      "num_tokens": 56634596.0,
      "step": 25850
    },
    {
      "entropy": 5.478093099594116,
      "epoch": 2.5659984120682813,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004345168744451395,
      "loss": 5.1471,
      "mean_token_accuracy": 0.1707705244421959,
      "num_tokens": 56646710.0,
      "step": 25855
    },
    {
      "entropy": 5.475478649139404,
      "epoch": 2.5664946407304488,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004344918845895364,
      "loss": 5.1278,
      "mean_token_accuracy": 0.16848275512456895,
      "num_tokens": 56658427.0,
      "step": 25860
    },
    {
      "entropy": 5.532387399673462,
      "epoch": 2.5669908693926162,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004344668907788782,
      "loss": 5.2026,
      "mean_token_accuracy": 0.16496639251708983,
      "num_tokens": 56669039.0,
      "step": 25865
    },
    {
      "entropy": 5.49155068397522,
      "epoch": 2.5674870980547837,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004344418930137848,
      "loss": 5.1075,
      "mean_token_accuracy": 0.17322129607200623,
      "num_tokens": 56679063.0,
      "step": 25870
    },
    {
      "entropy": 5.5149413585662845,
      "epoch": 2.567983326716951,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004344168912948759,
      "loss": 5.1558,
      "mean_token_accuracy": 0.16716093122959136,
      "num_tokens": 56690142.0,
      "step": 25875
    },
    {
      "entropy": 5.503774833679199,
      "epoch": 2.5684795553791187,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043439188562277174,
      "loss": 5.1769,
      "mean_token_accuracy": 0.16737297773361207,
      "num_tokens": 56700918.0,
      "step": 25880
    },
    {
      "entropy": 5.62050609588623,
      "epoch": 2.568975784041286,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043436687599809226,
      "loss": 5.3021,
      "mean_token_accuracy": 0.15982478260993957,
      "num_tokens": 56713520.0,
      "step": 25885
    },
    {
      "entropy": 5.439434480667114,
      "epoch": 2.5694720127034536,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043434186242145763,
      "loss": 5.1491,
      "mean_token_accuracy": 0.16882216930389404,
      "num_tokens": 56723428.0,
      "step": 25890
    },
    {
      "entropy": 5.476828336715698,
      "epoch": 2.569968241365621,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004343168448934881,
      "loss": 5.1138,
      "mean_token_accuracy": 0.17039965242147445,
      "num_tokens": 56734167.0,
      "step": 25895
    },
    {
      "entropy": 5.484786748886108,
      "epoch": 2.5704644700277886,
      "grad_norm": 3.328125,
      "learning_rate": 0.00043429182341480423,
      "loss": 5.0738,
      "mean_token_accuracy": 0.17245313972234727,
      "num_tokens": 56745196.0,
      "step": 25900
    },
    {
      "entropy": 5.444079780578614,
      "epoch": 2.5709606986899565,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004342667979860263,
      "loss": 5.0737,
      "mean_token_accuracy": 0.1657610833644867,
      "num_tokens": 56756166.0,
      "step": 25905
    },
    {
      "entropy": 5.453773117065429,
      "epoch": 2.571456927352124,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043424176860777495,
      "loss": 5.1041,
      "mean_token_accuracy": 0.16793687492609025,
      "num_tokens": 56766690.0,
      "step": 25910
    },
    {
      "entropy": 5.480613088607788,
      "epoch": 2.5719531560142914,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043421673528067095,
      "loss": 5.1693,
      "mean_token_accuracy": 0.16512681543827057,
      "num_tokens": 56776995.0,
      "step": 25915
    },
    {
      "entropy": 5.563981056213379,
      "epoch": 2.572449384676459,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043419169800533496,
      "loss": 5.2417,
      "mean_token_accuracy": 0.16664236783981323,
      "num_tokens": 56788509.0,
      "step": 25920
    },
    {
      "entropy": 5.517114114761353,
      "epoch": 2.5729456133386264,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004341666567823879,
      "loss": 5.1904,
      "mean_token_accuracy": 0.16096456944942475,
      "num_tokens": 56799027.0,
      "step": 25925
    },
    {
      "entropy": 5.538741254806519,
      "epoch": 2.573441842000794,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004341416116124508,
      "loss": 5.179,
      "mean_token_accuracy": 0.16692944690585138,
      "num_tokens": 56809796.0,
      "step": 25930
    },
    {
      "entropy": 5.495456409454346,
      "epoch": 2.5739380706629618,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004341165624961445,
      "loss": 5.1611,
      "mean_token_accuracy": 0.1682528704404831,
      "num_tokens": 56820523.0,
      "step": 25935
    },
    {
      "entropy": 5.493086910247802,
      "epoch": 2.5744342993251292,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043409150943409035,
      "loss": 5.1605,
      "mean_token_accuracy": 0.16122753471136092,
      "num_tokens": 56831668.0,
      "step": 25940
    },
    {
      "entropy": 5.400396680831909,
      "epoch": 2.5749305279872967,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043406645242690967,
      "loss": 5.0315,
      "mean_token_accuracy": 0.1733199581503868,
      "num_tokens": 56841950.0,
      "step": 25945
    },
    {
      "entropy": 5.469219017028808,
      "epoch": 2.575426756649464,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004340413914752237,
      "loss": 5.1507,
      "mean_token_accuracy": 0.16836681067943574,
      "num_tokens": 56852190.0,
      "step": 25950
    },
    {
      "entropy": 5.566040182113648,
      "epoch": 2.5759229853116317,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004340163265796539,
      "loss": 5.2044,
      "mean_token_accuracy": 0.16403166502714156,
      "num_tokens": 56864255.0,
      "step": 25955
    },
    {
      "entropy": 5.513579940795898,
      "epoch": 2.576419213973799,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004339912577408218,
      "loss": 5.0936,
      "mean_token_accuracy": 0.1722526580095291,
      "num_tokens": 56874711.0,
      "step": 25960
    },
    {
      "entropy": 5.405797719955444,
      "epoch": 2.5769154426359666,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004339661849593492,
      "loss": 5.0883,
      "mean_token_accuracy": 0.1721423625946045,
      "num_tokens": 56885790.0,
      "step": 25965
    },
    {
      "entropy": 5.47713270187378,
      "epoch": 2.577411671298134,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004339411082358577,
      "loss": 5.1528,
      "mean_token_accuracy": 0.17126154899597168,
      "num_tokens": 56896452.0,
      "step": 25970
    },
    {
      "entropy": 5.532026576995849,
      "epoch": 2.5779078999603016,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043391602757096923,
      "loss": 5.2564,
      "mean_token_accuracy": 0.1642930641770363,
      "num_tokens": 56908216.0,
      "step": 25975
    },
    {
      "entropy": 5.523578691482544,
      "epoch": 2.578404128622469,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043389094296530565,
      "loss": 5.1349,
      "mean_token_accuracy": 0.16599819511175157,
      "num_tokens": 56918346.0,
      "step": 25980
    },
    {
      "entropy": 5.490894317626953,
      "epoch": 2.5789003572846365,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004338658544194891,
      "loss": 5.2731,
      "mean_token_accuracy": 0.15528505593538283,
      "num_tokens": 56930950.0,
      "step": 25985
    },
    {
      "entropy": 5.515603542327881,
      "epoch": 2.5793965859468044,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043384076193414176,
      "loss": 5.1628,
      "mean_token_accuracy": 0.16958600282669067,
      "num_tokens": 56940622.0,
      "step": 25990
    },
    {
      "entropy": 5.491501903533935,
      "epoch": 2.579892814608972,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043381566550988564,
      "loss": 5.1358,
      "mean_token_accuracy": 0.17122194468975066,
      "num_tokens": 56952463.0,
      "step": 25995
    },
    {
      "entropy": 5.49561071395874,
      "epoch": 2.5803890432711394,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043379056514734334,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17213888317346573,
      "num_tokens": 56961705.0,
      "step": 26000
    },
    {
      "entropy": 5.4874265670776365,
      "epoch": 2.580885271933307,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004337654608471372,
      "loss": 5.1585,
      "mean_token_accuracy": 0.16250351071357727,
      "num_tokens": 56972359.0,
      "step": 26005
    },
    {
      "entropy": 5.490733957290649,
      "epoch": 2.5813815005954743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043374035260988964,
      "loss": 5.1456,
      "mean_token_accuracy": 0.17149323225021362,
      "num_tokens": 56983342.0,
      "step": 26010
    },
    {
      "entropy": 5.474743175506592,
      "epoch": 2.581877729257642,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043371524043622344,
      "loss": 5.0974,
      "mean_token_accuracy": 0.17366307973861694,
      "num_tokens": 56993859.0,
      "step": 26015
    },
    {
      "entropy": 5.5230786323547365,
      "epoch": 2.5823739579198093,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004336901243267614,
      "loss": 5.2355,
      "mean_token_accuracy": 0.16152314245700836,
      "num_tokens": 57006631.0,
      "step": 26020
    },
    {
      "entropy": 5.5174041271209715,
      "epoch": 2.582870186581977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000433665004282126,
      "loss": 5.1695,
      "mean_token_accuracy": 0.1633543774485588,
      "num_tokens": 57018121.0,
      "step": 26025
    },
    {
      "entropy": 5.46158242225647,
      "epoch": 2.5833664152441447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043363988030294055,
      "loss": 5.108,
      "mean_token_accuracy": 0.16511723548173904,
      "num_tokens": 57028127.0,
      "step": 26030
    },
    {
      "entropy": 5.4693992137908936,
      "epoch": 2.583862643906312,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004336147523898279,
      "loss": 5.095,
      "mean_token_accuracy": 0.17260410636663437,
      "num_tokens": 57038523.0,
      "step": 26035
    },
    {
      "entropy": 5.440100908279419,
      "epoch": 2.5843588725684796,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004335896205434111,
      "loss": 5.0993,
      "mean_token_accuracy": 0.17068338990211487,
      "num_tokens": 57048641.0,
      "step": 26040
    },
    {
      "entropy": 5.399348449707031,
      "epoch": 2.584855101230647,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043356448476431354,
      "loss": 5.0344,
      "mean_token_accuracy": 0.17417972683906555,
      "num_tokens": 57058794.0,
      "step": 26045
    },
    {
      "entropy": 5.496573114395142,
      "epoch": 2.5853513298928146,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004335393450531583,
      "loss": 5.1471,
      "mean_token_accuracy": 0.16600463390350342,
      "num_tokens": 57069814.0,
      "step": 26050
    },
    {
      "entropy": 5.614266777038575,
      "epoch": 2.585847558554982,
      "grad_norm": 1.265625,
      "learning_rate": 0.000433514201410569,
      "loss": 5.2153,
      "mean_token_accuracy": 0.1703847110271454,
      "num_tokens": 57081053.0,
      "step": 26055
    },
    {
      "entropy": 5.459247636795044,
      "epoch": 2.5863437872171495,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043348905383716905,
      "loss": 5.1165,
      "mean_token_accuracy": 0.17315464913845063,
      "num_tokens": 57091386.0,
      "step": 26060
    },
    {
      "entropy": 5.483805084228516,
      "epoch": 2.586840015879317,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043346390233358213,
      "loss": 5.1015,
      "mean_token_accuracy": 0.1681724578142166,
      "num_tokens": 57102659.0,
      "step": 26065
    },
    {
      "entropy": 5.544629716873169,
      "epoch": 2.5873362445414845,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004334387469004319,
      "loss": 5.1575,
      "mean_token_accuracy": 0.17150852382183074,
      "num_tokens": 57114257.0,
      "step": 26070
    },
    {
      "entropy": 5.477230262756348,
      "epoch": 2.5878324732036524,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004334135875383421,
      "loss": 5.1748,
      "mean_token_accuracy": 0.1662599414587021,
      "num_tokens": 57125037.0,
      "step": 26075
    },
    {
      "entropy": 5.460994052886963,
      "epoch": 2.58832870186582,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004333884242479367,
      "loss": 5.1538,
      "mean_token_accuracy": 0.16380897909402847,
      "num_tokens": 57137581.0,
      "step": 26080
    },
    {
      "entropy": 5.448868656158448,
      "epoch": 2.5888249305279873,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043336325702983966,
      "loss": 5.0974,
      "mean_token_accuracy": 0.17309724539518356,
      "num_tokens": 57148507.0,
      "step": 26085
    },
    {
      "entropy": 5.477474212646484,
      "epoch": 2.589321159190155,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043333808588467504,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1758417397737503,
      "num_tokens": 57159518.0,
      "step": 26090
    },
    {
      "entropy": 5.428126811981201,
      "epoch": 2.5898173878523223,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004333129108130671,
      "loss": 5.1083,
      "mean_token_accuracy": 0.17162669003009795,
      "num_tokens": 57170600.0,
      "step": 26095
    },
    {
      "entropy": 5.457097482681275,
      "epoch": 2.5903136165144898,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004332877318156402,
      "loss": 5.1352,
      "mean_token_accuracy": 0.1694913774728775,
      "num_tokens": 57181137.0,
      "step": 26100
    },
    {
      "entropy": 5.476279783248901,
      "epoch": 2.5908098451766572,
      "grad_norm": 1.0,
      "learning_rate": 0.00043326254889301856,
      "loss": 5.1256,
      "mean_token_accuracy": 0.16858483999967575,
      "num_tokens": 57191975.0,
      "step": 26105
    },
    {
      "entropy": 5.576321458816528,
      "epoch": 2.591306073838825,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004332373620458268,
      "loss": 5.1845,
      "mean_token_accuracy": 0.16065366119146346,
      "num_tokens": 57203139.0,
      "step": 26110
    },
    {
      "entropy": 5.484392786026001,
      "epoch": 2.5918023025009926,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043321217127468935,
      "loss": 5.2219,
      "mean_token_accuracy": 0.15444684475660325,
      "num_tokens": 57215542.0,
      "step": 26115
    },
    {
      "entropy": 5.50908842086792,
      "epoch": 2.59229853116316,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043318697658023097,
      "loss": 5.1416,
      "mean_token_accuracy": 0.1678308755159378,
      "num_tokens": 57226525.0,
      "step": 26120
    },
    {
      "entropy": 5.477509784698486,
      "epoch": 2.5927947598253276,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004331617779630765,
      "loss": 5.0833,
      "mean_token_accuracy": 0.1771938219666481,
      "num_tokens": 57236120.0,
      "step": 26125
    },
    {
      "entropy": 5.387144231796265,
      "epoch": 2.593290988487495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004331365754238507,
      "loss": 5.0598,
      "mean_token_accuracy": 0.17246893644332886,
      "num_tokens": 57246947.0,
      "step": 26130
    },
    {
      "entropy": 5.456854391098022,
      "epoch": 2.5937872171496625,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043311136896317855,
      "loss": 5.0868,
      "mean_token_accuracy": 0.16742828339338303,
      "num_tokens": 57258885.0,
      "step": 26135
    },
    {
      "entropy": 5.541700601577759,
      "epoch": 2.59428344581183,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043308615858168517,
      "loss": 5.1922,
      "mean_token_accuracy": 0.16770408004522325,
      "num_tokens": 57270047.0,
      "step": 26140
    },
    {
      "entropy": 5.568552446365357,
      "epoch": 2.5947796744739975,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004330609442799957,
      "loss": 5.2441,
      "mean_token_accuracy": 0.1685725063085556,
      "num_tokens": 57281405.0,
      "step": 26145
    },
    {
      "entropy": 5.522154951095581,
      "epoch": 2.595275903136165,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004330357260587355,
      "loss": 5.1453,
      "mean_token_accuracy": 0.17437975555658342,
      "num_tokens": 57292880.0,
      "step": 26150
    },
    {
      "entropy": 5.471359395980835,
      "epoch": 2.5957721317983324,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004330105039185297,
      "loss": 5.1129,
      "mean_token_accuracy": 0.17081110030412675,
      "num_tokens": 57304544.0,
      "step": 26155
    },
    {
      "entropy": 5.477781963348389,
      "epoch": 2.5962683604605004,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043298527786000393,
      "loss": 5.1158,
      "mean_token_accuracy": 0.168549644947052,
      "num_tokens": 57315869.0,
      "step": 26160
    },
    {
      "entropy": 5.481580781936645,
      "epoch": 2.596764589122668,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004329600478837837,
      "loss": 5.1506,
      "mean_token_accuracy": 0.167061448097229,
      "num_tokens": 57327933.0,
      "step": 26165
    },
    {
      "entropy": 5.5104857921600345,
      "epoch": 2.5972608177848353,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043293481399049474,
      "loss": 5.1435,
      "mean_token_accuracy": 0.1686350718140602,
      "num_tokens": 57338941.0,
      "step": 26170
    },
    {
      "entropy": 5.484836053848267,
      "epoch": 2.597757046447003,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004329095761807626,
      "loss": 5.1553,
      "mean_token_accuracy": 0.16733790040016175,
      "num_tokens": 57350128.0,
      "step": 26175
    },
    {
      "entropy": 5.421334266662598,
      "epoch": 2.5982532751091703,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004328843344552133,
      "loss": 5.109,
      "mean_token_accuracy": 0.16963373273611068,
      "num_tokens": 57361602.0,
      "step": 26180
    },
    {
      "entropy": 5.556619119644165,
      "epoch": 2.5987495037713377,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004328590888144726,
      "loss": 5.1982,
      "mean_token_accuracy": 0.16085220873355865,
      "num_tokens": 57373093.0,
      "step": 26185
    },
    {
      "entropy": 5.579673099517822,
      "epoch": 2.599245732433505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004328338392591668,
      "loss": 5.219,
      "mean_token_accuracy": 0.1629318341612816,
      "num_tokens": 57384799.0,
      "step": 26190
    },
    {
      "entropy": 5.435081624984742,
      "epoch": 2.599741961095673,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043280858578992185,
      "loss": 5.0775,
      "mean_token_accuracy": 0.17561801970005037,
      "num_tokens": 57395703.0,
      "step": 26195
    },
    {
      "entropy": 5.496482801437378,
      "epoch": 2.6002381897578406,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043278332840736397,
      "loss": 5.2347,
      "mean_token_accuracy": 0.16552955359220506,
      "num_tokens": 57406725.0,
      "step": 26200
    },
    {
      "entropy": 5.473227024078369,
      "epoch": 2.600734418420008,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043275806711211957,
      "loss": 5.0947,
      "mean_token_accuracy": 0.1720416083931923,
      "num_tokens": 57417362.0,
      "step": 26205
    },
    {
      "entropy": 5.545916509628296,
      "epoch": 2.6012306470821756,
      "grad_norm": 1.125,
      "learning_rate": 0.00043273280190481503,
      "loss": 5.1283,
      "mean_token_accuracy": 0.1690868079662323,
      "num_tokens": 57429178.0,
      "step": 26210
    },
    {
      "entropy": 5.482117128372193,
      "epoch": 2.601726875744343,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043270753278607687,
      "loss": 5.1193,
      "mean_token_accuracy": 0.17176716178655624,
      "num_tokens": 57439741.0,
      "step": 26215
    },
    {
      "entropy": 5.440033864974976,
      "epoch": 2.6022231044065105,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043268225975653176,
      "loss": 5.2014,
      "mean_token_accuracy": 0.1677505925297737,
      "num_tokens": 57451754.0,
      "step": 26220
    },
    {
      "entropy": 5.527003908157349,
      "epoch": 2.602719333068678,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004326569828168064,
      "loss": 5.1613,
      "mean_token_accuracy": 0.16736677289009094,
      "num_tokens": 57463396.0,
      "step": 26225
    },
    {
      "entropy": 5.4501368522644045,
      "epoch": 2.6032155617308455,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004326317019675275,
      "loss": 5.1253,
      "mean_token_accuracy": 0.1677732512354851,
      "num_tokens": 57474023.0,
      "step": 26230
    },
    {
      "entropy": 5.4938469409942625,
      "epoch": 2.603711790393013,
      "grad_norm": 1.125,
      "learning_rate": 0.00043260641720932214,
      "loss": 5.1172,
      "mean_token_accuracy": 0.1730557233095169,
      "num_tokens": 57484751.0,
      "step": 26235
    },
    {
      "entropy": 5.46383581161499,
      "epoch": 2.6042080190551804,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043258112854281726,
      "loss": 5.1125,
      "mean_token_accuracy": 0.16758394688367845,
      "num_tokens": 57496502.0,
      "step": 26240
    },
    {
      "entropy": 5.469937658309936,
      "epoch": 2.6047042477173483,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004325558359686399,
      "loss": 5.1496,
      "mean_token_accuracy": 0.16853010803461074,
      "num_tokens": 57507111.0,
      "step": 26245
    },
    {
      "entropy": 5.4946094989776615,
      "epoch": 2.605200476379516,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043253053948741734,
      "loss": 5.2224,
      "mean_token_accuracy": 0.15818072706460953,
      "num_tokens": 57517844.0,
      "step": 26250
    },
    {
      "entropy": 5.543411207199097,
      "epoch": 2.6056967050416833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043250523909977684,
      "loss": 5.1822,
      "mean_token_accuracy": 0.16756487488746644,
      "num_tokens": 57528047.0,
      "step": 26255
    },
    {
      "entropy": 5.553888750076294,
      "epoch": 2.6061929337038507,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004324799348063457,
      "loss": 5.1265,
      "mean_token_accuracy": 0.16592828929424286,
      "num_tokens": 57538489.0,
      "step": 26260
    },
    {
      "entropy": 5.480022239685058,
      "epoch": 2.606689162366018,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004324546266077518,
      "loss": 5.1366,
      "mean_token_accuracy": 0.17244988083839416,
      "num_tokens": 57549741.0,
      "step": 26265
    },
    {
      "entropy": 5.454023265838623,
      "epoch": 2.6071853910281857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004324293145046222,
      "loss": 5.0715,
      "mean_token_accuracy": 0.18007533997297287,
      "num_tokens": 57559803.0,
      "step": 26270
    },
    {
      "entropy": 5.474315547943116,
      "epoch": 2.607681619690353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043240399849758484,
      "loss": 5.1315,
      "mean_token_accuracy": 0.16763535439968108,
      "num_tokens": 57570609.0,
      "step": 26275
    },
    {
      "entropy": 5.589650106430054,
      "epoch": 2.608177848352521,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004323786785872675,
      "loss": 5.1554,
      "mean_token_accuracy": 0.1722498655319214,
      "num_tokens": 57580934.0,
      "step": 26280
    },
    {
      "entropy": 5.508875942230224,
      "epoch": 2.6086740770146886,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004323533547742981,
      "loss": 5.2596,
      "mean_token_accuracy": 0.16570896804332733,
      "num_tokens": 57594671.0,
      "step": 26285
    },
    {
      "entropy": 5.5372621536254885,
      "epoch": 2.609170305676856,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004323280270593046,
      "loss": 5.2077,
      "mean_token_accuracy": 0.1620237797498703,
      "num_tokens": 57605582.0,
      "step": 26290
    },
    {
      "entropy": 5.502821969985962,
      "epoch": 2.6096665343390235,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043230269544291503,
      "loss": 5.1268,
      "mean_token_accuracy": 0.16495517641305923,
      "num_tokens": 57617185.0,
      "step": 26295
    },
    {
      "entropy": 5.471487617492675,
      "epoch": 2.610162763001191,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004322773599257575,
      "loss": 5.0744,
      "mean_token_accuracy": 0.17665868252515793,
      "num_tokens": 57627205.0,
      "step": 26300
    },
    {
      "entropy": 5.397862243652344,
      "epoch": 2.6106589916633585,
      "grad_norm": 1.0,
      "learning_rate": 0.00043225202050846043,
      "loss": 5.0715,
      "mean_token_accuracy": 0.17036147713661193,
      "num_tokens": 57639496.0,
      "step": 26305
    },
    {
      "entropy": 5.553127288818359,
      "epoch": 2.611155220325526,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000432226677191652,
      "loss": 5.1958,
      "mean_token_accuracy": 0.16213841438293458,
      "num_tokens": 57650345.0,
      "step": 26310
    },
    {
      "entropy": 5.461737108230591,
      "epoch": 2.6116514489876934,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043220132997596084,
      "loss": 5.0805,
      "mean_token_accuracy": 0.174664106965065,
      "num_tokens": 57661409.0,
      "step": 26315
    },
    {
      "entropy": 5.565732097625732,
      "epoch": 2.612147677649861,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004321759788620155,
      "loss": 5.2461,
      "mean_token_accuracy": 0.15908790975809098,
      "num_tokens": 57672639.0,
      "step": 26320
    },
    {
      "entropy": 5.510355710983276,
      "epoch": 2.6126439063120284,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043215062385044446,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1602013237774372,
      "num_tokens": 57682249.0,
      "step": 26325
    },
    {
      "entropy": 5.443299245834351,
      "epoch": 2.613140134974196,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043212526494187665,
      "loss": 5.0482,
      "mean_token_accuracy": 0.17838987112045288,
      "num_tokens": 57692319.0,
      "step": 26330
    },
    {
      "entropy": 5.532259702682495,
      "epoch": 2.6136363636363638,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004320999021369408,
      "loss": 5.2322,
      "mean_token_accuracy": 0.1681890070438385,
      "num_tokens": 57702334.0,
      "step": 26335
    },
    {
      "entropy": 5.505326175689698,
      "epoch": 2.6141325922985312,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043207453543626587,
      "loss": 5.121,
      "mean_token_accuracy": 0.16794671714305878,
      "num_tokens": 57712407.0,
      "step": 26340
    },
    {
      "entropy": 5.4483846664428714,
      "epoch": 2.6146288209606987,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043204916484048094,
      "loss": 5.0993,
      "mean_token_accuracy": 0.1699115291237831,
      "num_tokens": 57724252.0,
      "step": 26345
    },
    {
      "entropy": 5.495292234420776,
      "epoch": 2.615125049622866,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043202379035021515,
      "loss": 5.172,
      "mean_token_accuracy": 0.16437052488327025,
      "num_tokens": 57735033.0,
      "step": 26350
    },
    {
      "entropy": 5.5561846733093265,
      "epoch": 2.6156212782850337,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004319984119660977,
      "loss": 5.1301,
      "mean_token_accuracy": 0.16878708451986313,
      "num_tokens": 57746182.0,
      "step": 26355
    },
    {
      "entropy": 5.479821014404297,
      "epoch": 2.616117506947201,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043197302968875795,
      "loss": 5.0841,
      "mean_token_accuracy": 0.1670690268278122,
      "num_tokens": 57756479.0,
      "step": 26360
    },
    {
      "entropy": 5.5279097080230715,
      "epoch": 2.616613735609369,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004319476435188254,
      "loss": 5.2005,
      "mean_token_accuracy": 0.16088382750749589,
      "num_tokens": 57767773.0,
      "step": 26365
    },
    {
      "entropy": 5.497365999221802,
      "epoch": 2.6171099642715365,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043192225345692933,
      "loss": 5.1064,
      "mean_token_accuracy": 0.16760663241147994,
      "num_tokens": 57777331.0,
      "step": 26370
    },
    {
      "entropy": 5.426784801483154,
      "epoch": 2.617606192933704,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043189685950369963,
      "loss": 5.1144,
      "mean_token_accuracy": 0.1701927587389946,
      "num_tokens": 57788538.0,
      "step": 26375
    },
    {
      "entropy": 5.562472343444824,
      "epoch": 2.6181024215958715,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043187146165976577,
      "loss": 5.2633,
      "mean_token_accuracy": 0.16502532213926316,
      "num_tokens": 57799983.0,
      "step": 26380
    },
    {
      "entropy": 5.535589075088501,
      "epoch": 2.618598650258039,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004318460599257578,
      "loss": 5.0571,
      "mean_token_accuracy": 0.1680133193731308,
      "num_tokens": 57810195.0,
      "step": 26385
    },
    {
      "entropy": 5.441228151321411,
      "epoch": 2.6190948789202064,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043182065430230545,
      "loss": 5.0429,
      "mean_token_accuracy": 0.1727302059531212,
      "num_tokens": 57821930.0,
      "step": 26390
    },
    {
      "entropy": 5.473917007446289,
      "epoch": 2.619591107582374,
      "grad_norm": 1.125,
      "learning_rate": 0.0004317952447900388,
      "loss": 5.0881,
      "mean_token_accuracy": 0.16784320324659346,
      "num_tokens": 57834272.0,
      "step": 26395
    },
    {
      "entropy": 5.5083959102630615,
      "epoch": 2.6200873362445414,
      "grad_norm": 1.59375,
      "learning_rate": 0.000431769831389588,
      "loss": 5.1268,
      "mean_token_accuracy": 0.16982325315475463,
      "num_tokens": 57845491.0,
      "step": 26400
    },
    {
      "entropy": 5.506555700302124,
      "epoch": 2.620583564906709,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004317444141015831,
      "loss": 5.1922,
      "mean_token_accuracy": 0.1683211237192154,
      "num_tokens": 57857816.0,
      "step": 26405
    },
    {
      "entropy": 5.469267177581787,
      "epoch": 2.6210797935688763,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043171899292665457,
      "loss": 5.1345,
      "mean_token_accuracy": 0.16509229689836502,
      "num_tokens": 57867769.0,
      "step": 26410
    },
    {
      "entropy": 5.5625128746032715,
      "epoch": 2.621576022231044,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004316935678654327,
      "loss": 5.2357,
      "mean_token_accuracy": 0.16157547533512115,
      "num_tokens": 57878027.0,
      "step": 26415
    },
    {
      "entropy": 5.485332059860229,
      "epoch": 2.6220722508932117,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043166813891854796,
      "loss": 5.1049,
      "mean_token_accuracy": 0.1732478365302086,
      "num_tokens": 57887938.0,
      "step": 26420
    },
    {
      "entropy": 5.510219097137451,
      "epoch": 2.622568479555379,
      "grad_norm": 1.03125,
      "learning_rate": 0.000431642706086631,
      "loss": 5.1179,
      "mean_token_accuracy": 0.17450241297483443,
      "num_tokens": 57899375.0,
      "step": 26425
    },
    {
      "entropy": 5.479735326766968,
      "epoch": 2.6230647082175467,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043161726937031243,
      "loss": 5.157,
      "mean_token_accuracy": 0.16948518753051758,
      "num_tokens": 57910393.0,
      "step": 26430
    },
    {
      "entropy": 5.4575690746307375,
      "epoch": 2.623560936879714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000431591828770223,
      "loss": 5.1734,
      "mean_token_accuracy": 0.1692090779542923,
      "num_tokens": 57922236.0,
      "step": 26435
    },
    {
      "entropy": 5.462493753433227,
      "epoch": 2.6240571655418816,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004315663842869938,
      "loss": 5.1214,
      "mean_token_accuracy": 0.17694670855998992,
      "num_tokens": 57931804.0,
      "step": 26440
    },
    {
      "entropy": 5.535282325744629,
      "epoch": 2.624553394204049,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043154093592125554,
      "loss": 5.1488,
      "mean_token_accuracy": 0.16140951514244078,
      "num_tokens": 57943259.0,
      "step": 26445
    },
    {
      "entropy": 5.416129493713379,
      "epoch": 2.625049622866217,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043151548367363943,
      "loss": 5.0199,
      "mean_token_accuracy": 0.17560317516326904,
      "num_tokens": 57952945.0,
      "step": 26450
    },
    {
      "entropy": 5.444956207275391,
      "epoch": 2.6255458515283845,
      "grad_norm": 1.40625,
      "learning_rate": 0.00043149002754477655,
      "loss": 5.1354,
      "mean_token_accuracy": 0.1665095493197441,
      "num_tokens": 57964030.0,
      "step": 26455
    },
    {
      "entropy": 5.464812564849853,
      "epoch": 2.626042080190552,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004314645675352982,
      "loss": 5.1059,
      "mean_token_accuracy": 0.17425035387277604,
      "num_tokens": 57974888.0,
      "step": 26460
    },
    {
      "entropy": 5.531467390060425,
      "epoch": 2.6265383088527194,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004314391036458357,
      "loss": 5.1449,
      "mean_token_accuracy": 0.16563370823860168,
      "num_tokens": 57984922.0,
      "step": 26465
    },
    {
      "entropy": 5.438196039199829,
      "epoch": 2.627034537514887,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043141363587702054,
      "loss": 5.0995,
      "mean_token_accuracy": 0.17055424600839614,
      "num_tokens": 57994881.0,
      "step": 26470
    },
    {
      "entropy": 5.522970914840698,
      "epoch": 2.6275307661770544,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004313881642294843,
      "loss": 5.178,
      "mean_token_accuracy": 0.17288727611303328,
      "num_tokens": 58005335.0,
      "step": 26475
    },
    {
      "entropy": 5.4896985530853275,
      "epoch": 2.628026994839222,
      "grad_norm": 1.0,
      "learning_rate": 0.0004313626887038585,
      "loss": 5.0719,
      "mean_token_accuracy": 0.176258385181427,
      "num_tokens": 58016692.0,
      "step": 26480
    },
    {
      "entropy": 5.494421291351318,
      "epoch": 2.6285232235013893,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000431337209300775,
      "loss": 5.1711,
      "mean_token_accuracy": 0.1714398443698883,
      "num_tokens": 58028282.0,
      "step": 26485
    },
    {
      "entropy": 5.435199546813965,
      "epoch": 2.629019452163557,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004313117260208656,
      "loss": 5.1189,
      "mean_token_accuracy": 0.17359686344861985,
      "num_tokens": 58040108.0,
      "step": 26490
    },
    {
      "entropy": 5.457900667190552,
      "epoch": 2.6295156808257243,
      "grad_norm": 1.125,
      "learning_rate": 0.0004312862388647622,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17333920001983644,
      "num_tokens": 58050156.0,
      "step": 26495
    },
    {
      "entropy": 5.449525499343872,
      "epoch": 2.6300119094878918,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043126074783309687,
      "loss": 5.1047,
      "mean_token_accuracy": 0.17071642428636552,
      "num_tokens": 58060929.0,
      "step": 26500
    },
    {
      "entropy": 5.492337560653686,
      "epoch": 2.6305081381500597,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004312352529265016,
      "loss": 5.1632,
      "mean_token_accuracy": 0.16524887084960938,
      "num_tokens": 58071684.0,
      "step": 26505
    },
    {
      "entropy": 5.550176525115967,
      "epoch": 2.631004366812227,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043120975414560886,
      "loss": 5.1903,
      "mean_token_accuracy": 0.1681920811533928,
      "num_tokens": 58082730.0,
      "step": 26510
    },
    {
      "entropy": 5.560279035568238,
      "epoch": 2.6315005954743946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043118425149105073,
      "loss": 5.2457,
      "mean_token_accuracy": 0.16601796448230743,
      "num_tokens": 58095278.0,
      "step": 26515
    },
    {
      "entropy": 5.601905012130738,
      "epoch": 2.631996824136562,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004311587449634597,
      "loss": 5.3113,
      "mean_token_accuracy": 0.16078627705574036,
      "num_tokens": 58107816.0,
      "step": 26520
    },
    {
      "entropy": 5.544784593582153,
      "epoch": 2.6324930527987296,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004311332345634683,
      "loss": 5.1942,
      "mean_token_accuracy": 0.16194585114717483,
      "num_tokens": 58117779.0,
      "step": 26525
    },
    {
      "entropy": 5.488578128814697,
      "epoch": 2.632989281460897,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043110772029170916,
      "loss": 5.1294,
      "mean_token_accuracy": 0.17485171854496,
      "num_tokens": 58128665.0,
      "step": 26530
    },
    {
      "entropy": 5.5546956062316895,
      "epoch": 2.633485510123065,
      "grad_norm": 1.125,
      "learning_rate": 0.00043108220214881494,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1682320922613144,
      "num_tokens": 58139658.0,
      "step": 26535
    },
    {
      "entropy": 5.569842910766601,
      "epoch": 2.6339817387852325,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004310566801354184,
      "loss": 5.1767,
      "mean_token_accuracy": 0.16830241680145264,
      "num_tokens": 58149980.0,
      "step": 26540
    },
    {
      "entropy": 5.448394536972046,
      "epoch": 2.6344779674474,
      "grad_norm": 1.125,
      "learning_rate": 0.00043103115425215246,
      "loss": 5.0402,
      "mean_token_accuracy": 0.1731789827346802,
      "num_tokens": 58159809.0,
      "step": 26545
    },
    {
      "entropy": 5.425151252746582,
      "epoch": 2.6349741961095674,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043100562449965016,
      "loss": 5.1275,
      "mean_token_accuracy": 0.1698968380689621,
      "num_tokens": 58169912.0,
      "step": 26550
    },
    {
      "entropy": 5.5288715839385985,
      "epoch": 2.635470424771735,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043098009087854446,
      "loss": 5.1813,
      "mean_token_accuracy": 0.16419783979654312,
      "num_tokens": 58179949.0,
      "step": 26555
    },
    {
      "entropy": 5.484056425094605,
      "epoch": 2.6359666534339024,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043095455338946863,
      "loss": 5.1295,
      "mean_token_accuracy": 0.17258311808109283,
      "num_tokens": 58189847.0,
      "step": 26560
    },
    {
      "entropy": 5.4325727939605715,
      "epoch": 2.63646288209607,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000430929012033056,
      "loss": 5.0369,
      "mean_token_accuracy": 0.17823905497789383,
      "num_tokens": 58201514.0,
      "step": 26565
    },
    {
      "entropy": 5.474339866638184,
      "epoch": 2.6369591107582373,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043090346680993983,
      "loss": 5.1135,
      "mean_token_accuracy": 0.18136129826307296,
      "num_tokens": 58211940.0,
      "step": 26570
    },
    {
      "entropy": 5.50669264793396,
      "epoch": 2.6374553394204048,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004308779177207536,
      "loss": 5.1545,
      "mean_token_accuracy": 0.16937684863805771,
      "num_tokens": 58222260.0,
      "step": 26575
    },
    {
      "entropy": 5.428654575347901,
      "epoch": 2.6379515680825723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004308523647661309,
      "loss": 5.0921,
      "mean_token_accuracy": 0.17311939895153045,
      "num_tokens": 58232649.0,
      "step": 26580
    },
    {
      "entropy": 5.442238283157349,
      "epoch": 2.6384477967447397,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043082680794670535,
      "loss": 5.0424,
      "mean_token_accuracy": 0.17477950006723403,
      "num_tokens": 58243373.0,
      "step": 26585
    },
    {
      "entropy": 5.382277631759644,
      "epoch": 2.6389440254069076,
      "grad_norm": 1.125,
      "learning_rate": 0.00043080124726311084,
      "loss": 5.1163,
      "mean_token_accuracy": 0.17402798384428025,
      "num_tokens": 58253863.0,
      "step": 26590
    },
    {
      "entropy": 5.465031290054322,
      "epoch": 2.639440254069075,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004307756827159811,
      "loss": 5.2046,
      "mean_token_accuracy": 0.17224040925502776,
      "num_tokens": 58265326.0,
      "step": 26595
    },
    {
      "entropy": 5.539064550399781,
      "epoch": 2.6399364827312426,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043075011430594994,
      "loss": 5.1782,
      "mean_token_accuracy": 0.1703191339969635,
      "num_tokens": 58276006.0,
      "step": 26600
    },
    {
      "entropy": 5.444818353652954,
      "epoch": 2.64043271139341,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043072454203365175,
      "loss": 4.9998,
      "mean_token_accuracy": 0.1771932065486908,
      "num_tokens": 58285932.0,
      "step": 26605
    },
    {
      "entropy": 5.4383433818817135,
      "epoch": 2.6409289400555775,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004306989658997204,
      "loss": 5.1049,
      "mean_token_accuracy": 0.17126367837190629,
      "num_tokens": 58297174.0,
      "step": 26610
    },
    {
      "entropy": 5.578279876708985,
      "epoch": 2.641425168717745,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043067338590479023,
      "loss": 5.2375,
      "mean_token_accuracy": 0.1604462295770645,
      "num_tokens": 58308509.0,
      "step": 26615
    },
    {
      "entropy": 5.447067832946777,
      "epoch": 2.641921397379913,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043064780204949543,
      "loss": 5.1447,
      "mean_token_accuracy": 0.16324626356363298,
      "num_tokens": 58318449.0,
      "step": 26620
    },
    {
      "entropy": 5.533356857299805,
      "epoch": 2.6424176260420804,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004306222143344706,
      "loss": 5.1234,
      "mean_token_accuracy": 0.17399397492408752,
      "num_tokens": 58329759.0,
      "step": 26625
    },
    {
      "entropy": 5.443163633346558,
      "epoch": 2.642913854704248,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004305966227603503,
      "loss": 5.0686,
      "mean_token_accuracy": 0.16978068202733992,
      "num_tokens": 58339363.0,
      "step": 26630
    },
    {
      "entropy": 5.412320995330811,
      "epoch": 2.6434100833664154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004305710273277689,
      "loss": 5.1463,
      "mean_token_accuracy": 0.1627653032541275,
      "num_tokens": 58349482.0,
      "step": 26635
    },
    {
      "entropy": 5.5105040073394775,
      "epoch": 2.643906312028583,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043054542803736126,
      "loss": 5.163,
      "mean_token_accuracy": 0.1686481460928917,
      "num_tokens": 58359879.0,
      "step": 26640
    },
    {
      "entropy": 5.544681406021118,
      "epoch": 2.6444025406907503,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043051982488976216,
      "loss": 5.1914,
      "mean_token_accuracy": 0.16446348279714584,
      "num_tokens": 58370475.0,
      "step": 26645
    },
    {
      "entropy": 5.462798357009888,
      "epoch": 2.644898769352918,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004304942178856066,
      "loss": 5.1369,
      "mean_token_accuracy": 0.1710032492876053,
      "num_tokens": 58381992.0,
      "step": 26650
    },
    {
      "entropy": 5.4790326118469235,
      "epoch": 2.6453949980150853,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043046860702552944,
      "loss": 5.0844,
      "mean_token_accuracy": 0.17764362096786498,
      "num_tokens": 58391616.0,
      "step": 26655
    },
    {
      "entropy": 5.5482337951660154,
      "epoch": 2.6458912266772527,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004304429923101658,
      "loss": 5.0993,
      "mean_token_accuracy": 0.17287470698356627,
      "num_tokens": 58401800.0,
      "step": 26660
    },
    {
      "entropy": 5.448855686187744,
      "epoch": 2.64638745533942,
      "grad_norm": 1.125,
      "learning_rate": 0.000430417373740151,
      "loss": 5.1039,
      "mean_token_accuracy": 0.17122768610715866,
      "num_tokens": 58413376.0,
      "step": 26665
    },
    {
      "entropy": 5.37421498298645,
      "epoch": 2.6468836840015877,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004303917513161201,
      "loss": 5.0597,
      "mean_token_accuracy": 0.17721162587404252,
      "num_tokens": 58423772.0,
      "step": 26670
    },
    {
      "entropy": 5.500373697280883,
      "epoch": 2.6473799126637556,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004303661250387088,
      "loss": 5.19,
      "mean_token_accuracy": 0.1650784730911255,
      "num_tokens": 58434838.0,
      "step": 26675
    },
    {
      "entropy": 5.526183557510376,
      "epoch": 2.647876141325923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004303404949085522,
      "loss": 5.1904,
      "mean_token_accuracy": 0.1676359385251999,
      "num_tokens": 58445036.0,
      "step": 26680
    },
    {
      "entropy": 5.498500871658325,
      "epoch": 2.6483723699880906,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004303148609262861,
      "loss": 5.1281,
      "mean_token_accuracy": 0.16988089829683303,
      "num_tokens": 58455558.0,
      "step": 26685
    },
    {
      "entropy": 5.4902873039245605,
      "epoch": 2.648868598650258,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004302892230925462,
      "loss": 5.1524,
      "mean_token_accuracy": 0.16351435631513594,
      "num_tokens": 58467716.0,
      "step": 26690
    },
    {
      "entropy": 5.449524784088135,
      "epoch": 2.6493648273124255,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043026358140796806,
      "loss": 5.1421,
      "mean_token_accuracy": 0.16490322947502137,
      "num_tokens": 58479421.0,
      "step": 26695
    },
    {
      "entropy": 5.410109567642212,
      "epoch": 2.649861055974593,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004302379358731878,
      "loss": 5.1064,
      "mean_token_accuracy": 0.17606889456510544,
      "num_tokens": 58490115.0,
      "step": 26700
    },
    {
      "entropy": 5.472385740280151,
      "epoch": 2.6503572846367605,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004302122864888411,
      "loss": 5.0816,
      "mean_token_accuracy": 0.17311066687107085,
      "num_tokens": 58501711.0,
      "step": 26705
    },
    {
      "entropy": 5.53517336845398,
      "epoch": 2.6508535132989284,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004301866332555643,
      "loss": 5.1437,
      "mean_token_accuracy": 0.16907930225133896,
      "num_tokens": 58513441.0,
      "step": 26710
    },
    {
      "entropy": 5.4745746612548825,
      "epoch": 2.651349741961096,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043016097617399327,
      "loss": 5.0749,
      "mean_token_accuracy": 0.17473644763231277,
      "num_tokens": 58523940.0,
      "step": 26715
    },
    {
      "entropy": 5.4573017120361325,
      "epoch": 2.6518459706232633,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004301353152447644,
      "loss": 5.1741,
      "mean_token_accuracy": 0.16287598609924317,
      "num_tokens": 58536354.0,
      "step": 26720
    },
    {
      "entropy": 5.48531322479248,
      "epoch": 2.652342199285431,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043010965046851404,
      "loss": 5.184,
      "mean_token_accuracy": 0.16329047530889512,
      "num_tokens": 58547096.0,
      "step": 26725
    },
    {
      "entropy": 5.504085636138916,
      "epoch": 2.6528384279475983,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043008398184587847,
      "loss": 5.1625,
      "mean_token_accuracy": 0.16790218353271485,
      "num_tokens": 58557951.0,
      "step": 26730
    },
    {
      "entropy": 5.447340965270996,
      "epoch": 2.6533346566097658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043005830937749447,
      "loss": 5.0717,
      "mean_token_accuracy": 0.17455704510211945,
      "num_tokens": 58569384.0,
      "step": 26735
    },
    {
      "entropy": 5.456131267547607,
      "epoch": 2.6538308852719332,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004300326330639984,
      "loss": 5.1363,
      "mean_token_accuracy": 0.17071943134069442,
      "num_tokens": 58580308.0,
      "step": 26740
    },
    {
      "entropy": 5.4797220706939695,
      "epoch": 2.6543271139341007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004300069529060272,
      "loss": 5.152,
      "mean_token_accuracy": 0.16508927196264267,
      "num_tokens": 58591780.0,
      "step": 26745
    },
    {
      "entropy": 5.501236772537231,
      "epoch": 2.654823342596268,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042998126890421754,
      "loss": 5.1481,
      "mean_token_accuracy": 0.17671160995960236,
      "num_tokens": 58601477.0,
      "step": 26750
    },
    {
      "entropy": 5.465973854064941,
      "epoch": 2.6553195712584357,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042995558105920627,
      "loss": 5.1353,
      "mean_token_accuracy": 0.1719975709915161,
      "num_tokens": 58612531.0,
      "step": 26755
    },
    {
      "entropy": 5.476226186752319,
      "epoch": 2.6558157999206036,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042992988937163066,
      "loss": 5.1773,
      "mean_token_accuracy": 0.16618674844503403,
      "num_tokens": 58624951.0,
      "step": 26760
    },
    {
      "entropy": 5.4933818817138675,
      "epoch": 2.656312028582771,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004299041938421275,
      "loss": 5.2078,
      "mean_token_accuracy": 0.16570258587598802,
      "num_tokens": 58636484.0,
      "step": 26765
    },
    {
      "entropy": 5.520197629928589,
      "epoch": 2.6568082572449385,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042987849447133423,
      "loss": 5.1291,
      "mean_token_accuracy": 0.16675496697425843,
      "num_tokens": 58647386.0,
      "step": 26770
    },
    {
      "entropy": 5.54072322845459,
      "epoch": 2.657304485907106,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004298527912598879,
      "loss": 5.1942,
      "mean_token_accuracy": 0.16811267286539078,
      "num_tokens": 58658656.0,
      "step": 26775
    },
    {
      "entropy": 5.466477394104004,
      "epoch": 2.6578007145692735,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004298270842084261,
      "loss": 5.0887,
      "mean_token_accuracy": 0.17541789561510085,
      "num_tokens": 58669901.0,
      "step": 26780
    },
    {
      "entropy": 5.463758516311645,
      "epoch": 2.658296943231441,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042980137331758626,
      "loss": 5.2022,
      "mean_token_accuracy": 0.16156850457191468,
      "num_tokens": 58680849.0,
      "step": 26785
    },
    {
      "entropy": 5.483605527877808,
      "epoch": 2.6587931718936084,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042977565858800593,
      "loss": 5.0911,
      "mean_token_accuracy": 0.16975382417440416,
      "num_tokens": 58690642.0,
      "step": 26790
    },
    {
      "entropy": 5.437202405929566,
      "epoch": 2.6592894005557763,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042974994002032275,
      "loss": 5.0404,
      "mean_token_accuracy": 0.17726391702890396,
      "num_tokens": 58700480.0,
      "step": 26795
    },
    {
      "entropy": 5.47571496963501,
      "epoch": 2.659785629217944,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042972421761517456,
      "loss": 5.1817,
      "mean_token_accuracy": 0.16195070445537568,
      "num_tokens": 58712443.0,
      "step": 26800
    },
    {
      "entropy": 5.433439636230469,
      "epoch": 2.6602818578801113,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004296984913731991,
      "loss": 5.0738,
      "mean_token_accuracy": 0.17152167558670045,
      "num_tokens": 58721172.0,
      "step": 26805
    },
    {
      "entropy": 5.441042757034301,
      "epoch": 2.6607780865422788,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004296727612950344,
      "loss": 5.1197,
      "mean_token_accuracy": 0.1742386668920517,
      "num_tokens": 58731060.0,
      "step": 26810
    },
    {
      "entropy": 5.475546741485596,
      "epoch": 2.6612743152044462,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004296470273813186,
      "loss": 5.1341,
      "mean_token_accuracy": 0.17080398499965668,
      "num_tokens": 58742189.0,
      "step": 26815
    },
    {
      "entropy": 5.547427368164063,
      "epoch": 2.6617705438666137,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004296212896326896,
      "loss": 5.1508,
      "mean_token_accuracy": 0.1696161150932312,
      "num_tokens": 58753617.0,
      "step": 26820
    },
    {
      "entropy": 5.460513544082642,
      "epoch": 2.662266772528781,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042959554804978596,
      "loss": 5.1545,
      "mean_token_accuracy": 0.17432137131690978,
      "num_tokens": 58765299.0,
      "step": 26825
    },
    {
      "entropy": 5.51094422340393,
      "epoch": 2.6627630011909487,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042956980263324576,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17059443145990372,
      "num_tokens": 58775280.0,
      "step": 26830
    },
    {
      "entropy": 5.505864429473877,
      "epoch": 2.663259229853116,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004295440533837075,
      "loss": 5.0789,
      "mean_token_accuracy": 0.1719609871506691,
      "num_tokens": 58786078.0,
      "step": 26835
    },
    {
      "entropy": 5.484742546081543,
      "epoch": 2.6637554585152836,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042951830030180983,
      "loss": 5.155,
      "mean_token_accuracy": 0.16989914774894715,
      "num_tokens": 58796665.0,
      "step": 26840
    },
    {
      "entropy": 5.5118671417236325,
      "epoch": 2.6642516871774515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042949254338819123,
      "loss": 5.1798,
      "mean_token_accuracy": 0.16307040750980378,
      "num_tokens": 58808150.0,
      "step": 26845
    },
    {
      "entropy": 5.51967225074768,
      "epoch": 2.664747915839619,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042946678264349036,
      "loss": 5.102,
      "mean_token_accuracy": 0.17475544959306716,
      "num_tokens": 58818278.0,
      "step": 26850
    },
    {
      "entropy": 5.465302515029907,
      "epoch": 2.6652441445017865,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042944101806834627,
      "loss": 5.1407,
      "mean_token_accuracy": 0.16648390293121337,
      "num_tokens": 58828847.0,
      "step": 26855
    },
    {
      "entropy": 5.455130577087402,
      "epoch": 2.665740373163954,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004294152496633977,
      "loss": 5.1371,
      "mean_token_accuracy": 0.16984124928712846,
      "num_tokens": 58839698.0,
      "step": 26860
    },
    {
      "entropy": 5.459395837783814,
      "epoch": 2.6662366018261214,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042938947742928357,
      "loss": 5.0946,
      "mean_token_accuracy": 0.1721004515886307,
      "num_tokens": 58850794.0,
      "step": 26865
    },
    {
      "entropy": 5.477340316772461,
      "epoch": 2.666732830488289,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042936370136664314,
      "loss": 5.1784,
      "mean_token_accuracy": 0.16539075821638108,
      "num_tokens": 58861590.0,
      "step": 26870
    },
    {
      "entropy": 5.541335153579712,
      "epoch": 2.6672290591504564,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004293379214761156,
      "loss": 5.2032,
      "mean_token_accuracy": 0.16312645077705384,
      "num_tokens": 58873796.0,
      "step": 26875
    },
    {
      "entropy": 5.565216970443726,
      "epoch": 2.6677252878126243,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004293121377583401,
      "loss": 5.157,
      "mean_token_accuracy": 0.16927811950445176,
      "num_tokens": 58885296.0,
      "step": 26880
    },
    {
      "entropy": 5.477684593200683,
      "epoch": 2.6682215164747918,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004292863502139561,
      "loss": 5.2083,
      "mean_token_accuracy": 0.16650303453207016,
      "num_tokens": 58897764.0,
      "step": 26885
    },
    {
      "entropy": 5.495551538467407,
      "epoch": 2.6687177451369593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004292605588436031,
      "loss": 5.1563,
      "mean_token_accuracy": 0.1655075266957283,
      "num_tokens": 58908964.0,
      "step": 26890
    },
    {
      "entropy": 5.375704097747803,
      "epoch": 2.6692139737991267,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042923476364792063,
      "loss": 5.0005,
      "mean_token_accuracy": 0.17701967805624008,
      "num_tokens": 58919152.0,
      "step": 26895
    },
    {
      "entropy": 5.405006170272827,
      "epoch": 2.669710202461294,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004292089646275484,
      "loss": 5.1451,
      "mean_token_accuracy": 0.1691821649670601,
      "num_tokens": 58928849.0,
      "step": 26900
    },
    {
      "entropy": 5.570833444595337,
      "epoch": 2.6702064311234617,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004291831617831261,
      "loss": 5.279,
      "mean_token_accuracy": 0.15910194516181947,
      "num_tokens": 58941560.0,
      "step": 26905
    },
    {
      "entropy": 5.57725977897644,
      "epoch": 2.670702659785629,
      "grad_norm": 1.0,
      "learning_rate": 0.00042915735511529356,
      "loss": 5.2061,
      "mean_token_accuracy": 0.15795112252235413,
      "num_tokens": 58953365.0,
      "step": 26910
    },
    {
      "entropy": 5.503256511688233,
      "epoch": 2.6711988884477966,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004291315446246909,
      "loss": 5.1091,
      "mean_token_accuracy": 0.17053559571504592,
      "num_tokens": 58964602.0,
      "step": 26915
    },
    {
      "entropy": 5.50094633102417,
      "epoch": 2.671695117109964,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042910573031195804,
      "loss": 5.1297,
      "mean_token_accuracy": 0.17260756343603134,
      "num_tokens": 58976154.0,
      "step": 26920
    },
    {
      "entropy": 5.496927881240845,
      "epoch": 2.6721913457721316,
      "grad_norm": 1.125,
      "learning_rate": 0.0004290799121777351,
      "loss": 5.1683,
      "mean_token_accuracy": 0.16924070566892624,
      "num_tokens": 58985989.0,
      "step": 26925
    },
    {
      "entropy": 5.566266679763794,
      "epoch": 2.672687574434299,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004290540902226624,
      "loss": 5.2221,
      "mean_token_accuracy": 0.16049024164676667,
      "num_tokens": 58997080.0,
      "step": 26930
    },
    {
      "entropy": 5.4953052520751955,
      "epoch": 2.673183803096467,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004290282644473802,
      "loss": 5.1432,
      "mean_token_accuracy": 0.1714162528514862,
      "num_tokens": 59008683.0,
      "step": 26935
    },
    {
      "entropy": 5.4771912574768065,
      "epoch": 2.6736800317586344,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004290024348525289,
      "loss": 5.071,
      "mean_token_accuracy": 0.17648707777261735,
      "num_tokens": 59019019.0,
      "step": 26940
    },
    {
      "entropy": 5.474037122726441,
      "epoch": 2.674176260420802,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004289766014387492,
      "loss": 5.1284,
      "mean_token_accuracy": 0.171190981566906,
      "num_tokens": 59029706.0,
      "step": 26945
    },
    {
      "entropy": 5.490227365493775,
      "epoch": 2.6746724890829694,
      "grad_norm": 2.25,
      "learning_rate": 0.0004289507642066813,
      "loss": 5.0585,
      "mean_token_accuracy": 0.1748732477426529,
      "num_tokens": 59040208.0,
      "step": 26950
    },
    {
      "entropy": 5.426524066925049,
      "epoch": 2.675168717745137,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004289249231569664,
      "loss": 5.109,
      "mean_token_accuracy": 0.1795317590236664,
      "num_tokens": 59050909.0,
      "step": 26955
    },
    {
      "entropy": 5.3747485160827635,
      "epoch": 2.6756649464073043,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042889907829024506,
      "loss": 5.0702,
      "mean_token_accuracy": 0.17429367303848267,
      "num_tokens": 59059834.0,
      "step": 26960
    },
    {
      "entropy": 5.520849418640137,
      "epoch": 2.6761611750694723,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042887322960715813,
      "loss": 5.152,
      "mean_token_accuracy": 0.16602381616830825,
      "num_tokens": 59071348.0,
      "step": 26965
    },
    {
      "entropy": 5.453034210205078,
      "epoch": 2.6766574037316397,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004288473771083468,
      "loss": 5.1048,
      "mean_token_accuracy": 0.16939321607351304,
      "num_tokens": 59083053.0,
      "step": 26970
    },
    {
      "entropy": 5.502906274795532,
      "epoch": 2.677153632393807,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004288215207944519,
      "loss": 5.1757,
      "mean_token_accuracy": 0.16568866223096848,
      "num_tokens": 59093400.0,
      "step": 26975
    },
    {
      "entropy": 5.505021619796753,
      "epoch": 2.6776498610559747,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042879566066611474,
      "loss": 5.1507,
      "mean_token_accuracy": 0.1669463708996773,
      "num_tokens": 59103653.0,
      "step": 26980
    },
    {
      "entropy": 5.527143144607544,
      "epoch": 2.678146089718142,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004287697967239767,
      "loss": 5.1752,
      "mean_token_accuracy": 0.1681654006242752,
      "num_tokens": 59116660.0,
      "step": 26985
    },
    {
      "entropy": 5.435150671005249,
      "epoch": 2.6786423183803096,
      "grad_norm": 1.109375,
      "learning_rate": 0.000428743928968679,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1794743597507477,
      "num_tokens": 59127424.0,
      "step": 26990
    },
    {
      "entropy": 5.466166973114014,
      "epoch": 2.679138547042477,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042871805740086317,
      "loss": 5.1504,
      "mean_token_accuracy": 0.16948517113924028,
      "num_tokens": 59137566.0,
      "step": 26995
    },
    {
      "entropy": 5.398263263702392,
      "epoch": 2.6796347757046446,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004286921820211707,
      "loss": 5.0467,
      "mean_token_accuracy": 0.1779508650302887,
      "num_tokens": 59148404.0,
      "step": 27000
    },
    {
      "epoch": 2.6796347757046446,
      "eval_entropy": 5.265685914842384,
      "eval_loss": 5.273345947265625,
      "eval_mean_token_accuracy": 0.16860176409740973,
      "eval_num_tokens": 59148404.0,
      "eval_runtime": 26.7312,
      "eval_samples_per_second": 1259.54,
      "eval_steps_per_second": 157.457,
      "step": 27000
    },
    {
      "entropy": 5.399322938919068,
      "epoch": 2.680131004366812,
      "grad_norm": 1.25,
      "learning_rate": 0.00042866630283024327,
      "loss": 5.1311,
      "mean_token_accuracy": 0.1688048928976059,
      "num_tokens": 59158097.0,
      "step": 27005
    },
    {
      "entropy": 5.42253999710083,
      "epoch": 2.6806272330289795,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004286404198287227,
      "loss": 4.9786,
      "mean_token_accuracy": 0.18141616880893707,
      "num_tokens": 59168940.0,
      "step": 27010
    },
    {
      "entropy": 5.474377298355103,
      "epoch": 2.681123461691147,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004286145330172508,
      "loss": 5.1224,
      "mean_token_accuracy": 0.1681310459971428,
      "num_tokens": 59180231.0,
      "step": 27015
    },
    {
      "entropy": 5.533919429779052,
      "epoch": 2.681619690353315,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004285886423964695,
      "loss": 5.2158,
      "mean_token_accuracy": 0.16379563957452775,
      "num_tokens": 59191565.0,
      "step": 27020
    },
    {
      "entropy": 5.477537679672241,
      "epoch": 2.6821159190154824,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004285627479670208,
      "loss": 5.2115,
      "mean_token_accuracy": 0.16681006401777268,
      "num_tokens": 59203394.0,
      "step": 27025
    },
    {
      "entropy": 5.373804569244385,
      "epoch": 2.68261214767765,
      "grad_norm": 1.25,
      "learning_rate": 0.0004285368497295468,
      "loss": 5.0824,
      "mean_token_accuracy": 0.1675624892115593,
      "num_tokens": 59213987.0,
      "step": 27030
    },
    {
      "entropy": 5.420059728622436,
      "epoch": 2.6831083763398174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004285109476846898,
      "loss": 5.0926,
      "mean_token_accuracy": 0.17746591866016387,
      "num_tokens": 59225825.0,
      "step": 27035
    },
    {
      "entropy": 5.395698976516724,
      "epoch": 2.683604605001985,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004284850418330921,
      "loss": 5.0756,
      "mean_token_accuracy": 0.17611055225133895,
      "num_tokens": 59235978.0,
      "step": 27040
    },
    {
      "entropy": 5.527285623550415,
      "epoch": 2.6841008336641523,
      "grad_norm": 1.3125,
      "learning_rate": 0.00042845913217539613,
      "loss": 5.2739,
      "mean_token_accuracy": 0.16267528086900712,
      "num_tokens": 59247505.0,
      "step": 27045
    },
    {
      "entropy": 5.440841579437256,
      "epoch": 2.6845970623263202,
      "grad_norm": 1.609375,
      "learning_rate": 0.00042843321871224426,
      "loss": 5.0576,
      "mean_token_accuracy": 0.17265614718198777,
      "num_tokens": 59258401.0,
      "step": 27050
    },
    {
      "entropy": 5.503900241851807,
      "epoch": 2.6850932909884877,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042840730144427927,
      "loss": 5.1133,
      "mean_token_accuracy": 0.16810685098171235,
      "num_tokens": 59268068.0,
      "step": 27055
    },
    {
      "entropy": 5.472047138214111,
      "epoch": 2.685589519650655,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042838138037214375,
      "loss": 5.1226,
      "mean_token_accuracy": 0.17068587392568588,
      "num_tokens": 59279567.0,
      "step": 27060
    },
    {
      "entropy": 5.45266900062561,
      "epoch": 2.6860857483128227,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042835545549648054,
      "loss": 5.051,
      "mean_token_accuracy": 0.173106849193573,
      "num_tokens": 59289862.0,
      "step": 27065
    },
    {
      "entropy": 5.45441312789917,
      "epoch": 2.68658197697499,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042832952681793235,
      "loss": 5.0788,
      "mean_token_accuracy": 0.17563043534755707,
      "num_tokens": 59301304.0,
      "step": 27070
    },
    {
      "entropy": 5.422416019439697,
      "epoch": 2.6870782056371576,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042830359433714243,
      "loss": 5.1307,
      "mean_token_accuracy": 0.17510637044906616,
      "num_tokens": 59312184.0,
      "step": 27075
    },
    {
      "entropy": 5.414707946777344,
      "epoch": 2.687574434299325,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042827765805475367,
      "loss": 5.0047,
      "mean_token_accuracy": 0.1769097551703453,
      "num_tokens": 59322688.0,
      "step": 27080
    },
    {
      "entropy": 5.411225128173828,
      "epoch": 2.6880706629614926,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042825171797140926,
      "loss": 5.0692,
      "mean_token_accuracy": 0.17453741729259492,
      "num_tokens": 59333349.0,
      "step": 27085
    },
    {
      "entropy": 5.446835470199585,
      "epoch": 2.68856689162366,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004282257740877525,
      "loss": 5.0653,
      "mean_token_accuracy": 0.17896694540977479,
      "num_tokens": 59344473.0,
      "step": 27090
    },
    {
      "entropy": 5.476280975341797,
      "epoch": 2.6890631202858275,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004281998264044267,
      "loss": 5.1713,
      "mean_token_accuracy": 0.16804344803094864,
      "num_tokens": 59355993.0,
      "step": 27095
    },
    {
      "entropy": 5.537916421890259,
      "epoch": 2.689559348947995,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004281738749220753,
      "loss": 5.193,
      "mean_token_accuracy": 0.1690578356385231,
      "num_tokens": 59367182.0,
      "step": 27100
    },
    {
      "entropy": 5.592710161209107,
      "epoch": 2.690055577610163,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042814791964134184,
      "loss": 5.1683,
      "mean_token_accuracy": 0.16527152359485625,
      "num_tokens": 59379235.0,
      "step": 27105
    },
    {
      "entropy": 5.452547550201416,
      "epoch": 2.6905518062723304,
      "grad_norm": 1.125,
      "learning_rate": 0.00042812196056287,
      "loss": 5.0872,
      "mean_token_accuracy": 0.17579531967639922,
      "num_tokens": 59388514.0,
      "step": 27110
    },
    {
      "entropy": 5.42165470123291,
      "epoch": 2.691048034934498,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042809599768730357,
      "loss": 5.1408,
      "mean_token_accuracy": 0.16910533607006073,
      "num_tokens": 59399496.0,
      "step": 27115
    },
    {
      "entropy": 5.474053239822387,
      "epoch": 2.6915442635966653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004280700310152861,
      "loss": 5.1056,
      "mean_token_accuracy": 0.1760031908750534,
      "num_tokens": 59411986.0,
      "step": 27120
    },
    {
      "entropy": 5.561061716079712,
      "epoch": 2.692040492258833,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004280440605474619,
      "loss": 5.3101,
      "mean_token_accuracy": 0.16042536944150926,
      "num_tokens": 59424232.0,
      "step": 27125
    },
    {
      "entropy": 5.557160663604736,
      "epoch": 2.6925367209210003,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004280180862844747,
      "loss": 5.185,
      "mean_token_accuracy": 0.1647045150399208,
      "num_tokens": 59435201.0,
      "step": 27130
    },
    {
      "entropy": 5.51388955116272,
      "epoch": 2.693032949583168,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042799210822696865,
      "loss": 5.197,
      "mean_token_accuracy": 0.16925347745418548,
      "num_tokens": 59445552.0,
      "step": 27135
    },
    {
      "entropy": 5.538771772384644,
      "epoch": 2.6935291782453357,
      "grad_norm": 1.140625,
      "learning_rate": 0.000427966126375588,
      "loss": 5.1831,
      "mean_token_accuracy": 0.15953448861837388,
      "num_tokens": 59455634.0,
      "step": 27140
    },
    {
      "entropy": 5.5451366901397705,
      "epoch": 2.694025406907503,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004279401407309771,
      "loss": 5.161,
      "mean_token_accuracy": 0.16927441209554672,
      "num_tokens": 59467955.0,
      "step": 27145
    },
    {
      "entropy": 5.537386512756347,
      "epoch": 2.6945216355696706,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004279141512937802,
      "loss": 5.1462,
      "mean_token_accuracy": 0.16949107348918915,
      "num_tokens": 59479566.0,
      "step": 27150
    },
    {
      "entropy": 5.386728763580322,
      "epoch": 2.695017864231838,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004278881580646418,
      "loss": 5.0911,
      "mean_token_accuracy": 0.17495038062334062,
      "num_tokens": 59489967.0,
      "step": 27155
    },
    {
      "entropy": 5.445855712890625,
      "epoch": 2.6955140928940056,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004278621610442066,
      "loss": 5.1348,
      "mean_token_accuracy": 0.16806379854679107,
      "num_tokens": 59501640.0,
      "step": 27160
    },
    {
      "entropy": 5.464929437637329,
      "epoch": 2.696010321556173,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004278361602331192,
      "loss": 5.0939,
      "mean_token_accuracy": 0.17651900202035903,
      "num_tokens": 59512767.0,
      "step": 27165
    },
    {
      "entropy": 5.446403551101684,
      "epoch": 2.6965065502183405,
      "grad_norm": 1.125,
      "learning_rate": 0.00042781015563202434,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1772410273551941,
      "num_tokens": 59522188.0,
      "step": 27170
    },
    {
      "entropy": 5.426087045669556,
      "epoch": 2.697002778880508,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004277841472415669,
      "loss": 5.1043,
      "mean_token_accuracy": 0.17350180298089982,
      "num_tokens": 59534871.0,
      "step": 27175
    },
    {
      "entropy": 5.579172229766845,
      "epoch": 2.6974990075426755,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004277581350623918,
      "loss": 5.2232,
      "mean_token_accuracy": 0.15857364237308502,
      "num_tokens": 59546858.0,
      "step": 27180
    },
    {
      "entropy": 5.44722580909729,
      "epoch": 2.697995236204843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004277321190951441,
      "loss": 5.0524,
      "mean_token_accuracy": 0.16923653483390808,
      "num_tokens": 59557022.0,
      "step": 27185
    },
    {
      "entropy": 5.397711944580078,
      "epoch": 2.698491464867011,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042770609934046904,
      "loss": 5.09,
      "mean_token_accuracy": 0.17626593112945557,
      "num_tokens": 59567564.0,
      "step": 27190
    },
    {
      "entropy": 5.523210906982422,
      "epoch": 2.6989876935291783,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042768007579901176,
      "loss": 5.2297,
      "mean_token_accuracy": 0.15777265429496765,
      "num_tokens": 59579122.0,
      "step": 27195
    },
    {
      "entropy": 5.595355463027954,
      "epoch": 2.699483922191346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042765404847141755,
      "loss": 5.2134,
      "mean_token_accuracy": 0.17067782580852509,
      "num_tokens": 59590477.0,
      "step": 27200
    },
    {
      "entropy": 5.4955121040344235,
      "epoch": 2.6999801508535133,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004276280173583319,
      "loss": 5.1032,
      "mean_token_accuracy": 0.17185094207525253,
      "num_tokens": 59601217.0,
      "step": 27205
    },
    {
      "entropy": 5.411932134628296,
      "epoch": 2.7004763795156808,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042760198246040025,
      "loss": 5.1069,
      "mean_token_accuracy": 0.1740837037563324,
      "num_tokens": 59612271.0,
      "step": 27210
    },
    {
      "entropy": 5.506404447555542,
      "epoch": 2.7009726081778482,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004275759437782683,
      "loss": 5.1955,
      "mean_token_accuracy": 0.16518528014421463,
      "num_tokens": 59623948.0,
      "step": 27215
    },
    {
      "entropy": 5.481752872467041,
      "epoch": 2.701468836840016,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042754990131258175,
      "loss": 5.0518,
      "mean_token_accuracy": 0.17356666326522827,
      "num_tokens": 59635720.0,
      "step": 27220
    },
    {
      "entropy": 5.497309970855713,
      "epoch": 2.7019650655021836,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004275238550639863,
      "loss": 5.1453,
      "mean_token_accuracy": 0.17210535407066346,
      "num_tokens": 59645171.0,
      "step": 27225
    },
    {
      "entropy": 5.522292566299439,
      "epoch": 2.702461294164351,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000427497805033128,
      "loss": 5.2233,
      "mean_token_accuracy": 0.16382930278778077,
      "num_tokens": 59655266.0,
      "step": 27230
    },
    {
      "entropy": 5.45999550819397,
      "epoch": 2.7029575228265186,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004274717512206527,
      "loss": 5.1222,
      "mean_token_accuracy": 0.16849755644798278,
      "num_tokens": 59667055.0,
      "step": 27235
    },
    {
      "entropy": 5.465108251571655,
      "epoch": 2.703453751488686,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004274456936272065,
      "loss": 5.0843,
      "mean_token_accuracy": 0.17385688424110413,
      "num_tokens": 59678621.0,
      "step": 27240
    },
    {
      "entropy": 5.450790357589722,
      "epoch": 2.7039499801508535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042741963225343554,
      "loss": 5.101,
      "mean_token_accuracy": 0.17740917652845384,
      "num_tokens": 59690514.0,
      "step": 27245
    },
    {
      "entropy": 5.5248634815216064,
      "epoch": 2.704446208813021,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004273935670999861,
      "loss": 5.151,
      "mean_token_accuracy": 0.1650402620434761,
      "num_tokens": 59701271.0,
      "step": 27250
    },
    {
      "entropy": 5.499176263809204,
      "epoch": 2.7049424374751885,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042736749816750476,
      "loss": 5.2163,
      "mean_token_accuracy": 0.15958122313022613,
      "num_tokens": 59712787.0,
      "step": 27255
    },
    {
      "entropy": 5.448222064971924,
      "epoch": 2.705438666137356,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004273414254566376,
      "loss": 5.0434,
      "mean_token_accuracy": 0.1738167628645897,
      "num_tokens": 59722787.0,
      "step": 27260
    },
    {
      "entropy": 5.491036796569825,
      "epoch": 2.7059348947995234,
      "grad_norm": 1.125,
      "learning_rate": 0.0004273153489680315,
      "loss": 5.1302,
      "mean_token_accuracy": 0.17609838992357255,
      "num_tokens": 59733302.0,
      "step": 27265
    },
    {
      "entropy": 5.443224239349365,
      "epoch": 2.706431123461691,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042728926870233287,
      "loss": 5.1202,
      "mean_token_accuracy": 0.16833443343639373,
      "num_tokens": 59743290.0,
      "step": 27270
    },
    {
      "entropy": 5.518179321289063,
      "epoch": 2.706927352123859,
      "grad_norm": 1.25,
      "learning_rate": 0.00042726318466018855,
      "loss": 5.1884,
      "mean_token_accuracy": 0.16715065091848375,
      "num_tokens": 59754145.0,
      "step": 27275
    },
    {
      "entropy": 5.522227191925049,
      "epoch": 2.7074235807860263,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004272370968422454,
      "loss": 5.1509,
      "mean_token_accuracy": 0.170865061879158,
      "num_tokens": 59764075.0,
      "step": 27280
    },
    {
      "entropy": 5.4730394840240475,
      "epoch": 2.7079198094481938,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042721100524915017,
      "loss": 5.1699,
      "mean_token_accuracy": 0.16712161600589753,
      "num_tokens": 59775495.0,
      "step": 27285
    },
    {
      "entropy": 5.4413238048553465,
      "epoch": 2.7084160381103612,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042718490988155014,
      "loss": 5.0364,
      "mean_token_accuracy": 0.19394786953926085,
      "num_tokens": 59787006.0,
      "step": 27290
    },
    {
      "entropy": 5.538814067840576,
      "epoch": 2.7089122667725287,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042715881074009216,
      "loss": 5.1426,
      "mean_token_accuracy": 0.16909931749105453,
      "num_tokens": 59797680.0,
      "step": 27295
    },
    {
      "entropy": 5.439060163497925,
      "epoch": 2.709408495434696,
      "grad_norm": 1.0,
      "learning_rate": 0.00042713270782542364,
      "loss": 5.1144,
      "mean_token_accuracy": 0.170664943754673,
      "num_tokens": 59809666.0,
      "step": 27300
    },
    {
      "entropy": 5.405431509017944,
      "epoch": 2.7099047240968637,
      "grad_norm": 1.25,
      "learning_rate": 0.0004271066011381917,
      "loss": 5.0866,
      "mean_token_accuracy": 0.17656137347221373,
      "num_tokens": 59819889.0,
      "step": 27305
    },
    {
      "entropy": 5.43539571762085,
      "epoch": 2.7104009527590316,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004270804906790438,
      "loss": 5.0378,
      "mean_token_accuracy": 0.1778980165719986,
      "num_tokens": 59830344.0,
      "step": 27310
    },
    {
      "entropy": 5.4810826778411865,
      "epoch": 2.710897181421199,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004270543764486275,
      "loss": 5.0894,
      "mean_token_accuracy": 0.17707936465740204,
      "num_tokens": 59843498.0,
      "step": 27315
    },
    {
      "entropy": 5.522764825820923,
      "epoch": 2.7113934100833665,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004270282584475903,
      "loss": 5.1474,
      "mean_token_accuracy": 0.16799701303243636,
      "num_tokens": 59854633.0,
      "step": 27320
    },
    {
      "entropy": 5.475446176528931,
      "epoch": 2.711889638745534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042700213667657983,
      "loss": 5.2154,
      "mean_token_accuracy": 0.16825246959924697,
      "num_tokens": 59865925.0,
      "step": 27325
    },
    {
      "entropy": 5.476767253875733,
      "epoch": 2.7123858674077015,
      "grad_norm": 1.15625,
      "learning_rate": 0.000426976011136244,
      "loss": 5.0937,
      "mean_token_accuracy": 0.1781882479786873,
      "num_tokens": 59877196.0,
      "step": 27330
    },
    {
      "entropy": 5.5053528308868405,
      "epoch": 2.712882096069869,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042694988182723047,
      "loss": 5.1611,
      "mean_token_accuracy": 0.16164173632860185,
      "num_tokens": 59887411.0,
      "step": 27335
    },
    {
      "entropy": 5.457582664489746,
      "epoch": 2.7133783247320364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042692374875018735,
      "loss": 5.1104,
      "mean_token_accuracy": 0.1793423429131508,
      "num_tokens": 59899123.0,
      "step": 27340
    },
    {
      "entropy": 5.43488826751709,
      "epoch": 2.713874553394204,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042689761190576263,
      "loss": 5.0667,
      "mean_token_accuracy": 0.17761276811361312,
      "num_tokens": 59910326.0,
      "step": 27345
    },
    {
      "entropy": 5.479538822174073,
      "epoch": 2.7143707820563714,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042687147129460447,
      "loss": 5.1734,
      "mean_token_accuracy": 0.1648683175444603,
      "num_tokens": 59920722.0,
      "step": 27350
    },
    {
      "entropy": 5.48493013381958,
      "epoch": 2.714867010718539,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042684532691736104,
      "loss": 5.1533,
      "mean_token_accuracy": 0.16580720394849777,
      "num_tokens": 59932568.0,
      "step": 27355
    },
    {
      "entropy": 5.470747470855713,
      "epoch": 2.715363239380707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042681917877468076,
      "loss": 5.1901,
      "mean_token_accuracy": 0.16223060637712478,
      "num_tokens": 59943805.0,
      "step": 27360
    },
    {
      "entropy": 5.477178287506104,
      "epoch": 2.7158594680428743,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042679302686721187,
      "loss": 5.0867,
      "mean_token_accuracy": 0.17851133346557618,
      "num_tokens": 59954455.0,
      "step": 27365
    },
    {
      "entropy": 5.460833024978638,
      "epoch": 2.7163556967050417,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004267668711956031,
      "loss": 5.084,
      "mean_token_accuracy": 0.1750555783510208,
      "num_tokens": 59966722.0,
      "step": 27370
    },
    {
      "entropy": 5.456364107131958,
      "epoch": 2.716851925367209,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042674071176050296,
      "loss": 5.1046,
      "mean_token_accuracy": 0.17216793745756148,
      "num_tokens": 59976702.0,
      "step": 27375
    },
    {
      "entropy": 5.432805919647217,
      "epoch": 2.7173481540293767,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042671454856256,
      "loss": 5.0726,
      "mean_token_accuracy": 0.17196227014064788,
      "num_tokens": 59987038.0,
      "step": 27380
    },
    {
      "entropy": 5.5090954303741455,
      "epoch": 2.717844382691544,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004266883816024233,
      "loss": 5.1419,
      "mean_token_accuracy": 0.170354962348938,
      "num_tokens": 59996755.0,
      "step": 27385
    },
    {
      "entropy": 5.469521427154541,
      "epoch": 2.7183406113537116,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004266622108807415,
      "loss": 5.0818,
      "mean_token_accuracy": 0.16695159375667573,
      "num_tokens": 60006419.0,
      "step": 27390
    },
    {
      "entropy": 5.5022862434387205,
      "epoch": 2.7188368400158796,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004266360363981637,
      "loss": 5.1952,
      "mean_token_accuracy": 0.1710411787033081,
      "num_tokens": 60017793.0,
      "step": 27395
    },
    {
      "entropy": 5.442492246627808,
      "epoch": 2.719333068678047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042660985815533905,
      "loss": 5.051,
      "mean_token_accuracy": 0.17312813103199004,
      "num_tokens": 60029392.0,
      "step": 27400
    },
    {
      "entropy": 5.490962934494019,
      "epoch": 2.7198292973402145,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042658367615291644,
      "loss": 5.1985,
      "mean_token_accuracy": 0.17197677940130235,
      "num_tokens": 60040467.0,
      "step": 27405
    },
    {
      "entropy": 5.452605438232422,
      "epoch": 2.720325526002382,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004265574903915454,
      "loss": 5.1232,
      "mean_token_accuracy": 0.16906119436025618,
      "num_tokens": 60051583.0,
      "step": 27410
    },
    {
      "entropy": 5.485458469390869,
      "epoch": 2.7208217546645495,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004265313008718751,
      "loss": 5.1281,
      "mean_token_accuracy": 0.1689688131213188,
      "num_tokens": 60062218.0,
      "step": 27415
    },
    {
      "entropy": 5.485013866424561,
      "epoch": 2.721317983326717,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042650510759455516,
      "loss": 5.1224,
      "mean_token_accuracy": 0.17654032111167908,
      "num_tokens": 60074455.0,
      "step": 27420
    },
    {
      "entropy": 5.478207588195801,
      "epoch": 2.7218142119888844,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042647891056023495,
      "loss": 5.1051,
      "mean_token_accuracy": 0.1675885111093521,
      "num_tokens": 60085442.0,
      "step": 27425
    },
    {
      "entropy": 5.4540478706359865,
      "epoch": 2.722310440651052,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042645270976956426,
      "loss": 5.1659,
      "mean_token_accuracy": 0.17346219569444657,
      "num_tokens": 60096441.0,
      "step": 27430
    },
    {
      "entropy": 5.509996461868286,
      "epoch": 2.7228066693132194,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042642650522319253,
      "loss": 5.1304,
      "mean_token_accuracy": 0.17440126836299896,
      "num_tokens": 60107399.0,
      "step": 27435
    },
    {
      "entropy": 5.508813428878784,
      "epoch": 2.723302897975387,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004264002969217699,
      "loss": 5.1004,
      "mean_token_accuracy": 0.17620668262243272,
      "num_tokens": 60117568.0,
      "step": 27440
    },
    {
      "entropy": 5.501161003112793,
      "epoch": 2.7237991266375547,
      "grad_norm": 1.125,
      "learning_rate": 0.0004263740848659461,
      "loss": 5.1618,
      "mean_token_accuracy": 0.17150973826646804,
      "num_tokens": 60127214.0,
      "step": 27445
    },
    {
      "entropy": 5.42465295791626,
      "epoch": 2.724295355299722,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042634786905637123,
      "loss": 5.0959,
      "mean_token_accuracy": 0.17285579442977905,
      "num_tokens": 60138167.0,
      "step": 27450
    },
    {
      "entropy": 5.4119157791137695,
      "epoch": 2.7247915839618897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004263216494936953,
      "loss": 5.0658,
      "mean_token_accuracy": 0.1722798228263855,
      "num_tokens": 60148841.0,
      "step": 27455
    },
    {
      "entropy": 5.448369264602661,
      "epoch": 2.725287812624057,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042629542617856855,
      "loss": 5.1673,
      "mean_token_accuracy": 0.17189225405454636,
      "num_tokens": 60159512.0,
      "step": 27460
    },
    {
      "entropy": 5.4380316734313965,
      "epoch": 2.7257840412862246,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004262691991116412,
      "loss": 5.1045,
      "mean_token_accuracy": 0.1675738975405693,
      "num_tokens": 60170463.0,
      "step": 27465
    },
    {
      "entropy": 5.361341190338135,
      "epoch": 2.726280269948392,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004262429682935637,
      "loss": 4.9785,
      "mean_token_accuracy": 0.17910560816526414,
      "num_tokens": 60180431.0,
      "step": 27470
    },
    {
      "entropy": 5.473160171508789,
      "epoch": 2.7267764986105596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004262167337249865,
      "loss": 5.13,
      "mean_token_accuracy": 0.1687505602836609,
      "num_tokens": 60190410.0,
      "step": 27475
    },
    {
      "entropy": 5.468656301498413,
      "epoch": 2.7272727272727275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004261904954065601,
      "loss": 5.0728,
      "mean_token_accuracy": 0.16982356309890748,
      "num_tokens": 60201909.0,
      "step": 27480
    },
    {
      "entropy": 5.526486110687256,
      "epoch": 2.727768955934895,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004261642533389352,
      "loss": 5.0772,
      "mean_token_accuracy": 0.16937623023986817,
      "num_tokens": 60212577.0,
      "step": 27485
    },
    {
      "entropy": 5.480685234069824,
      "epoch": 2.7282651845970625,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042613800752276264,
      "loss": 5.1684,
      "mean_token_accuracy": 0.16537186354398728,
      "num_tokens": 60222615.0,
      "step": 27490
    },
    {
      "entropy": 5.456792306900025,
      "epoch": 2.72876141325923,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042611175795869306,
      "loss": 5.1879,
      "mean_token_accuracy": 0.16926994323730468,
      "num_tokens": 60234338.0,
      "step": 27495
    },
    {
      "entropy": 5.521838569641114,
      "epoch": 2.7292576419213974,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004260855046473775,
      "loss": 5.0787,
      "mean_token_accuracy": 0.16985243409872056,
      "num_tokens": 60244781.0,
      "step": 27500
    },
    {
      "entropy": 5.492384719848633,
      "epoch": 2.729753870583565,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042605924758946706,
      "loss": 5.1552,
      "mean_token_accuracy": 0.1721781685948372,
      "num_tokens": 60255150.0,
      "step": 27505
    },
    {
      "entropy": 5.398989534378051,
      "epoch": 2.7302500992457324,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042603298678561277,
      "loss": 5.1062,
      "mean_token_accuracy": 0.17410674095153808,
      "num_tokens": 60266073.0,
      "step": 27510
    },
    {
      "entropy": 5.497126865386963,
      "epoch": 2.7307463279079,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004260067222364659,
      "loss": 5.1544,
      "mean_token_accuracy": 0.16680903881788253,
      "num_tokens": 60277520.0,
      "step": 27515
    },
    {
      "entropy": 5.516575813293457,
      "epoch": 2.7312425565700673,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042598045394267757,
      "loss": 5.1198,
      "mean_token_accuracy": 0.17188170105218886,
      "num_tokens": 60288530.0,
      "step": 27520
    },
    {
      "entropy": 5.471673679351807,
      "epoch": 2.731738785232235,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042595418190489944,
      "loss": 5.0669,
      "mean_token_accuracy": 0.18000012934207915,
      "num_tokens": 60299926.0,
      "step": 27525
    },
    {
      "entropy": 5.493292760848999,
      "epoch": 2.7322350138944023,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042592790612378277,
      "loss": 5.2357,
      "mean_token_accuracy": 0.16906825304031373,
      "num_tokens": 60312129.0,
      "step": 27530
    },
    {
      "entropy": 5.473834800720215,
      "epoch": 2.73273124255657,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042590162659997933,
      "loss": 5.131,
      "mean_token_accuracy": 0.16935312151908874,
      "num_tokens": 60322405.0,
      "step": 27535
    },
    {
      "entropy": 5.526683330535889,
      "epoch": 2.7332274712187377,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004258753433341407,
      "loss": 5.139,
      "mean_token_accuracy": 0.1682938516139984,
      "num_tokens": 60333112.0,
      "step": 27540
    },
    {
      "entropy": 5.49456262588501,
      "epoch": 2.733723699880905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042584905632691867,
      "loss": 5.1335,
      "mean_token_accuracy": 0.17038790434598922,
      "num_tokens": 60343500.0,
      "step": 27545
    },
    {
      "entropy": 5.451054191589355,
      "epoch": 2.7342199285430726,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004258227655789651,
      "loss": 5.1104,
      "mean_token_accuracy": 0.16986193656921386,
      "num_tokens": 60354711.0,
      "step": 27550
    },
    {
      "entropy": 5.476953840255737,
      "epoch": 2.73471615720524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004257964710909318,
      "loss": 5.1027,
      "mean_token_accuracy": 0.17312677949666977,
      "num_tokens": 60365064.0,
      "step": 27555
    },
    {
      "entropy": 5.4260790824890135,
      "epoch": 2.7352123858674076,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042577017286347116,
      "loss": 5.1507,
      "mean_token_accuracy": 0.16614237874746324,
      "num_tokens": 60376425.0,
      "step": 27560
    },
    {
      "entropy": 5.4818628311157225,
      "epoch": 2.7357086145295755,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042574387089723504,
      "loss": 5.0334,
      "mean_token_accuracy": 0.17790220081806182,
      "num_tokens": 60386745.0,
      "step": 27565
    },
    {
      "entropy": 5.466653156280517,
      "epoch": 2.736204843191743,
      "grad_norm": 1.4375,
      "learning_rate": 0.00042571756519287567,
      "loss": 5.0835,
      "mean_token_accuracy": 0.1709902599453926,
      "num_tokens": 60397038.0,
      "step": 27570
    },
    {
      "entropy": 5.389129781723023,
      "epoch": 2.7367010718539104,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042569125575104557,
      "loss": 5.0917,
      "mean_token_accuracy": 0.17283748984336852,
      "num_tokens": 60406955.0,
      "step": 27575
    },
    {
      "entropy": 5.488987970352173,
      "epoch": 2.737197300516078,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042566494257239695,
      "loss": 5.1113,
      "mean_token_accuracy": 0.16941978484392167,
      "num_tokens": 60418245.0,
      "step": 27580
    },
    {
      "entropy": 5.495974063873291,
      "epoch": 2.7376935291782454,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004256386256575824,
      "loss": 5.1267,
      "mean_token_accuracy": 0.16500394195318221,
      "num_tokens": 60429310.0,
      "step": 27585
    },
    {
      "entropy": 5.488760185241699,
      "epoch": 2.738189757840413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004256123050072546,
      "loss": 5.1687,
      "mean_token_accuracy": 0.17090535759925843,
      "num_tokens": 60441355.0,
      "step": 27590
    },
    {
      "entropy": 5.582341098785401,
      "epoch": 2.7386859865025803,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042558598062206616,
      "loss": 5.313,
      "mean_token_accuracy": 0.1599149614572525,
      "num_tokens": 60452853.0,
      "step": 27595
    },
    {
      "entropy": 5.526364135742187,
      "epoch": 2.739182215164748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042555965250266984,
      "loss": 5.1106,
      "mean_token_accuracy": 0.168956758081913,
      "num_tokens": 60462934.0,
      "step": 27600
    },
    {
      "entropy": 5.514018106460571,
      "epoch": 2.7396784438269153,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042553332064971866,
      "loss": 5.2081,
      "mean_token_accuracy": 0.163618303835392,
      "num_tokens": 60475575.0,
      "step": 27605
    },
    {
      "entropy": 5.36696343421936,
      "epoch": 2.7401746724890828,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042550698506386536,
      "loss": 4.9706,
      "mean_token_accuracy": 0.18319304138422013,
      "num_tokens": 60486495.0,
      "step": 27610
    },
    {
      "entropy": 5.427762794494629,
      "epoch": 2.7406709011512502,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042548064574576323,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17948891669511796,
      "num_tokens": 60496806.0,
      "step": 27615
    },
    {
      "entropy": 5.462457656860352,
      "epoch": 2.741167129813418,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004254543026960654,
      "loss": 5.1047,
      "mean_token_accuracy": 0.17192971855401992,
      "num_tokens": 60508905.0,
      "step": 27620
    },
    {
      "entropy": 5.368332862854004,
      "epoch": 2.7416633584755856,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004254279559154249,
      "loss": 5.0504,
      "mean_token_accuracy": 0.17058863937854768,
      "num_tokens": 60519603.0,
      "step": 27625
    },
    {
      "entropy": 5.457321977615356,
      "epoch": 2.742159587137753,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042540160540449536,
      "loss": 5.1374,
      "mean_token_accuracy": 0.16968114972114562,
      "num_tokens": 60530977.0,
      "step": 27630
    },
    {
      "entropy": 5.487758922576904,
      "epoch": 2.7426558157999206,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004253752511639301,
      "loss": 5.1507,
      "mean_token_accuracy": 0.16729033291339873,
      "num_tokens": 60541354.0,
      "step": 27635
    },
    {
      "entropy": 5.441304397583008,
      "epoch": 2.743152044462088,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004253488931943826,
      "loss": 5.114,
      "mean_token_accuracy": 0.175933738052845,
      "num_tokens": 60552531.0,
      "step": 27640
    },
    {
      "entropy": 5.404482221603393,
      "epoch": 2.7436482731242555,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042532253149650653,
      "loss": 5.0799,
      "mean_token_accuracy": 0.17382190376520157,
      "num_tokens": 60562552.0,
      "step": 27645
    },
    {
      "entropy": 5.492954874038697,
      "epoch": 2.7441445017864234,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004252961660709556,
      "loss": 5.1383,
      "mean_token_accuracy": 0.16861938834190368,
      "num_tokens": 60572849.0,
      "step": 27650
    },
    {
      "entropy": 5.459127426147461,
      "epoch": 2.744640730448591,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042526979691838353,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17106774002313613,
      "num_tokens": 60584095.0,
      "step": 27655
    },
    {
      "entropy": 5.467094755172729,
      "epoch": 2.7451369591107584,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042524342403944446,
      "loss": 5.1392,
      "mean_token_accuracy": 0.17051862925291061,
      "num_tokens": 60594353.0,
      "step": 27660
    },
    {
      "entropy": 5.482387733459473,
      "epoch": 2.745633187772926,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042521704743479213,
      "loss": 5.1317,
      "mean_token_accuracy": 0.16595661491155625,
      "num_tokens": 60604558.0,
      "step": 27665
    },
    {
      "entropy": 5.419806575775146,
      "epoch": 2.7461294164350933,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042519066710508064,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17323573678731918,
      "num_tokens": 60615670.0,
      "step": 27670
    },
    {
      "entropy": 5.377108335494995,
      "epoch": 2.746625645097261,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042516428305096435,
      "loss": 5.022,
      "mean_token_accuracy": 0.1766265571117401,
      "num_tokens": 60626452.0,
      "step": 27675
    },
    {
      "entropy": 5.476261997222901,
      "epoch": 2.7471218737594283,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004251378952730973,
      "loss": 5.1124,
      "mean_token_accuracy": 0.17529001086950302,
      "num_tokens": 60637122.0,
      "step": 27680
    },
    {
      "entropy": 5.495320510864258,
      "epoch": 2.7476181024215958,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004251115037721341,
      "loss": 5.0987,
      "mean_token_accuracy": 0.17677818089723588,
      "num_tokens": 60648835.0,
      "step": 27685
    },
    {
      "entropy": 5.447657108306885,
      "epoch": 2.7481143310837632,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042508510854872905,
      "loss": 5.1413,
      "mean_token_accuracy": 0.16832537651062013,
      "num_tokens": 60658537.0,
      "step": 27690
    },
    {
      "entropy": 5.5038402557373045,
      "epoch": 2.7486105597459307,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042505870960353665,
      "loss": 5.1704,
      "mean_token_accuracy": 0.1693202570080757,
      "num_tokens": 60670450.0,
      "step": 27695
    },
    {
      "entropy": 5.5589793682098385,
      "epoch": 2.749106788408098,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004250323069372116,
      "loss": 5.1227,
      "mean_token_accuracy": 0.17503261268138887,
      "num_tokens": 60680539.0,
      "step": 27700
    },
    {
      "entropy": 5.470003652572632,
      "epoch": 2.749603017070266,
      "grad_norm": 1.125,
      "learning_rate": 0.00042500590055040866,
      "loss": 5.2176,
      "mean_token_accuracy": 0.16523246616125106,
      "num_tokens": 60693106.0,
      "step": 27705
    },
    {
      "entropy": 5.547760772705078,
      "epoch": 2.7500992457324336,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042497949044378267,
      "loss": 5.1864,
      "mean_token_accuracy": 0.16476092785596846,
      "num_tokens": 60703509.0,
      "step": 27710
    },
    {
      "entropy": 5.497804403305054,
      "epoch": 2.750595474394601,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042495307661798837,
      "loss": 5.1565,
      "mean_token_accuracy": 0.168133644759655,
      "num_tokens": 60714113.0,
      "step": 27715
    },
    {
      "entropy": 5.538295316696167,
      "epoch": 2.7510917030567685,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042492665907368094,
      "loss": 5.2221,
      "mean_token_accuracy": 0.16531198918819429,
      "num_tokens": 60726826.0,
      "step": 27720
    },
    {
      "entropy": 5.490433120727539,
      "epoch": 2.751587931718936,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042490023781151546,
      "loss": 5.1269,
      "mean_token_accuracy": 0.16791064739227296,
      "num_tokens": 60738443.0,
      "step": 27725
    },
    {
      "entropy": 5.436273765563965,
      "epoch": 2.7520841603811035,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042487381283214706,
      "loss": 5.0816,
      "mean_token_accuracy": 0.17295931577682494,
      "num_tokens": 60748865.0,
      "step": 27730
    },
    {
      "entropy": 5.4342780113220215,
      "epoch": 2.7525803890432714,
      "grad_norm": 1.0,
      "learning_rate": 0.0004248473841362311,
      "loss": 5.1278,
      "mean_token_accuracy": 0.17487930804491042,
      "num_tokens": 60760001.0,
      "step": 27735
    },
    {
      "entropy": 5.558310508728027,
      "epoch": 2.753076617705439,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042482095172442286,
      "loss": 5.1891,
      "mean_token_accuracy": 0.1707563355565071,
      "num_tokens": 60772003.0,
      "step": 27740
    },
    {
      "entropy": 5.459653997421265,
      "epoch": 2.7535728463676064,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004247945155973778,
      "loss": 5.0564,
      "mean_token_accuracy": 0.17814452350139617,
      "num_tokens": 60783167.0,
      "step": 27745
    },
    {
      "entropy": 5.442328548431396,
      "epoch": 2.754069075029774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004247680757557516,
      "loss": 5.1553,
      "mean_token_accuracy": 0.16840662211179733,
      "num_tokens": 60793917.0,
      "step": 27750
    },
    {
      "entropy": 5.554823017120361,
      "epoch": 2.7545653036919413,
      "grad_norm": 1.125,
      "learning_rate": 0.00042474163220019987,
      "loss": 5.2444,
      "mean_token_accuracy": 0.1624252274632454,
      "num_tokens": 60805166.0,
      "step": 27755
    },
    {
      "entropy": 5.584030961990356,
      "epoch": 2.7550615323541088,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042471518493137836,
      "loss": 5.1661,
      "mean_token_accuracy": 0.17054525911808013,
      "num_tokens": 60816984.0,
      "step": 27760
    },
    {
      "entropy": 5.455572080612183,
      "epoch": 2.7555577610162763,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004246887339499427,
      "loss": 5.1002,
      "mean_token_accuracy": 0.17574168592691422,
      "num_tokens": 60829481.0,
      "step": 27765
    },
    {
      "entropy": 5.4625670433044435,
      "epoch": 2.7560539896784437,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042466227925654917,
      "loss": 5.0745,
      "mean_token_accuracy": 0.1734531581401825,
      "num_tokens": 60839376.0,
      "step": 27770
    },
    {
      "entropy": 5.430452966690064,
      "epoch": 2.756550218340611,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004246358208518535,
      "loss": 5.1267,
      "mean_token_accuracy": 0.17310923486948013,
      "num_tokens": 60850383.0,
      "step": 27775
    },
    {
      "entropy": 5.454214000701905,
      "epoch": 2.7570464470027787,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00042460935873651196,
      "loss": 5.0924,
      "mean_token_accuracy": 0.17178258001804353,
      "num_tokens": 60861249.0,
      "step": 27780
    },
    {
      "entropy": 5.51443133354187,
      "epoch": 2.757542675664946,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042458289291118066,
      "loss": 5.1353,
      "mean_token_accuracy": 0.17565696835517883,
      "num_tokens": 60872220.0,
      "step": 27785
    },
    {
      "entropy": 5.471701383590698,
      "epoch": 2.758038904327114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042455642337651595,
      "loss": 5.1215,
      "mean_token_accuracy": 0.16786434203386308,
      "num_tokens": 60882006.0,
      "step": 27790
    },
    {
      "entropy": 5.461632585525512,
      "epoch": 2.7585351329892815,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004245299501331743,
      "loss": 5.0409,
      "mean_token_accuracy": 0.18220600485801697,
      "num_tokens": 60891732.0,
      "step": 27795
    },
    {
      "entropy": 5.487818622589112,
      "epoch": 2.759031361651449,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042450347318181194,
      "loss": 5.1488,
      "mean_token_accuracy": 0.17227755188941957,
      "num_tokens": 60902990.0,
      "step": 27800
    },
    {
      "entropy": 5.479510116577148,
      "epoch": 2.7595275903136165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042447699252308566,
      "loss": 5.1161,
      "mean_token_accuracy": 0.16751719266176224,
      "num_tokens": 60914673.0,
      "step": 27805
    },
    {
      "entropy": 5.450253486633301,
      "epoch": 2.760023818975784,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004244505081576521,
      "loss": 5.0799,
      "mean_token_accuracy": 0.17779861092567445,
      "num_tokens": 60925872.0,
      "step": 27810
    },
    {
      "entropy": 5.441878080368042,
      "epoch": 2.7605200476379514,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042442402008616793,
      "loss": 5.1763,
      "mean_token_accuracy": 0.16904534697532653,
      "num_tokens": 60936589.0,
      "step": 27815
    },
    {
      "entropy": 5.453408908843994,
      "epoch": 2.7610162763001194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042439752830929,
      "loss": 5.1121,
      "mean_token_accuracy": 0.1685281962156296,
      "num_tokens": 60947213.0,
      "step": 27820
    },
    {
      "entropy": 5.54884147644043,
      "epoch": 2.761512504962287,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004243710328276754,
      "loss": 5.1291,
      "mean_token_accuracy": 0.16600079983472824,
      "num_tokens": 60958615.0,
      "step": 27825
    },
    {
      "entropy": 5.447000789642334,
      "epoch": 2.7620087336244543,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000424344533641981,
      "loss": 5.068,
      "mean_token_accuracy": 0.1739522784948349,
      "num_tokens": 60968830.0,
      "step": 27830
    },
    {
      "entropy": 5.408438491821289,
      "epoch": 2.762504962286622,
      "grad_norm": 1.046875,
      "learning_rate": 0.000424318030752864,
      "loss": 5.0571,
      "mean_token_accuracy": 0.17458976656198502,
      "num_tokens": 60979744.0,
      "step": 27835
    },
    {
      "entropy": 5.504576730728149,
      "epoch": 2.7630011909487893,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004242915241609815,
      "loss": 5.1446,
      "mean_token_accuracy": 0.16834144443273544,
      "num_tokens": 60990291.0,
      "step": 27840
    },
    {
      "entropy": 5.480570363998413,
      "epoch": 2.7634974196109567,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000424265013866991,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17506101727485657,
      "num_tokens": 61002170.0,
      "step": 27845
    },
    {
      "entropy": 5.417062044143677,
      "epoch": 2.763993648273124,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004242384998715497,
      "loss": 5.1475,
      "mean_token_accuracy": 0.16897553354501724,
      "num_tokens": 61012639.0,
      "step": 27850
    },
    {
      "entropy": 5.503439283370971,
      "epoch": 2.7644898769352917,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004242119821753153,
      "loss": 5.0981,
      "mean_token_accuracy": 0.1728903904557228,
      "num_tokens": 61022657.0,
      "step": 27855
    },
    {
      "entropy": 5.434080505371094,
      "epoch": 2.764986105597459,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004241854607789452,
      "loss": 5.0629,
      "mean_token_accuracy": 0.1745052829384804,
      "num_tokens": 61033108.0,
      "step": 27860
    },
    {
      "entropy": 5.385993528366089,
      "epoch": 2.7654823342596266,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042415893568309714,
      "loss": 5.0992,
      "mean_token_accuracy": 0.1731051728129387,
      "num_tokens": 61043651.0,
      "step": 27865
    },
    {
      "entropy": 5.377015447616577,
      "epoch": 2.765978562921794,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042413240688842884,
      "loss": 5.0235,
      "mean_token_accuracy": 0.1778884455561638,
      "num_tokens": 61053234.0,
      "step": 27870
    },
    {
      "entropy": 5.427054500579834,
      "epoch": 2.766474791583962,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004241058743955982,
      "loss": 5.0878,
      "mean_token_accuracy": 0.17177530527114868,
      "num_tokens": 61065178.0,
      "step": 27875
    },
    {
      "entropy": 5.446893310546875,
      "epoch": 2.7669710202461295,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004240793382052633,
      "loss": 5.0876,
      "mean_token_accuracy": 0.17880483120679855,
      "num_tokens": 61075709.0,
      "step": 27880
    },
    {
      "entropy": 5.485193157196045,
      "epoch": 2.767467248908297,
      "grad_norm": 1.109375,
      "learning_rate": 0.000424052798318082,
      "loss": 5.1636,
      "mean_token_accuracy": 0.16437891870737076,
      "num_tokens": 61086880.0,
      "step": 27885
    },
    {
      "entropy": 5.418242931365967,
      "epoch": 2.7679634775704645,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004240262547347124,
      "loss": 5.0894,
      "mean_token_accuracy": 0.17649204432964324,
      "num_tokens": 61097008.0,
      "step": 27890
    },
    {
      "entropy": 5.432586717605591,
      "epoch": 2.768459706232632,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042399970745581283,
      "loss": 5.0777,
      "mean_token_accuracy": 0.17331922948360443,
      "num_tokens": 61107844.0,
      "step": 27895
    },
    {
      "entropy": 5.376350688934326,
      "epoch": 2.7689559348947994,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042397315648204164,
      "loss": 5.0526,
      "mean_token_accuracy": 0.17908426821231843,
      "num_tokens": 61118652.0,
      "step": 27900
    },
    {
      "entropy": 5.5167131423950195,
      "epoch": 2.7694521635569673,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042394660181405716,
      "loss": 5.1883,
      "mean_token_accuracy": 0.16769085675477982,
      "num_tokens": 61129143.0,
      "step": 27905
    },
    {
      "entropy": 5.478622579574585,
      "epoch": 2.769948392219135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004239200434525179,
      "loss": 5.1295,
      "mean_token_accuracy": 0.1701539561152458,
      "num_tokens": 61140623.0,
      "step": 27910
    },
    {
      "entropy": 5.400295925140381,
      "epoch": 2.7704446208813023,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004238934813980824,
      "loss": 5.0769,
      "mean_token_accuracy": 0.16992235481739043,
      "num_tokens": 61150998.0,
      "step": 27915
    },
    {
      "entropy": 5.501615381240844,
      "epoch": 2.7709408495434698,
      "grad_norm": 1.0,
      "learning_rate": 0.00042386691565140947,
      "loss": 5.1191,
      "mean_token_accuracy": 0.17036477029323577,
      "num_tokens": 61162086.0,
      "step": 27920
    },
    {
      "entropy": 5.477650690078735,
      "epoch": 2.7714370782056372,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00042384034621315773,
      "loss": 5.0962,
      "mean_token_accuracy": 0.17159136533737182,
      "num_tokens": 61171806.0,
      "step": 27925
    },
    {
      "entropy": 5.418765211105347,
      "epoch": 2.7719333068678047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004238137730839862,
      "loss": 5.0605,
      "mean_token_accuracy": 0.17298756092786788,
      "num_tokens": 61182056.0,
      "step": 27930
    },
    {
      "entropy": 5.4366843700408936,
      "epoch": 2.772429535529972,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004237871962645538,
      "loss": 5.1134,
      "mean_token_accuracy": 0.17595559507608413,
      "num_tokens": 61193231.0,
      "step": 27935
    },
    {
      "entropy": 5.430489206314087,
      "epoch": 2.7729257641921397,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042376061575551946,
      "loss": 5.0789,
      "mean_token_accuracy": 0.17309942096471786,
      "num_tokens": 61203785.0,
      "step": 27940
    },
    {
      "entropy": 5.447842359542847,
      "epoch": 2.773421992854307,
      "grad_norm": 1.0,
      "learning_rate": 0.0004237340315575425,
      "loss": 5.1044,
      "mean_token_accuracy": 0.17260647863149642,
      "num_tokens": 61215244.0,
      "step": 27945
    },
    {
      "entropy": 5.5153991222381595,
      "epoch": 2.7739182215164746,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004237074436712819,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1730697736144066,
      "num_tokens": 61227147.0,
      "step": 27950
    },
    {
      "entropy": 5.5336809158325195,
      "epoch": 2.774414450178642,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042368085209739715,
      "loss": 5.1642,
      "mean_token_accuracy": 0.1695493459701538,
      "num_tokens": 61237687.0,
      "step": 27955
    },
    {
      "entropy": 5.379493141174317,
      "epoch": 2.77491067884081,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004236542568365478,
      "loss": 5.0388,
      "mean_token_accuracy": 0.1750490665435791,
      "num_tokens": 61247700.0,
      "step": 27960
    },
    {
      "entropy": 5.4336631298065186,
      "epoch": 2.7754069075029775,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042362765788939307,
      "loss": 5.0616,
      "mean_token_accuracy": 0.17034044712781907,
      "num_tokens": 61258407.0,
      "step": 27965
    },
    {
      "entropy": 5.508750534057617,
      "epoch": 2.775903136165145,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004236010552565927,
      "loss": 5.2264,
      "mean_token_accuracy": 0.1595378577709198,
      "num_tokens": 61269065.0,
      "step": 27970
    },
    {
      "entropy": 5.525210857391357,
      "epoch": 2.7763993648273124,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042357444893880645,
      "loss": 5.1233,
      "mean_token_accuracy": 0.17014157623052598,
      "num_tokens": 61280767.0,
      "step": 27975
    },
    {
      "entropy": 5.407552003860474,
      "epoch": 2.77689559348948,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042354783893669385,
      "loss": 5.0508,
      "mean_token_accuracy": 0.17168679237365722,
      "num_tokens": 61291967.0,
      "step": 27980
    },
    {
      "entropy": 5.494279289245606,
      "epoch": 2.7773918221516474,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004235212252509151,
      "loss": 5.2249,
      "mean_token_accuracy": 0.1629580646753311,
      "num_tokens": 61303572.0,
      "step": 27985
    },
    {
      "entropy": 5.51927170753479,
      "epoch": 2.777888050813815,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042349460788212995,
      "loss": 5.145,
      "mean_token_accuracy": 0.16700749695301056,
      "num_tokens": 61315674.0,
      "step": 27990
    },
    {
      "entropy": 5.458497142791748,
      "epoch": 2.7783842794759828,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042346798683099844,
      "loss": 5.0657,
      "mean_token_accuracy": 0.17250966578722,
      "num_tokens": 61326334.0,
      "step": 27995
    },
    {
      "entropy": 5.453553295135498,
      "epoch": 2.7788805081381502,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004234413620981809,
      "loss": 5.1836,
      "mean_token_accuracy": 0.1653658479452133,
      "num_tokens": 61337505.0,
      "step": 28000
    },
    {
      "entropy": 5.463118267059326,
      "epoch": 2.7793767368003177,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042341473368433735,
      "loss": 5.1557,
      "mean_token_accuracy": 0.17021935135126115,
      "num_tokens": 61348451.0,
      "step": 28005
    },
    {
      "entropy": 5.520605039596558,
      "epoch": 2.779872965462485,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004233881015901282,
      "loss": 5.2887,
      "mean_token_accuracy": 0.15823209136724473,
      "num_tokens": 61358356.0,
      "step": 28010
    },
    {
      "entropy": 5.568944644927979,
      "epoch": 2.7803691941246527,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004233614658162139,
      "loss": 5.1383,
      "mean_token_accuracy": 0.1707230880856514,
      "num_tokens": 61370996.0,
      "step": 28015
    },
    {
      "entropy": 5.579865837097168,
      "epoch": 2.78086542278682,
      "grad_norm": 1.421875,
      "learning_rate": 0.00042333482636325504,
      "loss": 5.1824,
      "mean_token_accuracy": 0.16914692968130113,
      "num_tokens": 61383928.0,
      "step": 28020
    },
    {
      "entropy": 5.474665021896362,
      "epoch": 2.7813616514489876,
      "grad_norm": 1.140625,
      "learning_rate": 0.000423308183231912,
      "loss": 5.1354,
      "mean_token_accuracy": 0.17074507027864455,
      "num_tokens": 61394277.0,
      "step": 28025
    },
    {
      "entropy": 5.4219786643981935,
      "epoch": 2.781857880111155,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042328153642284567,
      "loss": 5.1269,
      "mean_token_accuracy": 0.17878834754228592,
      "num_tokens": 61405928.0,
      "step": 28030
    },
    {
      "entropy": 5.488362789154053,
      "epoch": 2.7823541087733226,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004232548859367167,
      "loss": 5.1743,
      "mean_token_accuracy": 0.1681109443306923,
      "num_tokens": 61417893.0,
      "step": 28035
    },
    {
      "entropy": 5.482503032684326,
      "epoch": 2.78285033743549,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000423228231774186,
      "loss": 5.1265,
      "mean_token_accuracy": 0.17361967414617538,
      "num_tokens": 61428099.0,
      "step": 28040
    },
    {
      "entropy": 5.456796598434448,
      "epoch": 2.783346566097658,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004232015739359146,
      "loss": 5.0461,
      "mean_token_accuracy": 0.1778467997908592,
      "num_tokens": 61437636.0,
      "step": 28045
    },
    {
      "entropy": 5.411783885955811,
      "epoch": 2.7838427947598254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042317491242256346,
      "loss": 5.0325,
      "mean_token_accuracy": 0.17163978666067123,
      "num_tokens": 61448038.0,
      "step": 28050
    },
    {
      "entropy": 5.424706506729126,
      "epoch": 2.784339023421993,
      "grad_norm": 1.0,
      "learning_rate": 0.0004231482472347938,
      "loss": 5.1124,
      "mean_token_accuracy": 0.1669310837984085,
      "num_tokens": 61460258.0,
      "step": 28055
    },
    {
      "entropy": 5.461479949951172,
      "epoch": 2.7848352520841604,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004231215783732668,
      "loss": 5.1321,
      "mean_token_accuracy": 0.17254331558942795,
      "num_tokens": 61472274.0,
      "step": 28060
    },
    {
      "entropy": 5.45784101486206,
      "epoch": 2.785331480746328,
      "grad_norm": 1.125,
      "learning_rate": 0.00042309490583864395,
      "loss": 5.1147,
      "mean_token_accuracy": 0.17443737238645554,
      "num_tokens": 61482975.0,
      "step": 28065
    },
    {
      "entropy": 5.48743166923523,
      "epoch": 2.7858277094084953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004230682296315865,
      "loss": 5.1341,
      "mean_token_accuracy": 0.16580336093902587,
      "num_tokens": 61494223.0,
      "step": 28070
    },
    {
      "entropy": 5.531915092468262,
      "epoch": 2.786323938070663,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000423041549752756,
      "loss": 5.1855,
      "mean_token_accuracy": 0.17205650210380555,
      "num_tokens": 61505981.0,
      "step": 28075
    },
    {
      "entropy": 5.398397827148438,
      "epoch": 2.7868201667328307,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004230148662028141,
      "loss": 5.0741,
      "mean_token_accuracy": 0.1747310370206833,
      "num_tokens": 61516410.0,
      "step": 28080
    },
    {
      "entropy": 5.436687755584717,
      "epoch": 2.787316395394998,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042298817898242237,
      "loss": 5.0772,
      "mean_token_accuracy": 0.1710105925798416,
      "num_tokens": 61526551.0,
      "step": 28085
    },
    {
      "entropy": 5.4900243282318115,
      "epoch": 2.7878126240571657,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042296148809224275,
      "loss": 5.1975,
      "mean_token_accuracy": 0.16580548286437988,
      "num_tokens": 61538214.0,
      "step": 28090
    },
    {
      "entropy": 5.502868700027466,
      "epoch": 2.788308852719333,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004229347935329371,
      "loss": 5.1652,
      "mean_token_accuracy": 0.16959122121334075,
      "num_tokens": 61549726.0,
      "step": 28095
    },
    {
      "entropy": 5.600861978530884,
      "epoch": 2.7888050813815006,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042290809530516724,
      "loss": 5.2431,
      "mean_token_accuracy": 0.15913015753030776,
      "num_tokens": 61560070.0,
      "step": 28100
    },
    {
      "entropy": 5.583103132247925,
      "epoch": 2.789301310043668,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004228813934095954,
      "loss": 5.1759,
      "mean_token_accuracy": 0.16879895329475403,
      "num_tokens": 61571348.0,
      "step": 28105
    },
    {
      "entropy": 5.456355810165405,
      "epoch": 2.7897975387058356,
      "grad_norm": 1.125,
      "learning_rate": 0.00042285468784688363,
      "loss": 5.0457,
      "mean_token_accuracy": 0.17731528133153915,
      "num_tokens": 61581981.0,
      "step": 28110
    },
    {
      "entropy": 5.4488355159759525,
      "epoch": 2.790293767368003,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042282797861769423,
      "loss": 5.1395,
      "mean_token_accuracy": 0.16891846656799317,
      "num_tokens": 61591248.0,
      "step": 28115
    },
    {
      "entropy": 5.4661112308502195,
      "epoch": 2.7907899960301705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004228012657226895,
      "loss": 5.1511,
      "mean_token_accuracy": 0.16896012425422668,
      "num_tokens": 61602432.0,
      "step": 28120
    },
    {
      "entropy": 5.414995098114014,
      "epoch": 2.791286224692338,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042277454916253186,
      "loss": 5.1126,
      "mean_token_accuracy": 0.17025501728057862,
      "num_tokens": 61613418.0,
      "step": 28125
    },
    {
      "entropy": 5.460662031173706,
      "epoch": 2.791782453354506,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004227478289378838,
      "loss": 5.0889,
      "mean_token_accuracy": 0.16925387978553771,
      "num_tokens": 61623783.0,
      "step": 28130
    },
    {
      "entropy": 5.550783252716064,
      "epoch": 2.7922786820166734,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042272110504940795,
      "loss": 5.1807,
      "mean_token_accuracy": 0.1668537437915802,
      "num_tokens": 61633838.0,
      "step": 28135
    },
    {
      "entropy": 5.44984188079834,
      "epoch": 2.792774910678841,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042269437749776703,
      "loss": 5.0595,
      "mean_token_accuracy": 0.17637290060520172,
      "num_tokens": 61645982.0,
      "step": 28140
    },
    {
      "entropy": 5.380315160751342,
      "epoch": 2.7932711393410083,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004226676462836238,
      "loss": 5.0334,
      "mean_token_accuracy": 0.17695634365081786,
      "num_tokens": 61656030.0,
      "step": 28145
    },
    {
      "entropy": 5.423975563049316,
      "epoch": 2.793767368003176,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004226409114076411,
      "loss": 5.045,
      "mean_token_accuracy": 0.17367569506168365,
      "num_tokens": 61666746.0,
      "step": 28150
    },
    {
      "entropy": 5.536663436889649,
      "epoch": 2.7942635966653433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042261417287048196,
      "loss": 5.2199,
      "mean_token_accuracy": 0.1676828846335411,
      "num_tokens": 61677681.0,
      "step": 28155
    },
    {
      "entropy": 5.53027491569519,
      "epoch": 2.7947598253275108,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004225874306728094,
      "loss": 5.1854,
      "mean_token_accuracy": 0.16718290597200394,
      "num_tokens": 61689564.0,
      "step": 28160
    },
    {
      "entropy": 5.467653894424439,
      "epoch": 2.7952560539896787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042256068481528666,
      "loss": 5.0968,
      "mean_token_accuracy": 0.17398060858249664,
      "num_tokens": 61701074.0,
      "step": 28165
    },
    {
      "entropy": 5.413700246810913,
      "epoch": 2.795752282651846,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004225339352985768,
      "loss": 5.0847,
      "mean_token_accuracy": 0.17463179677724838,
      "num_tokens": 61712203.0,
      "step": 28170
    },
    {
      "entropy": 5.407344579696655,
      "epoch": 2.7962485113140136,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042250718212334334,
      "loss": 5.0347,
      "mean_token_accuracy": 0.17474863678216934,
      "num_tokens": 61723362.0,
      "step": 28175
    },
    {
      "entropy": 5.461672687530518,
      "epoch": 2.796744739976181,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042248042529024956,
      "loss": 5.0572,
      "mean_token_accuracy": 0.17400580793619155,
      "num_tokens": 61733454.0,
      "step": 28180
    },
    {
      "entropy": 5.533198404312134,
      "epoch": 2.7972409686383486,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042245366479995894,
      "loss": 5.181,
      "mean_token_accuracy": 0.16883738338947296,
      "num_tokens": 61745014.0,
      "step": 28185
    },
    {
      "entropy": 5.42902774810791,
      "epoch": 2.797737197300516,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042242690065313535,
      "loss": 5.1134,
      "mean_token_accuracy": 0.17180700302124025,
      "num_tokens": 61755334.0,
      "step": 28190
    },
    {
      "entropy": 5.378274536132812,
      "epoch": 2.7982334259626835,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042240013285044215,
      "loss": 5.0558,
      "mean_token_accuracy": 0.1781066343188286,
      "num_tokens": 61766024.0,
      "step": 28195
    },
    {
      "entropy": 5.50236964225769,
      "epoch": 2.798729654624851,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042237336139254336,
      "loss": 5.1231,
      "mean_token_accuracy": 0.17053781747817992,
      "num_tokens": 61778228.0,
      "step": 28200
    },
    {
      "entropy": 5.518052673339843,
      "epoch": 2.7992258832870185,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042234658628010274,
      "loss": 5.1444,
      "mean_token_accuracy": 0.1673691377043724,
      "num_tokens": 61789072.0,
      "step": 28205
    },
    {
      "entropy": 5.465368556976318,
      "epoch": 2.799722111949186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042231980751378435,
      "loss": 5.103,
      "mean_token_accuracy": 0.17574398070573807,
      "num_tokens": 61799503.0,
      "step": 28210
    },
    {
      "entropy": 5.434835195541382,
      "epoch": 2.8002183406113534,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042229302509425205,
      "loss": 5.0639,
      "mean_token_accuracy": 0.1757493644952774,
      "num_tokens": 61811063.0,
      "step": 28215
    },
    {
      "entropy": 5.457683897018432,
      "epoch": 2.8007145692735214,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004222662390221702,
      "loss": 5.1773,
      "mean_token_accuracy": 0.16746330857276917,
      "num_tokens": 61822832.0,
      "step": 28220
    },
    {
      "entropy": 5.544493770599365,
      "epoch": 2.801210797935689,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004222394492982029,
      "loss": 5.1391,
      "mean_token_accuracy": 0.1647442251443863,
      "num_tokens": 61832456.0,
      "step": 28225
    },
    {
      "entropy": 5.477757978439331,
      "epoch": 2.8017070265978563,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004222126559230146,
      "loss": 5.0704,
      "mean_token_accuracy": 0.16981329917907714,
      "num_tokens": 61843043.0,
      "step": 28230
    },
    {
      "entropy": 5.395609664916992,
      "epoch": 2.802203255260024,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004221858588972696,
      "loss": 5.1114,
      "mean_token_accuracy": 0.17036620974540712,
      "num_tokens": 61853249.0,
      "step": 28235
    },
    {
      "entropy": 5.3665382862091064,
      "epoch": 2.8026994839221913,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042215905822163245,
      "loss": 5.0478,
      "mean_token_accuracy": 0.1785596489906311,
      "num_tokens": 61863842.0,
      "step": 28240
    },
    {
      "entropy": 5.438422775268554,
      "epoch": 2.8031957125843587,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004221322538967678,
      "loss": 5.1404,
      "mean_token_accuracy": 0.17405532896518708,
      "num_tokens": 61876924.0,
      "step": 28245
    },
    {
      "entropy": 5.563470363616943,
      "epoch": 2.8036919412465267,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004221054459233403,
      "loss": 5.1676,
      "mean_token_accuracy": 0.16787475794553758,
      "num_tokens": 61886766.0,
      "step": 28250
    },
    {
      "entropy": 5.516751766204834,
      "epoch": 2.804188169908694,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042207863430201466,
      "loss": 5.1273,
      "mean_token_accuracy": 0.17408113181591034,
      "num_tokens": 61896538.0,
      "step": 28255
    },
    {
      "entropy": 5.364685344696045,
      "epoch": 2.8046843985708616,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042205181903345586,
      "loss": 5.1054,
      "mean_token_accuracy": 0.16500231772661209,
      "num_tokens": 61907328.0,
      "step": 28260
    },
    {
      "entropy": 5.472604513168335,
      "epoch": 2.805180627233029,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004220250001183288,
      "loss": 5.136,
      "mean_token_accuracy": 0.16505857259035112,
      "num_tokens": 61918797.0,
      "step": 28265
    },
    {
      "entropy": 5.523580741882324,
      "epoch": 2.8056768558951966,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004219981775572986,
      "loss": 5.1355,
      "mean_token_accuracy": 0.17546123415231704,
      "num_tokens": 61928755.0,
      "step": 28270
    },
    {
      "entropy": 5.447868776321411,
      "epoch": 2.806173084557364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004219713513510303,
      "loss": 5.1341,
      "mean_token_accuracy": 0.1697654977440834,
      "num_tokens": 61939993.0,
      "step": 28275
    },
    {
      "entropy": 5.404237604141235,
      "epoch": 2.8066693132195315,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042194452150018917,
      "loss": 5.0088,
      "mean_token_accuracy": 0.18431564569473266,
      "num_tokens": 61950982.0,
      "step": 28280
    },
    {
      "entropy": 5.455752325057984,
      "epoch": 2.807165541881699,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042191768800544055,
      "loss": 5.0882,
      "mean_token_accuracy": 0.17096222639083863,
      "num_tokens": 61961786.0,
      "step": 28285
    },
    {
      "entropy": 5.441495752334594,
      "epoch": 2.8076617705438665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042189085086744986,
      "loss": 5.0978,
      "mean_token_accuracy": 0.1792908400297165,
      "num_tokens": 61972548.0,
      "step": 28290
    },
    {
      "entropy": 5.440370559692383,
      "epoch": 2.808157999206034,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004218640100868826,
      "loss": 5.0663,
      "mean_token_accuracy": 0.17463562041521072,
      "num_tokens": 61981822.0,
      "step": 28295
    },
    {
      "entropy": 5.479927730560303,
      "epoch": 2.8086542278682014,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042183716566440433,
      "loss": 5.2034,
      "mean_token_accuracy": 0.15787188708782196,
      "num_tokens": 61992997.0,
      "step": 28300
    },
    {
      "entropy": 5.478485250473023,
      "epoch": 2.8091504565303693,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004218103176006808,
      "loss": 5.1404,
      "mean_token_accuracy": 0.16722093224525453,
      "num_tokens": 62004391.0,
      "step": 28305
    },
    {
      "entropy": 5.4674324035644535,
      "epoch": 2.809646685192537,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004217834658963777,
      "loss": 5.1028,
      "mean_token_accuracy": 0.1794124349951744,
      "num_tokens": 62014099.0,
      "step": 28310
    },
    {
      "entropy": 5.477788019180298,
      "epoch": 2.8101429138547043,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042175661055216093,
      "loss": 5.0756,
      "mean_token_accuracy": 0.17358602285385133,
      "num_tokens": 62023872.0,
      "step": 28315
    },
    {
      "entropy": 5.4834206104278564,
      "epoch": 2.8106391425168717,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004217297515686965,
      "loss": 5.0731,
      "mean_token_accuracy": 0.17100338637828827,
      "num_tokens": 62035566.0,
      "step": 28320
    },
    {
      "entropy": 5.478532171249389,
      "epoch": 2.811135371179039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004217028889466503,
      "loss": 5.1148,
      "mean_token_accuracy": 0.16858033686876298,
      "num_tokens": 62046742.0,
      "step": 28325
    },
    {
      "entropy": 5.444647407531738,
      "epoch": 2.8116315998412067,
      "grad_norm": 1.125,
      "learning_rate": 0.0004216760226866887,
      "loss": 5.0636,
      "mean_token_accuracy": 0.1705285593867302,
      "num_tokens": 62056545.0,
      "step": 28330
    },
    {
      "entropy": 5.464788961410522,
      "epoch": 2.8121278285033746,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042164915278947775,
      "loss": 5.1794,
      "mean_token_accuracy": 0.16808093190193177,
      "num_tokens": 62066414.0,
      "step": 28335
    },
    {
      "entropy": 5.407382774353027,
      "epoch": 2.812624057165542,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004216222792556838,
      "loss": 5.0706,
      "mean_token_accuracy": 0.17247609049081802,
      "num_tokens": 62078406.0,
      "step": 28340
    },
    {
      "entropy": 5.45408034324646,
      "epoch": 2.8131202858277096,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004215954020859732,
      "loss": 5.1222,
      "mean_token_accuracy": 0.16855347156524658,
      "num_tokens": 62089516.0,
      "step": 28345
    },
    {
      "entropy": 5.497828531265259,
      "epoch": 2.813616514489877,
      "grad_norm": 1.0,
      "learning_rate": 0.00042156852128101264,
      "loss": 5.0858,
      "mean_token_accuracy": 0.1780945688486099,
      "num_tokens": 62099287.0,
      "step": 28350
    },
    {
      "entropy": 5.439867877960205,
      "epoch": 2.8141127431520445,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042154163684146855,
      "loss": 5.0436,
      "mean_token_accuracy": 0.1779022827744484,
      "num_tokens": 62111110.0,
      "step": 28355
    },
    {
      "entropy": 5.568574476242065,
      "epoch": 2.814608971814212,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004215147487680076,
      "loss": 5.2296,
      "mean_token_accuracy": 0.15969937890768052,
      "num_tokens": 62123614.0,
      "step": 28360
    },
    {
      "entropy": 5.4243998527526855,
      "epoch": 2.8151052004763795,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042148785706129657,
      "loss": 5.1716,
      "mean_token_accuracy": 0.16929552257061004,
      "num_tokens": 62133682.0,
      "step": 28365
    },
    {
      "entropy": 5.437830305099487,
      "epoch": 2.815601429138547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004214609617220024,
      "loss": 5.0913,
      "mean_token_accuracy": 0.1738521561026573,
      "num_tokens": 62144802.0,
      "step": 28370
    },
    {
      "entropy": 5.479652070999146,
      "epoch": 2.8160976578007144,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042143406275079186,
      "loss": 5.0543,
      "mean_token_accuracy": 0.17173270285129547,
      "num_tokens": 62154205.0,
      "step": 28375
    },
    {
      "entropy": 5.427470397949219,
      "epoch": 2.816593886462882,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004214071601483322,
      "loss": 4.9976,
      "mean_token_accuracy": 0.18574858456850052,
      "num_tokens": 62164648.0,
      "step": 28380
    },
    {
      "entropy": 5.447363424301147,
      "epoch": 2.8170901151250494,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004213802539152905,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17512689679861068,
      "num_tokens": 62176073.0,
      "step": 28385
    },
    {
      "entropy": 5.499179029464722,
      "epoch": 2.8175863437872173,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042135334405233384,
      "loss": 5.124,
      "mean_token_accuracy": 0.1672062948346138,
      "num_tokens": 62187350.0,
      "step": 28390
    },
    {
      "entropy": 5.424795150756836,
      "epoch": 2.8180825724493848,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004213264305601296,
      "loss": 5.0682,
      "mean_token_accuracy": 0.1707600921392441,
      "num_tokens": 62197729.0,
      "step": 28395
    },
    {
      "entropy": 5.446064138412476,
      "epoch": 2.8185788011115522,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042129951343934527,
      "loss": 5.0462,
      "mean_token_accuracy": 0.17746407836675643,
      "num_tokens": 62209047.0,
      "step": 28400
    },
    {
      "entropy": 5.444215106964111,
      "epoch": 2.8190750297737197,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004212725926906482,
      "loss": 5.0664,
      "mean_token_accuracy": 0.17244853377342223,
      "num_tokens": 62218859.0,
      "step": 28405
    },
    {
      "entropy": 5.412169599533081,
      "epoch": 2.819571258435887,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042124566831470604,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1752679944038391,
      "num_tokens": 62230396.0,
      "step": 28410
    },
    {
      "entropy": 5.440034198760986,
      "epoch": 2.8200674870980547,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004212187403121864,
      "loss": 5.0591,
      "mean_token_accuracy": 0.17750144749879837,
      "num_tokens": 62241790.0,
      "step": 28415
    },
    {
      "entropy": 5.4391865730285645,
      "epoch": 2.8205637157602226,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042119180868375707,
      "loss": 5.0489,
      "mean_token_accuracy": 0.1672515869140625,
      "num_tokens": 62253820.0,
      "step": 28420
    },
    {
      "entropy": 5.416459798812866,
      "epoch": 2.82105994442239,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004211648734300859,
      "loss": 5.0865,
      "mean_token_accuracy": 0.1721369966864586,
      "num_tokens": 62263875.0,
      "step": 28425
    },
    {
      "entropy": 5.473156499862671,
      "epoch": 2.8215561730845575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042113793455184084,
      "loss": 5.129,
      "mean_token_accuracy": 0.16724544614553452,
      "num_tokens": 62275390.0,
      "step": 28430
    },
    {
      "entropy": 5.464316654205322,
      "epoch": 2.822052401746725,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004211109920496899,
      "loss": 5.1538,
      "mean_token_accuracy": 0.16743456125259398,
      "num_tokens": 62286969.0,
      "step": 28435
    },
    {
      "entropy": 5.483538246154785,
      "epoch": 2.8225486304088925,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042108404592430126,
      "loss": 5.0917,
      "mean_token_accuracy": 0.17318332195281982,
      "num_tokens": 62297810.0,
      "step": 28440
    },
    {
      "entropy": 5.45084056854248,
      "epoch": 2.82304485907106,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000421057096176343,
      "loss": 5.0884,
      "mean_token_accuracy": 0.1727159634232521,
      "num_tokens": 62309474.0,
      "step": 28445
    },
    {
      "entropy": 5.4141950607299805,
      "epoch": 2.8235410877332274,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004210301428064835,
      "loss": 5.1088,
      "mean_token_accuracy": 0.1724878281354904,
      "num_tokens": 62320022.0,
      "step": 28450
    },
    {
      "entropy": 5.504427528381347,
      "epoch": 2.824037316395395,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004210031858153911,
      "loss": 5.1293,
      "mean_token_accuracy": 0.17871991991996766,
      "num_tokens": 62330450.0,
      "step": 28455
    },
    {
      "entropy": 5.498330879211426,
      "epoch": 2.8245335450575624,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042097622520373434,
      "loss": 5.1285,
      "mean_token_accuracy": 0.16908367723226547,
      "num_tokens": 62341419.0,
      "step": 28460
    },
    {
      "entropy": 5.473728513717651,
      "epoch": 2.82502977371973,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004209492609721817,
      "loss": 5.1602,
      "mean_token_accuracy": 0.16758544892072677,
      "num_tokens": 62353312.0,
      "step": 28465
    },
    {
      "entropy": 5.503668022155762,
      "epoch": 2.8255260023818973,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004209222931214019,
      "loss": 5.1079,
      "mean_token_accuracy": 0.1746346265077591,
      "num_tokens": 62364838.0,
      "step": 28470
    },
    {
      "entropy": 5.443870210647583,
      "epoch": 2.8260222310440652,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042089532165206367,
      "loss": 5.0371,
      "mean_token_accuracy": 0.17969904243946075,
      "num_tokens": 62376443.0,
      "step": 28475
    },
    {
      "entropy": 5.51093225479126,
      "epoch": 2.8265184597062327,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004208683465648358,
      "loss": 5.1999,
      "mean_token_accuracy": 0.16524693816900254,
      "num_tokens": 62387263.0,
      "step": 28480
    },
    {
      "entropy": 5.461005544662475,
      "epoch": 2.8270146883684,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042084136786038737,
      "loss": 5.0379,
      "mean_token_accuracy": 0.17223937064409256,
      "num_tokens": 62398929.0,
      "step": 28485
    },
    {
      "entropy": 5.450663900375366,
      "epoch": 2.8275109170305677,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004208143855393872,
      "loss": 5.1499,
      "mean_token_accuracy": 0.16695887148380278,
      "num_tokens": 62409516.0,
      "step": 28490
    },
    {
      "entropy": 5.564820718765259,
      "epoch": 2.828007145692735,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042078739960250446,
      "loss": 5.1232,
      "mean_token_accuracy": 0.17345669269561767,
      "num_tokens": 62421856.0,
      "step": 28495
    },
    {
      "entropy": 5.508591985702514,
      "epoch": 2.8285033743549026,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042076041005040835,
      "loss": 5.1061,
      "mean_token_accuracy": 0.1743642657995224,
      "num_tokens": 62432340.0,
      "step": 28500
    },
    {
      "entropy": 5.422167778015137,
      "epoch": 2.8289996030170705,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004207334168837682,
      "loss": 5.1259,
      "mean_token_accuracy": 0.16557691395282745,
      "num_tokens": 62442652.0,
      "step": 28505
    },
    {
      "entropy": 5.505354928970337,
      "epoch": 2.829495831679238,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004207064201032533,
      "loss": 5.0446,
      "mean_token_accuracy": 0.17968766242265702,
      "num_tokens": 62454503.0,
      "step": 28510
    },
    {
      "entropy": 5.4105370998382565,
      "epoch": 2.8299920603414055,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042067941970953316,
      "loss": 5.043,
      "mean_token_accuracy": 0.17613838911056517,
      "num_tokens": 62464858.0,
      "step": 28515
    },
    {
      "entropy": 5.485928964614868,
      "epoch": 2.830488289003573,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042065241570327734,
      "loss": 5.1211,
      "mean_token_accuracy": 0.171981143951416,
      "num_tokens": 62475153.0,
      "step": 28520
    },
    {
      "entropy": 5.479932641983032,
      "epoch": 2.8309845176657404,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004206254080851555,
      "loss": 5.1724,
      "mean_token_accuracy": 0.16528775840997695,
      "num_tokens": 62487883.0,
      "step": 28525
    },
    {
      "entropy": 5.430444622039795,
      "epoch": 2.831480746327908,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042059839685583726,
      "loss": 5.0812,
      "mean_token_accuracy": 0.1752985492348671,
      "num_tokens": 62499905.0,
      "step": 28530
    },
    {
      "entropy": 5.480695152282715,
      "epoch": 2.8319769749900754,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042057138201599247,
      "loss": 5.1425,
      "mean_token_accuracy": 0.16592279076576233,
      "num_tokens": 62511403.0,
      "step": 28535
    },
    {
      "entropy": 5.506823492050171,
      "epoch": 2.832473203652243,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042054436356629113,
      "loss": 5.0996,
      "mean_token_accuracy": 0.17641578018665313,
      "num_tokens": 62522543.0,
      "step": 28540
    },
    {
      "entropy": 5.434993076324463,
      "epoch": 2.8329694323144103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004205173415074032,
      "loss": 5.0941,
      "mean_token_accuracy": 0.17132963985204697,
      "num_tokens": 62533053.0,
      "step": 28545
    },
    {
      "entropy": 5.491401529312133,
      "epoch": 2.833465660976578,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004204903158399989,
      "loss": 5.1859,
      "mean_token_accuracy": 0.16497590988874436,
      "num_tokens": 62543699.0,
      "step": 28550
    },
    {
      "entropy": 5.438759279251099,
      "epoch": 2.8339618896387453,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004204632865647481,
      "loss": 5.1353,
      "mean_token_accuracy": 0.17424153685569763,
      "num_tokens": 62554121.0,
      "step": 28555
    },
    {
      "entropy": 5.449421358108521,
      "epoch": 2.834458118300913,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004204362536823213,
      "loss": 5.0695,
      "mean_token_accuracy": 0.1741992190480232,
      "num_tokens": 62564010.0,
      "step": 28560
    },
    {
      "entropy": 5.433414697647095,
      "epoch": 2.8349543469630807,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042040921719338877,
      "loss": 5.0389,
      "mean_token_accuracy": 0.1787270113825798,
      "num_tokens": 62574451.0,
      "step": 28565
    },
    {
      "entropy": 5.441277694702149,
      "epoch": 2.835450575625248,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000420382177098621,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17146547585725785,
      "num_tokens": 62585298.0,
      "step": 28570
    },
    {
      "entropy": 5.450202465057373,
      "epoch": 2.8359468042874156,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042035513339868855,
      "loss": 5.1762,
      "mean_token_accuracy": 0.17270709127187728,
      "num_tokens": 62596248.0,
      "step": 28575
    },
    {
      "entropy": 5.518754529953003,
      "epoch": 2.836443032949583,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004203280860942619,
      "loss": 5.1147,
      "mean_token_accuracy": 0.1689852923154831,
      "num_tokens": 62607233.0,
      "step": 28580
    },
    {
      "entropy": 5.453936290740967,
      "epoch": 2.8369392616117506,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004203010351860119,
      "loss": 5.0368,
      "mean_token_accuracy": 0.18131799548864364,
      "num_tokens": 62618564.0,
      "step": 28585
    },
    {
      "entropy": 5.428332090377808,
      "epoch": 2.837435490273918,
      "grad_norm": 1.125,
      "learning_rate": 0.0004202739806746094,
      "loss": 5.1235,
      "mean_token_accuracy": 0.16745442897081375,
      "num_tokens": 62630049.0,
      "step": 28590
    },
    {
      "entropy": 5.480254220962524,
      "epoch": 2.837931718936086,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004202469225607252,
      "loss": 5.166,
      "mean_token_accuracy": 0.17049575597047806,
      "num_tokens": 62640887.0,
      "step": 28595
    },
    {
      "entropy": 5.508724784851074,
      "epoch": 2.8384279475982535,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004202198608450302,
      "loss": 5.1082,
      "mean_token_accuracy": 0.16587840914726257,
      "num_tokens": 62651816.0,
      "step": 28600
    },
    {
      "entropy": 5.472384405136109,
      "epoch": 2.838924176260421,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004201927955281957,
      "loss": 5.0734,
      "mean_token_accuracy": 0.17440974712371826,
      "num_tokens": 62662601.0,
      "step": 28605
    },
    {
      "entropy": 5.415324258804321,
      "epoch": 2.8394204049225884,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004201657266108926,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17656342536211014,
      "num_tokens": 62673661.0,
      "step": 28610
    },
    {
      "entropy": 5.405610656738281,
      "epoch": 2.839916633584756,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042013865409379244,
      "loss": 5.0451,
      "mean_token_accuracy": 0.1766412228345871,
      "num_tokens": 62683984.0,
      "step": 28615
    },
    {
      "entropy": 5.475344038009643,
      "epoch": 2.8404128622469234,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004201115779775663,
      "loss": 5.1362,
      "mean_token_accuracy": 0.1696288362145424,
      "num_tokens": 62695338.0,
      "step": 28620
    },
    {
      "entropy": 5.500081205368042,
      "epoch": 2.840909090909091,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004200844982628857,
      "loss": 5.1597,
      "mean_token_accuracy": 0.16935080736875535,
      "num_tokens": 62706341.0,
      "step": 28625
    },
    {
      "entropy": 5.502814865112304,
      "epoch": 2.8414053195712583,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004200574149504222,
      "loss": 5.1651,
      "mean_token_accuracy": 0.17702099233865737,
      "num_tokens": 62717895.0,
      "step": 28630
    },
    {
      "entropy": 5.423112916946411,
      "epoch": 2.841901548233426,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004200303280408474,
      "loss": 5.1079,
      "mean_token_accuracy": 0.16968428045511247,
      "num_tokens": 62729745.0,
      "step": 28635
    },
    {
      "entropy": 5.4326262950897215,
      "epoch": 2.8423977768955933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042000323753483296,
      "loss": 5.0935,
      "mean_token_accuracy": 0.17901426702737808,
      "num_tokens": 62740298.0,
      "step": 28640
    },
    {
      "entropy": 5.430450916290283,
      "epoch": 2.842894005557761,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041997614343305067,
      "loss": 5.0779,
      "mean_token_accuracy": 0.17826662659645082,
      "num_tokens": 62751292.0,
      "step": 28645
    },
    {
      "entropy": 5.511986827850341,
      "epoch": 2.8433902342199286,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041994904573617247,
      "loss": 5.2035,
      "mean_token_accuracy": 0.16418781727552414,
      "num_tokens": 62761789.0,
      "step": 28650
    },
    {
      "entropy": 5.455896091461182,
      "epoch": 2.843886462882096,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041992194444487026,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17685592025518418,
      "num_tokens": 62773317.0,
      "step": 28655
    },
    {
      "entropy": 5.433033466339111,
      "epoch": 2.8443826915442636,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041989483955981607,
      "loss": 5.0678,
      "mean_token_accuracy": 0.1764715939760208,
      "num_tokens": 62783780.0,
      "step": 28660
    },
    {
      "entropy": 5.480651712417602,
      "epoch": 2.844878920206431,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041986773108168216,
      "loss": 5.0875,
      "mean_token_accuracy": 0.17018549144268036,
      "num_tokens": 62794535.0,
      "step": 28665
    },
    {
      "entropy": 5.418764305114746,
      "epoch": 2.8453751488685985,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041984061901114057,
      "loss": 5.0754,
      "mean_token_accuracy": 0.17186801135540009,
      "num_tokens": 62804410.0,
      "step": 28670
    },
    {
      "entropy": 5.455482482910156,
      "epoch": 2.845871377530766,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004198135033488638,
      "loss": 5.0655,
      "mean_token_accuracy": 0.1815451681613922,
      "num_tokens": 62814693.0,
      "step": 28675
    },
    {
      "entropy": 5.448589658737182,
      "epoch": 2.846367606192934,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041978638409552417,
      "loss": 5.0756,
      "mean_token_accuracy": 0.17521673440933228,
      "num_tokens": 62826114.0,
      "step": 28680
    },
    {
      "entropy": 5.392014646530152,
      "epoch": 2.8468638348551014,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004197592612517942,
      "loss": 5.1191,
      "mean_token_accuracy": 0.17158463299274446,
      "num_tokens": 62838130.0,
      "step": 28685
    },
    {
      "entropy": 5.552255010604858,
      "epoch": 2.847360063517269,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004197321348183465,
      "loss": 5.1601,
      "mean_token_accuracy": 0.16662283092737198,
      "num_tokens": 62849290.0,
      "step": 28690
    },
    {
      "entropy": 5.527254247665406,
      "epoch": 2.8478562921794364,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041970500479585364,
      "loss": 5.2314,
      "mean_token_accuracy": 0.17016611844301224,
      "num_tokens": 62860888.0,
      "step": 28695
    },
    {
      "entropy": 5.4940873146057125,
      "epoch": 2.848352520841604,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041967787118498854,
      "loss": 5.135,
      "mean_token_accuracy": 0.1783927157521248,
      "num_tokens": 62870231.0,
      "step": 28700
    },
    {
      "entropy": 5.454318618774414,
      "epoch": 2.8488487495037713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004196507339864241,
      "loss": 5.0398,
      "mean_token_accuracy": 0.17601525485515596,
      "num_tokens": 62881147.0,
      "step": 28705
    },
    {
      "entropy": 5.45527868270874,
      "epoch": 2.849344978165939,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000419623593200833,
      "loss": 5.0656,
      "mean_token_accuracy": 0.1774405851960182,
      "num_tokens": 62891272.0,
      "step": 28710
    },
    {
      "entropy": 5.392216777801513,
      "epoch": 2.8498412068281063,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041959644882888844,
      "loss": 5.0623,
      "mean_token_accuracy": 0.17533062547445297,
      "num_tokens": 62902376.0,
      "step": 28715
    },
    {
      "entropy": 5.418603134155274,
      "epoch": 2.8503374354902737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041956930087126357,
      "loss": 5.0623,
      "mean_token_accuracy": 0.1776852563023567,
      "num_tokens": 62913811.0,
      "step": 28720
    },
    {
      "entropy": 5.440509986877442,
      "epoch": 2.850833664152441,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041954214932863156,
      "loss": 5.1132,
      "mean_token_accuracy": 0.17037979066371917,
      "num_tokens": 62926001.0,
      "step": 28725
    },
    {
      "entropy": 5.430136394500733,
      "epoch": 2.851329892814609,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004195149942016657,
      "loss": 5.0726,
      "mean_token_accuracy": 0.1710929110646248,
      "num_tokens": 62936183.0,
      "step": 28730
    },
    {
      "entropy": 5.431177854537964,
      "epoch": 2.8518261214767766,
      "grad_norm": 1.125,
      "learning_rate": 0.0004194878354910394,
      "loss": 5.0716,
      "mean_token_accuracy": 0.17507654428482056,
      "num_tokens": 62947427.0,
      "step": 28735
    },
    {
      "entropy": 5.484645175933838,
      "epoch": 2.852322350138944,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041946067319742605,
      "loss": 5.1456,
      "mean_token_accuracy": 0.1710016116499901,
      "num_tokens": 62958630.0,
      "step": 28740
    },
    {
      "entropy": 5.422136402130127,
      "epoch": 2.8528185788011116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041943350732149935,
      "loss": 5.0675,
      "mean_token_accuracy": 0.17519067227840424,
      "num_tokens": 62968856.0,
      "step": 28745
    },
    {
      "entropy": 5.422419738769531,
      "epoch": 2.853314807463279,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041940633786393295,
      "loss": 4.9784,
      "mean_token_accuracy": 0.1811966121196747,
      "num_tokens": 62978393.0,
      "step": 28750
    },
    {
      "entropy": 5.423931884765625,
      "epoch": 2.8538110361254465,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041937916482540045,
      "loss": 5.1078,
      "mean_token_accuracy": 0.17664023041725158,
      "num_tokens": 62989786.0,
      "step": 28755
    },
    {
      "entropy": 5.4463695049285885,
      "epoch": 2.854307264787614,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041935198820657575,
      "loss": 5.0917,
      "mean_token_accuracy": 0.17437742352485658,
      "num_tokens": 63000021.0,
      "step": 28760
    },
    {
      "entropy": 5.472069835662841,
      "epoch": 2.854803493449782,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004193248080081329,
      "loss": 5.1171,
      "mean_token_accuracy": 0.16755940467119218,
      "num_tokens": 63011398.0,
      "step": 28765
    },
    {
      "entropy": 5.445945405960083,
      "epoch": 2.8552997221119494,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004192976242307457,
      "loss": 5.0835,
      "mean_token_accuracy": 0.17155231088399886,
      "num_tokens": 63021155.0,
      "step": 28770
    },
    {
      "entropy": 5.469005060195923,
      "epoch": 2.855795950774117,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004192704368750884,
      "loss": 5.1183,
      "mean_token_accuracy": 0.1709684818983078,
      "num_tokens": 63031510.0,
      "step": 28775
    },
    {
      "entropy": 5.526885557174682,
      "epoch": 2.8562921794362843,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004192432459418351,
      "loss": 5.1205,
      "mean_token_accuracy": 0.17284403294324874,
      "num_tokens": 63041978.0,
      "step": 28780
    },
    {
      "entropy": 5.500367593765259,
      "epoch": 2.856788408098452,
      "grad_norm": 1.0,
      "learning_rate": 0.0004192160514316602,
      "loss": 5.1573,
      "mean_token_accuracy": 0.17283862382173537,
      "num_tokens": 63052194.0,
      "step": 28785
    },
    {
      "entropy": 5.3944731712341305,
      "epoch": 2.8572846367606193,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004191888533452379,
      "loss": 5.0051,
      "mean_token_accuracy": 0.17668139338493347,
      "num_tokens": 63062144.0,
      "step": 28790
    },
    {
      "entropy": 5.406002378463745,
      "epoch": 2.8577808654227868,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004191616516832428,
      "loss": 5.1247,
      "mean_token_accuracy": 0.16992093473672867,
      "num_tokens": 63072403.0,
      "step": 28795
    },
    {
      "entropy": 5.434264087677002,
      "epoch": 2.8582770940849542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004191344464463494,
      "loss": 5.0547,
      "mean_token_accuracy": 0.17670615166425704,
      "num_tokens": 63084799.0,
      "step": 28800
    },
    {
      "entropy": 5.441890001296997,
      "epoch": 2.8587733227471217,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041910723763523224,
      "loss": 5.1541,
      "mean_token_accuracy": 0.1822491019964218,
      "num_tokens": 63095437.0,
      "step": 28805
    },
    {
      "entropy": 5.41574764251709,
      "epoch": 2.859269551409289,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004190800252505661,
      "loss": 5.1627,
      "mean_token_accuracy": 0.16624983996152878,
      "num_tokens": 63106743.0,
      "step": 28810
    },
    {
      "entropy": 5.438851022720337,
      "epoch": 2.8597657800714567,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004190528092930259,
      "loss": 5.0713,
      "mean_token_accuracy": 0.1725921928882599,
      "num_tokens": 63120525.0,
      "step": 28815
    },
    {
      "entropy": 5.467798089981079,
      "epoch": 2.8602620087336246,
      "grad_norm": 1.0,
      "learning_rate": 0.0004190255897632863,
      "loss": 5.0845,
      "mean_token_accuracy": 0.16949854046106339,
      "num_tokens": 63131510.0,
      "step": 28820
    },
    {
      "entropy": 5.471356344223023,
      "epoch": 2.860758237395792,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041899836666202254,
      "loss": 5.1458,
      "mean_token_accuracy": 0.17910135090351104,
      "num_tokens": 63143094.0,
      "step": 28825
    },
    {
      "entropy": 5.3653534889221195,
      "epoch": 2.8612544660579595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004189711399899095,
      "loss": 5.0515,
      "mean_token_accuracy": 0.17987437844276427,
      "num_tokens": 63155159.0,
      "step": 28830
    },
    {
      "entropy": 5.413580226898193,
      "epoch": 2.861750694720127,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004189439097476224,
      "loss": 5.1023,
      "mean_token_accuracy": 0.1773876279592514,
      "num_tokens": 63167132.0,
      "step": 28835
    },
    {
      "entropy": 5.4499814987182615,
      "epoch": 2.8622469233822945,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004189166759358366,
      "loss": 5.1023,
      "mean_token_accuracy": 0.17330080568790435,
      "num_tokens": 63176864.0,
      "step": 28840
    },
    {
      "entropy": 5.48391547203064,
      "epoch": 2.862743152044462,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041888943855522727,
      "loss": 5.0962,
      "mean_token_accuracy": 0.1730571821331978,
      "num_tokens": 63188003.0,
      "step": 28845
    },
    {
      "entropy": 5.409011888504028,
      "epoch": 2.86323938070663,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041886219760646997,
      "loss": 5.0251,
      "mean_token_accuracy": 0.17509617358446122,
      "num_tokens": 63197634.0,
      "step": 28850
    },
    {
      "entropy": 5.378947162628174,
      "epoch": 2.8637356093687973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004188349530902402,
      "loss": 5.0096,
      "mean_token_accuracy": 0.17281535267829895,
      "num_tokens": 63207441.0,
      "step": 28855
    },
    {
      "entropy": 5.391799259185791,
      "epoch": 2.864231838030965,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041880770500721346,
      "loss": 5.0595,
      "mean_token_accuracy": 0.17692960649728776,
      "num_tokens": 63217541.0,
      "step": 28860
    },
    {
      "entropy": 5.400323820114136,
      "epoch": 2.8647280666931323,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004187804533580655,
      "loss": 5.0176,
      "mean_token_accuracy": 0.18104435354471207,
      "num_tokens": 63228391.0,
      "step": 28865
    },
    {
      "entropy": 5.456394481658935,
      "epoch": 2.8652242953552998,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004187531981434721,
      "loss": 5.0568,
      "mean_token_accuracy": 0.17470857948064805,
      "num_tokens": 63238462.0,
      "step": 28870
    },
    {
      "entropy": 5.441836547851563,
      "epoch": 2.8657205240174672,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041872593936410923,
      "loss": 5.1181,
      "mean_token_accuracy": 0.16948676854372025,
      "num_tokens": 63249305.0,
      "step": 28875
    },
    {
      "entropy": 5.492992353439331,
      "epoch": 2.8662167526796347,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041869867702065266,
      "loss": 5.1456,
      "mean_token_accuracy": 0.17188394963741302,
      "num_tokens": 63260832.0,
      "step": 28880
    },
    {
      "entropy": 5.441061019897461,
      "epoch": 2.866712981341802,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004186714111137787,
      "loss": 5.1094,
      "mean_token_accuracy": 0.16379343420267106,
      "num_tokens": 63273689.0,
      "step": 28885
    },
    {
      "entropy": 5.517810249328614,
      "epoch": 2.8672092100039697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004186441416441632,
      "loss": 5.2479,
      "mean_token_accuracy": 0.16502172648906707,
      "num_tokens": 63284912.0,
      "step": 28890
    },
    {
      "entropy": 5.5108729839324955,
      "epoch": 2.867705438666137,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004186168686124825,
      "loss": 5.1408,
      "mean_token_accuracy": 0.16764615923166276,
      "num_tokens": 63296428.0,
      "step": 28895
    },
    {
      "entropy": 5.397865533828735,
      "epoch": 2.8682016673283046,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041858959201941307,
      "loss": 5.0511,
      "mean_token_accuracy": 0.18103874921798707,
      "num_tokens": 63307509.0,
      "step": 28900
    },
    {
      "entropy": 5.481018877029419,
      "epoch": 2.8686978959904725,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041856231186563107,
      "loss": 5.1917,
      "mean_token_accuracy": 0.16337432116270065,
      "num_tokens": 63319194.0,
      "step": 28905
    },
    {
      "entropy": 5.409933996200562,
      "epoch": 2.86919412465264,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041853502815181306,
      "loss": 5.0122,
      "mean_token_accuracy": 0.1773390457034111,
      "num_tokens": 63329122.0,
      "step": 28910
    },
    {
      "entropy": 5.417974472045898,
      "epoch": 2.8696903533148075,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004185077408786357,
      "loss": 5.1048,
      "mean_token_accuracy": 0.17595039904117585,
      "num_tokens": 63340304.0,
      "step": 28915
    },
    {
      "entropy": 5.470542526245117,
      "epoch": 2.870186581976975,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004184804500467756,
      "loss": 5.0894,
      "mean_token_accuracy": 0.17143539488315582,
      "num_tokens": 63351029.0,
      "step": 28920
    },
    {
      "entropy": 5.42136173248291,
      "epoch": 2.8706828106391424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041845315565690945,
      "loss": 5.0373,
      "mean_token_accuracy": 0.1742078885436058,
      "num_tokens": 63361339.0,
      "step": 28925
    },
    {
      "entropy": 5.508452129364014,
      "epoch": 2.87117903930131,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041842585770971417,
      "loss": 5.1086,
      "mean_token_accuracy": 0.17583802342414856,
      "num_tokens": 63373431.0,
      "step": 28930
    },
    {
      "entropy": 5.396961116790772,
      "epoch": 2.871675267963478,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004183985562058667,
      "loss": 5.0609,
      "mean_token_accuracy": 0.17368265390396118,
      "num_tokens": 63384096.0,
      "step": 28935
    },
    {
      "entropy": 5.48833155632019,
      "epoch": 2.8721714966256453,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000418371251146044,
      "loss": 5.1269,
      "mean_token_accuracy": 0.16896763890981675,
      "num_tokens": 63394852.0,
      "step": 28940
    },
    {
      "entropy": 5.447500944137573,
      "epoch": 2.872667725287813,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004183439425309232,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17904231697320938,
      "num_tokens": 63405575.0,
      "step": 28945
    },
    {
      "entropy": 5.466448593139648,
      "epoch": 2.8731639539499803,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004183166303611816,
      "loss": 5.1287,
      "mean_token_accuracy": 0.17205386161804198,
      "num_tokens": 63416677.0,
      "step": 28950
    },
    {
      "entropy": 5.49888768196106,
      "epoch": 2.8736601826121477,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004182893146374963,
      "loss": 5.1974,
      "mean_token_accuracy": 0.1731965348124504,
      "num_tokens": 63428613.0,
      "step": 28955
    },
    {
      "entropy": 5.458173036575317,
      "epoch": 2.874156411274315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041826199536054476,
      "loss": 5.1331,
      "mean_token_accuracy": 0.16693205535411834,
      "num_tokens": 63439321.0,
      "step": 28960
    },
    {
      "entropy": 5.470850276947021,
      "epoch": 2.8746526399364827,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004182346725310045,
      "loss": 5.2075,
      "mean_token_accuracy": 0.16770905256271362,
      "num_tokens": 63451279.0,
      "step": 28965
    },
    {
      "entropy": 5.509257841110229,
      "epoch": 2.87514886859865,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041820734614955295,
      "loss": 5.1125,
      "mean_token_accuracy": 0.17593871355056762,
      "num_tokens": 63460801.0,
      "step": 28970
    },
    {
      "entropy": 5.4568140506744385,
      "epoch": 2.8756450972608176,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004181800162168678,
      "loss": 5.0729,
      "mean_token_accuracy": 0.17266490757465364,
      "num_tokens": 63470232.0,
      "step": 28975
    },
    {
      "entropy": 5.431046628952027,
      "epoch": 2.876141325922985,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004181526827336268,
      "loss": 5.1138,
      "mean_token_accuracy": 0.17438317388296126,
      "num_tokens": 63480247.0,
      "step": 28980
    },
    {
      "entropy": 5.444803714752197,
      "epoch": 2.8766375545851526,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004181253457005077,
      "loss": 5.0984,
      "mean_token_accuracy": 0.16665793359279632,
      "num_tokens": 63492104.0,
      "step": 28985
    },
    {
      "entropy": 5.4555436134338375,
      "epoch": 2.8771337832473205,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041809800511818853,
      "loss": 5.0948,
      "mean_token_accuracy": 0.17539948523044585,
      "num_tokens": 63502784.0,
      "step": 28990
    },
    {
      "entropy": 5.498994064331055,
      "epoch": 2.877630011909488,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041807066098734704,
      "loss": 5.178,
      "mean_token_accuracy": 0.17159930020570754,
      "num_tokens": 63514150.0,
      "step": 28995
    },
    {
      "entropy": 5.490581130981445,
      "epoch": 2.8781262405716554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004180433133086615,
      "loss": 5.147,
      "mean_token_accuracy": 0.17508909702301026,
      "num_tokens": 63526682.0,
      "step": 29000
    },
    {
      "entropy": 5.441305160522461,
      "epoch": 2.878622469233823,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004180159620828101,
      "loss": 5.1315,
      "mean_token_accuracy": 0.1685812920331955,
      "num_tokens": 63537448.0,
      "step": 29005
    },
    {
      "entropy": 5.426985263824463,
      "epoch": 2.8791186978959904,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004179886073104709,
      "loss": 5.0532,
      "mean_token_accuracy": 0.1735990047454834,
      "num_tokens": 63547489.0,
      "step": 29010
    },
    {
      "entropy": 5.450059795379639,
      "epoch": 2.879614926558158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004179612489923224,
      "loss": 5.0581,
      "mean_token_accuracy": 0.17823182940483093,
      "num_tokens": 63558324.0,
      "step": 29015
    },
    {
      "entropy": 5.463821601867676,
      "epoch": 2.880111155220326,
      "grad_norm": 1.0,
      "learning_rate": 0.000417933887129043,
      "loss": 5.0973,
      "mean_token_accuracy": 0.17169996947050095,
      "num_tokens": 63568630.0,
      "step": 29020
    },
    {
      "entropy": 5.350571584701538,
      "epoch": 2.8806073838824933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004179065217213111,
      "loss": 5.0261,
      "mean_token_accuracy": 0.17466460317373275,
      "num_tokens": 63578068.0,
      "step": 29025
    },
    {
      "entropy": 5.418845224380493,
      "epoch": 2.8811036125446607,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041787915276980556,
      "loss": 5.1423,
      "mean_token_accuracy": 0.1671348035335541,
      "num_tokens": 63589103.0,
      "step": 29030
    },
    {
      "entropy": 5.455748462677002,
      "epoch": 2.881599841206828,
      "grad_norm": 1.0,
      "learning_rate": 0.00041785178027520486,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1758216455578804,
      "num_tokens": 63600305.0,
      "step": 29035
    },
    {
      "entropy": 5.520303440093994,
      "epoch": 2.8820960698689957,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041782440423818775,
      "loss": 5.1634,
      "mean_token_accuracy": 0.1712069481611252,
      "num_tokens": 63611910.0,
      "step": 29040
    },
    {
      "entropy": 5.4806914806365965,
      "epoch": 2.882592298531163,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004177970246594332,
      "loss": 5.0355,
      "mean_token_accuracy": 0.1759425163269043,
      "num_tokens": 63622674.0,
      "step": 29045
    },
    {
      "entropy": 5.493885183334351,
      "epoch": 2.8830885271933306,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041776964153962017,
      "loss": 5.1534,
      "mean_token_accuracy": 0.1647355243563652,
      "num_tokens": 63633274.0,
      "step": 29050
    },
    {
      "entropy": 5.466422748565674,
      "epoch": 2.883584755855498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041774225487942765,
      "loss": 5.1046,
      "mean_token_accuracy": 0.17656143009662628,
      "num_tokens": 63644738.0,
      "step": 29055
    },
    {
      "entropy": 5.44983925819397,
      "epoch": 2.8840809845176656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004177148646795348,
      "loss": 5.1252,
      "mean_token_accuracy": 0.17150243371725082,
      "num_tokens": 63655468.0,
      "step": 29060
    },
    {
      "entropy": 5.417666578292847,
      "epoch": 2.884577213179833,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041768747094062084,
      "loss": 4.9935,
      "mean_token_accuracy": 0.174552258849144,
      "num_tokens": 63667675.0,
      "step": 29065
    },
    {
      "entropy": 5.432532215118409,
      "epoch": 2.8850734418420005,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041766007366336494,
      "loss": 5.1384,
      "mean_token_accuracy": 0.16940570622682571,
      "num_tokens": 63679251.0,
      "step": 29070
    },
    {
      "entropy": 5.413814210891724,
      "epoch": 2.8855696705041685,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041763267284844675,
      "loss": 5.071,
      "mean_token_accuracy": 0.17700104862451554,
      "num_tokens": 63690809.0,
      "step": 29075
    },
    {
      "entropy": 5.4655519962310795,
      "epoch": 2.886065899166336,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004176052684965456,
      "loss": 5.0494,
      "mean_token_accuracy": 0.17646438032388687,
      "num_tokens": 63702434.0,
      "step": 29080
    },
    {
      "entropy": 5.43359694480896,
      "epoch": 2.8865621278285034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004175778606083411,
      "loss": 5.1628,
      "mean_token_accuracy": 0.17094054520130159,
      "num_tokens": 63713173.0,
      "step": 29085
    },
    {
      "entropy": 5.438838005065918,
      "epoch": 2.887058356490671,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004175504491845128,
      "loss": 5.1258,
      "mean_token_accuracy": 0.16919476240873338,
      "num_tokens": 63723956.0,
      "step": 29090
    },
    {
      "entropy": 5.437114000320435,
      "epoch": 2.8875545851528384,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041752303422574067,
      "loss": 5.1396,
      "mean_token_accuracy": 0.16954636722803115,
      "num_tokens": 63735280.0,
      "step": 29095
    },
    {
      "entropy": 5.533269214630127,
      "epoch": 2.888050813815006,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004174956157327043,
      "loss": 5.1272,
      "mean_token_accuracy": 0.16601199209690093,
      "num_tokens": 63746075.0,
      "step": 29100
    },
    {
      "entropy": 5.517338466644287,
      "epoch": 2.8885470424771738,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004174681937060837,
      "loss": 5.112,
      "mean_token_accuracy": 0.177194444835186,
      "num_tokens": 63758152.0,
      "step": 29105
    },
    {
      "entropy": 5.421280145645142,
      "epoch": 2.8890432711393412,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041744076814655893,
      "loss": 5.1113,
      "mean_token_accuracy": 0.1727056547999382,
      "num_tokens": 63768855.0,
      "step": 29110
    },
    {
      "entropy": 5.4417102336883545,
      "epoch": 2.8895394998015087,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041741333905481,
      "loss": 5.1315,
      "mean_token_accuracy": 0.16387836933135985,
      "num_tokens": 63780258.0,
      "step": 29115
    },
    {
      "entropy": 5.515068483352661,
      "epoch": 2.890035728463676,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004173859064315172,
      "loss": 5.1772,
      "mean_token_accuracy": 0.1710044339299202,
      "num_tokens": 63791959.0,
      "step": 29120
    },
    {
      "entropy": 5.482991933822632,
      "epoch": 2.8905319571258437,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041735847027736067,
      "loss": 5.1337,
      "mean_token_accuracy": 0.16840649545192718,
      "num_tokens": 63802859.0,
      "step": 29125
    },
    {
      "entropy": 5.4269389629364015,
      "epoch": 2.891028185788011,
      "grad_norm": 1.0,
      "learning_rate": 0.0004173310305930209,
      "loss": 5.0663,
      "mean_token_accuracy": 0.16869206875562667,
      "num_tokens": 63814275.0,
      "step": 29130
    },
    {
      "entropy": 5.548747873306274,
      "epoch": 2.8915244144501786,
      "grad_norm": 1.125,
      "learning_rate": 0.00041730358737917823,
      "loss": 5.118,
      "mean_token_accuracy": 0.17366829067468642,
      "num_tokens": 63824633.0,
      "step": 29135
    },
    {
      "entropy": 5.494446420669556,
      "epoch": 2.892020643112346,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041727614063651313,
      "loss": 5.1012,
      "mean_token_accuracy": 0.1707159847021103,
      "num_tokens": 63836478.0,
      "step": 29140
    },
    {
      "entropy": 5.485472774505615,
      "epoch": 2.8925168717745136,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004172486903657065,
      "loss": 5.1495,
      "mean_token_accuracy": 0.16634923666715623,
      "num_tokens": 63847917.0,
      "step": 29145
    },
    {
      "entropy": 5.448352193832397,
      "epoch": 2.893013100436681,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004172212365674388,
      "loss": 5.1328,
      "mean_token_accuracy": 0.17571888417005538,
      "num_tokens": 63857258.0,
      "step": 29150
    },
    {
      "entropy": 5.431738567352295,
      "epoch": 2.8935093290988485,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041719377924239083,
      "loss": 5.1744,
      "mean_token_accuracy": 0.16880376935005187,
      "num_tokens": 63868627.0,
      "step": 29155
    },
    {
      "entropy": 5.495714950561523,
      "epoch": 2.8940055577610164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041716631839124356,
      "loss": 5.1532,
      "mean_token_accuracy": 0.17306388169527054,
      "num_tokens": 63879954.0,
      "step": 29160
    },
    {
      "entropy": 5.6038415908813475,
      "epoch": 2.894501786423184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000417138854014678,
      "loss": 5.2326,
      "mean_token_accuracy": 0.17197195887565614,
      "num_tokens": 63891147.0,
      "step": 29165
    },
    {
      "entropy": 5.439862155914307,
      "epoch": 2.8949980150853514,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004171113861133751,
      "loss": 5.075,
      "mean_token_accuracy": 0.17069433033466339,
      "num_tokens": 63902380.0,
      "step": 29170
    },
    {
      "entropy": 5.4327545166015625,
      "epoch": 2.895494243747519,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004170839146880161,
      "loss": 5.124,
      "mean_token_accuracy": 0.17437343299388885,
      "num_tokens": 63912557.0,
      "step": 29175
    },
    {
      "entropy": 5.4847711563110355,
      "epoch": 2.8959904724096863,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004170564397392821,
      "loss": 5.158,
      "mean_token_accuracy": 0.172648523747921,
      "num_tokens": 63923959.0,
      "step": 29180
    },
    {
      "entropy": 5.470569467544555,
      "epoch": 2.896486701071854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004170289612678545,
      "loss": 5.071,
      "mean_token_accuracy": 0.1761628657579422,
      "num_tokens": 63935600.0,
      "step": 29185
    },
    {
      "entropy": 5.468276262283325,
      "epoch": 2.8969829297340213,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041700147927441477,
      "loss": 5.2199,
      "mean_token_accuracy": 0.1644713744521141,
      "num_tokens": 63946371.0,
      "step": 29190
    },
    {
      "entropy": 5.465210628509522,
      "epoch": 2.897479158396189,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041697399375964435,
      "loss": 5.1725,
      "mean_token_accuracy": 0.16633804142475128,
      "num_tokens": 63956019.0,
      "step": 29195
    },
    {
      "entropy": 5.454396152496338,
      "epoch": 2.8979753870583567,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041694650472422467,
      "loss": 5.1043,
      "mean_token_accuracy": 0.16748059839010238,
      "num_tokens": 63967330.0,
      "step": 29200
    },
    {
      "entropy": 5.474564027786255,
      "epoch": 2.898471615720524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041691901216883763,
      "loss": 5.0901,
      "mean_token_accuracy": 0.17477495223283768,
      "num_tokens": 63978952.0,
      "step": 29205
    },
    {
      "entropy": 5.464394187927246,
      "epoch": 2.8989678443826916,
      "grad_norm": 1.078125,
      "learning_rate": 0.000416891516094165,
      "loss": 5.1229,
      "mean_token_accuracy": 0.16851274222135543,
      "num_tokens": 63989913.0,
      "step": 29210
    },
    {
      "entropy": 5.381395435333252,
      "epoch": 2.899464073044859,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004168640165008884,
      "loss": 5.0326,
      "mean_token_accuracy": 0.179513081908226,
      "num_tokens": 64000033.0,
      "step": 29215
    },
    {
      "entropy": 5.417944812774659,
      "epoch": 2.8999603017070266,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041683651338968994,
      "loss": 5.0964,
      "mean_token_accuracy": 0.168350251019001,
      "num_tokens": 64010746.0,
      "step": 29220
    },
    {
      "entropy": 5.4479594230651855,
      "epoch": 2.900456530369194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041680900676125147,
      "loss": 5.0678,
      "mean_token_accuracy": 0.173532535135746,
      "num_tokens": 64021179.0,
      "step": 29225
    },
    {
      "entropy": 5.50103063583374,
      "epoch": 2.9009527590313615,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041678149661625525,
      "loss": 5.1332,
      "mean_token_accuracy": 0.17033475935459136,
      "num_tokens": 64032953.0,
      "step": 29230
    },
    {
      "entropy": 5.421468687057495,
      "epoch": 2.901448987693529,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004167539829553835,
      "loss": 4.9802,
      "mean_token_accuracy": 0.18539279699325562,
      "num_tokens": 64042548.0,
      "step": 29235
    },
    {
      "entropy": 5.397246170043945,
      "epoch": 2.9019452163556965,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041672646577931835,
      "loss": 5.0275,
      "mean_token_accuracy": 0.17372252196073532,
      "num_tokens": 64053254.0,
      "step": 29240
    },
    {
      "entropy": 5.3959388256073,
      "epoch": 2.9024414450178644,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041669894508874225,
      "loss": 5.0863,
      "mean_token_accuracy": 0.1680764377117157,
      "num_tokens": 64063307.0,
      "step": 29245
    },
    {
      "entropy": 5.471155452728271,
      "epoch": 2.902937673680032,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041667142088433766,
      "loss": 5.0809,
      "mean_token_accuracy": 0.18004562109708785,
      "num_tokens": 64074274.0,
      "step": 29250
    },
    {
      "entropy": 5.333494377136231,
      "epoch": 2.9034339023421993,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041664389316678713,
      "loss": 4.9503,
      "mean_token_accuracy": 0.1957260012626648,
      "num_tokens": 64084243.0,
      "step": 29255
    },
    {
      "entropy": 5.420783042907715,
      "epoch": 2.903930131004367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004166163619367733,
      "loss": 5.0705,
      "mean_token_accuracy": 0.17850443422794343,
      "num_tokens": 64095899.0,
      "step": 29260
    },
    {
      "entropy": 5.4241101264953615,
      "epoch": 2.9044263596665343,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041658882719497876,
      "loss": 4.9913,
      "mean_token_accuracy": 0.17821164280176163,
      "num_tokens": 64107408.0,
      "step": 29265
    },
    {
      "entropy": 5.49098916053772,
      "epoch": 2.9049225883287018,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004165612889420864,
      "loss": 5.1501,
      "mean_token_accuracy": 0.16673435866832734,
      "num_tokens": 64118935.0,
      "step": 29270
    },
    {
      "entropy": 5.468742609024048,
      "epoch": 2.9054188169908692,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004165337471787792,
      "loss": 5.1203,
      "mean_token_accuracy": 0.17524151355028153,
      "num_tokens": 64129085.0,
      "step": 29275
    },
    {
      "entropy": 5.395329141616822,
      "epoch": 2.905915045653037,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041650620190574004,
      "loss": 5.0089,
      "mean_token_accuracy": 0.1789089947938919,
      "num_tokens": 64139605.0,
      "step": 29280
    },
    {
      "entropy": 5.494961881637574,
      "epoch": 2.9064112743152046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041647865312365186,
      "loss": 5.1516,
      "mean_token_accuracy": 0.16849828660488128,
      "num_tokens": 64151171.0,
      "step": 29285
    },
    {
      "entropy": 5.518099927902222,
      "epoch": 2.906907502977372,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041645110083319805,
      "loss": 5.1302,
      "mean_token_accuracy": 0.1641354590654373,
      "num_tokens": 64162417.0,
      "step": 29290
    },
    {
      "entropy": 5.502341651916504,
      "epoch": 2.9074037316395396,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041642354503506175,
      "loss": 5.0982,
      "mean_token_accuracy": 0.17002911418676375,
      "num_tokens": 64172755.0,
      "step": 29295
    },
    {
      "entropy": 5.44690203666687,
      "epoch": 2.907899960301707,
      "grad_norm": 1.0,
      "learning_rate": 0.00041639598572992625,
      "loss": 5.0852,
      "mean_token_accuracy": 0.17460645586252213,
      "num_tokens": 64184270.0,
      "step": 29300
    },
    {
      "entropy": 5.508746242523193,
      "epoch": 2.9083961889638745,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041636842291847495,
      "loss": 5.167,
      "mean_token_accuracy": 0.16518789380788804,
      "num_tokens": 64195400.0,
      "step": 29305
    },
    {
      "entropy": 5.50805253982544,
      "epoch": 2.908892417626042,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041634085660139145,
      "loss": 5.1324,
      "mean_token_accuracy": 0.17262187153100966,
      "num_tokens": 64205101.0,
      "step": 29310
    },
    {
      "entropy": 5.405797052383423,
      "epoch": 2.9093886462882095,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004163132867793592,
      "loss": 4.9988,
      "mean_token_accuracy": 0.17412517070770264,
      "num_tokens": 64216047.0,
      "step": 29315
    },
    {
      "entropy": 5.383992147445679,
      "epoch": 2.909884874950377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000416285713453062,
      "loss": 5.0946,
      "mean_token_accuracy": 0.17861682921648026,
      "num_tokens": 64226195.0,
      "step": 29320
    },
    {
      "entropy": 5.449498653411865,
      "epoch": 2.9103811036125444,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004162581366231835,
      "loss": 5.0878,
      "mean_token_accuracy": 0.18057252764701842,
      "num_tokens": 64236405.0,
      "step": 29325
    },
    {
      "entropy": 5.398318815231323,
      "epoch": 2.9108773322747123,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041623055629040764,
      "loss": 5.0394,
      "mean_token_accuracy": 0.17789310663938523,
      "num_tokens": 64246989.0,
      "step": 29330
    },
    {
      "entropy": 5.417199325561524,
      "epoch": 2.91137356093688,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004162029724554183,
      "loss": 5.0749,
      "mean_token_accuracy": 0.17317272424697877,
      "num_tokens": 64257928.0,
      "step": 29335
    },
    {
      "entropy": 5.4394909858703615,
      "epoch": 2.9118697895990473,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041617538511889946,
      "loss": 5.1119,
      "mean_token_accuracy": 0.17389601469039917,
      "num_tokens": 64268985.0,
      "step": 29340
    },
    {
      "entropy": 5.6029150009155275,
      "epoch": 2.9123660182612148,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004161477942815353,
      "loss": 5.2494,
      "mean_token_accuracy": 0.16612409055233002,
      "num_tokens": 64279907.0,
      "step": 29345
    },
    {
      "entropy": 5.516003370285034,
      "epoch": 2.9128622469233822,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041612019994401,
      "loss": 5.1261,
      "mean_token_accuracy": 0.16610352843999862,
      "num_tokens": 64291496.0,
      "step": 29350
    },
    {
      "entropy": 5.420829010009766,
      "epoch": 2.9133584755855497,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004160926021070078,
      "loss": 5.0796,
      "mean_token_accuracy": 0.1812197744846344,
      "num_tokens": 64302160.0,
      "step": 29355
    },
    {
      "entropy": 5.494960021972656,
      "epoch": 2.913854704247717,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004160650007712132,
      "loss": 5.146,
      "mean_token_accuracy": 0.17095946073532103,
      "num_tokens": 64313671.0,
      "step": 29360
    },
    {
      "entropy": 5.471186780929566,
      "epoch": 2.914350932909885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041603739593731046,
      "loss": 5.1308,
      "mean_token_accuracy": 0.1685690090060234,
      "num_tokens": 64325376.0,
      "step": 29365
    },
    {
      "entropy": 5.497774410247803,
      "epoch": 2.9148471615720526,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004160097876059842,
      "loss": 5.1149,
      "mean_token_accuracy": 0.17296502441167833,
      "num_tokens": 64336048.0,
      "step": 29370
    },
    {
      "entropy": 5.489615869522095,
      "epoch": 2.91534339023422,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004159821757779191,
      "loss": 5.1556,
      "mean_token_accuracy": 0.1652045100927353,
      "num_tokens": 64346833.0,
      "step": 29375
    },
    {
      "entropy": 5.446487045288086,
      "epoch": 2.9158396188963875,
      "grad_norm": 1.28125,
      "learning_rate": 0.00041595456045379975,
      "loss": 5.1018,
      "mean_token_accuracy": 0.16907763928174974,
      "num_tokens": 64359656.0,
      "step": 29380
    },
    {
      "entropy": 5.423136806488037,
      "epoch": 2.916335847558555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004159269416343111,
      "loss": 5.0371,
      "mean_token_accuracy": 0.1722923368215561,
      "num_tokens": 64371130.0,
      "step": 29385
    },
    {
      "entropy": 5.407291746139526,
      "epoch": 2.9168320762207225,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000415899319320138,
      "loss": 5.0926,
      "mean_token_accuracy": 0.17806855589151382,
      "num_tokens": 64381826.0,
      "step": 29390
    },
    {
      "entropy": 5.513131904602051,
      "epoch": 2.91732830488289,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041587169351196536,
      "loss": 5.2034,
      "mean_token_accuracy": 0.16620451360940933,
      "num_tokens": 64392725.0,
      "step": 29395
    },
    {
      "entropy": 5.497659397125244,
      "epoch": 2.9178245335450574,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041584406421047825,
      "loss": 5.0853,
      "mean_token_accuracy": 0.1736529663205147,
      "num_tokens": 64402587.0,
      "step": 29400
    },
    {
      "entropy": 5.465218925476075,
      "epoch": 2.918320762207225,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041581643141636184,
      "loss": 5.1522,
      "mean_token_accuracy": 0.17277072072029115,
      "num_tokens": 64414692.0,
      "step": 29405
    },
    {
      "entropy": 5.436130523681641,
      "epoch": 2.9188169908693924,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004157887951303014,
      "loss": 5.1545,
      "mean_token_accuracy": 0.17153491079807281,
      "num_tokens": 64427278.0,
      "step": 29410
    },
    {
      "entropy": 5.44992241859436,
      "epoch": 2.91931321953156,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004157611553529822,
      "loss": 5.0417,
      "mean_token_accuracy": 0.1733646050095558,
      "num_tokens": 64438221.0,
      "step": 29415
    },
    {
      "entropy": 5.460375165939331,
      "epoch": 2.919809448193728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041573351208508966,
      "loss": 5.2015,
      "mean_token_accuracy": 0.1693061336874962,
      "num_tokens": 64450139.0,
      "step": 29420
    },
    {
      "entropy": 5.458996295928955,
      "epoch": 2.9203056768558953,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041570586532730926,
      "loss": 5.0948,
      "mean_token_accuracy": 0.17151543349027634,
      "num_tokens": 64461523.0,
      "step": 29425
    },
    {
      "entropy": 5.464106225967408,
      "epoch": 2.9208019055180627,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004156782150803266,
      "loss": 5.1698,
      "mean_token_accuracy": 0.16327131390571595,
      "num_tokens": 64474206.0,
      "step": 29430
    },
    {
      "entropy": 5.467969226837158,
      "epoch": 2.92129813418023,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041565056134482733,
      "loss": 5.1231,
      "mean_token_accuracy": 0.16781802773475646,
      "num_tokens": 64485585.0,
      "step": 29435
    },
    {
      "entropy": 5.562425661087036,
      "epoch": 2.9217943628423977,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004156229041214972,
      "loss": 5.1604,
      "mean_token_accuracy": 0.170232655107975,
      "num_tokens": 64497237.0,
      "step": 29440
    },
    {
      "entropy": 5.4473906517028805,
      "epoch": 2.922290591504565,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004155952434110221,
      "loss": 5.1062,
      "mean_token_accuracy": 0.17307801991701127,
      "num_tokens": 64508235.0,
      "step": 29445
    },
    {
      "entropy": 5.4873692989349365,
      "epoch": 2.922786820166733,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004155675792140879,
      "loss": 5.1465,
      "mean_token_accuracy": 0.17237553894519805,
      "num_tokens": 64520345.0,
      "step": 29450
    },
    {
      "entropy": 5.502862882614136,
      "epoch": 2.9232830488289006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004155399115313806,
      "loss": 5.1376,
      "mean_token_accuracy": 0.1706508681178093,
      "num_tokens": 64531428.0,
      "step": 29455
    },
    {
      "entropy": 5.516918277740478,
      "epoch": 2.923779277491068,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041551224036358644,
      "loss": 5.147,
      "mean_token_accuracy": 0.16842904090881347,
      "num_tokens": 64541489.0,
      "step": 29460
    },
    {
      "entropy": 5.380442190170288,
      "epoch": 2.9242755061532355,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041548456571139145,
      "loss": 5.0292,
      "mean_token_accuracy": 0.18139352053403854,
      "num_tokens": 64553026.0,
      "step": 29465
    },
    {
      "entropy": 5.494320058822632,
      "epoch": 2.924771734815403,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041545688757548184,
      "loss": 5.216,
      "mean_token_accuracy": 0.16686538159847258,
      "num_tokens": 64565988.0,
      "step": 29470
    },
    {
      "entropy": 5.488052797317505,
      "epoch": 2.9252679634775705,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041542920595654423,
      "loss": 5.0884,
      "mean_token_accuracy": 0.1727609559893608,
      "num_tokens": 64578098.0,
      "step": 29475
    },
    {
      "entropy": 5.419833040237426,
      "epoch": 2.925764192139738,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041540152085526476,
      "loss": 5.0027,
      "mean_token_accuracy": 0.17195311784744263,
      "num_tokens": 64588931.0,
      "step": 29480
    },
    {
      "entropy": 5.356493520736694,
      "epoch": 2.9262604208019054,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041537383227233023,
      "loss": 5.0113,
      "mean_token_accuracy": 0.18751323223114014,
      "num_tokens": 64599636.0,
      "step": 29485
    },
    {
      "entropy": 5.429446125030518,
      "epoch": 2.926756649464073,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004153461402084271,
      "loss": 5.1124,
      "mean_token_accuracy": 0.17077947407960892,
      "num_tokens": 64609355.0,
      "step": 29490
    },
    {
      "entropy": 5.4042003631591795,
      "epoch": 2.9272528781262404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004153184446642421,
      "loss": 5.0689,
      "mean_token_accuracy": 0.1789991706609726,
      "num_tokens": 64619969.0,
      "step": 29495
    },
    {
      "entropy": 5.450017309188842,
      "epoch": 2.927749106788408,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041529074564046204,
      "loss": 5.0457,
      "mean_token_accuracy": 0.18524824678897858,
      "num_tokens": 64629538.0,
      "step": 29500
    },
    {
      "entropy": 5.49266619682312,
      "epoch": 2.9282453354505757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004152630431377738,
      "loss": 5.1002,
      "mean_token_accuracy": 0.1774833008646965,
      "num_tokens": 64640255.0,
      "step": 29505
    },
    {
      "entropy": 5.357032299041748,
      "epoch": 2.9287415641127432,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004152353371568643,
      "loss": 4.9969,
      "mean_token_accuracy": 0.17939275056123732,
      "num_tokens": 64650051.0,
      "step": 29510
    },
    {
      "entropy": 5.391980838775635,
      "epoch": 2.9292377927749107,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041520762769842065,
      "loss": 5.0779,
      "mean_token_accuracy": 0.17721532583236693,
      "num_tokens": 64660575.0,
      "step": 29515
    },
    {
      "entropy": 5.531826448440552,
      "epoch": 2.929734021437078,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041517991476312994,
      "loss": 5.1422,
      "mean_token_accuracy": 0.17248581498861312,
      "num_tokens": 64671486.0,
      "step": 29520
    },
    {
      "entropy": 5.488949346542358,
      "epoch": 2.9302302500992456,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004151521983516794,
      "loss": 5.0254,
      "mean_token_accuracy": 0.17784492671489716,
      "num_tokens": 64682315.0,
      "step": 29525
    },
    {
      "entropy": 5.448695516586303,
      "epoch": 2.930726478761413,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004151244784647563,
      "loss": 5.1598,
      "mean_token_accuracy": 0.17516190856695174,
      "num_tokens": 64693940.0,
      "step": 29530
    },
    {
      "entropy": 5.458058834075928,
      "epoch": 2.931222707423581,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004150967551030482,
      "loss": 5.0699,
      "mean_token_accuracy": 0.17666810303926467,
      "num_tokens": 64703347.0,
      "step": 29535
    },
    {
      "entropy": 5.4190412044525145,
      "epoch": 2.9317189360857485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041506902826724234,
      "loss": 5.132,
      "mean_token_accuracy": 0.16858236938714982,
      "num_tokens": 64714115.0,
      "step": 29540
    },
    {
      "entropy": 5.398310804367066,
      "epoch": 2.932215164747916,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004150412979580264,
      "loss": 5.02,
      "mean_token_accuracy": 0.1822185695171356,
      "num_tokens": 64724888.0,
      "step": 29545
    },
    {
      "entropy": 5.511121463775635,
      "epoch": 2.9327113934100835,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041501356417608816,
      "loss": 5.1275,
      "mean_token_accuracy": 0.17167562395334243,
      "num_tokens": 64735626.0,
      "step": 29550
    },
    {
      "entropy": 5.462503099441529,
      "epoch": 2.933207622072251,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041498582692211513,
      "loss": 5.0706,
      "mean_token_accuracy": 0.1749953120946884,
      "num_tokens": 64747807.0,
      "step": 29555
    },
    {
      "entropy": 5.404977512359619,
      "epoch": 2.9337038507344184,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041495808619679525,
      "loss": 5.0361,
      "mean_token_accuracy": 0.17745941132307053,
      "num_tokens": 64756845.0,
      "step": 29560
    },
    {
      "entropy": 5.419396829605103,
      "epoch": 2.934200079396586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041493034200081643,
      "loss": 5.1246,
      "mean_token_accuracy": 0.17180739790201188,
      "num_tokens": 64767316.0,
      "step": 29565
    },
    {
      "entropy": 5.502083492279053,
      "epoch": 2.9346963080587534,
      "grad_norm": 1.125,
      "learning_rate": 0.0004149025943348666,
      "loss": 5.0526,
      "mean_token_accuracy": 0.1808272197842598,
      "num_tokens": 64778329.0,
      "step": 29570
    },
    {
      "entropy": 5.458843231201172,
      "epoch": 2.935192536720921,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041487484319963395,
      "loss": 5.1043,
      "mean_token_accuracy": 0.17504890263080597,
      "num_tokens": 64789215.0,
      "step": 29575
    },
    {
      "entropy": 5.393897771835327,
      "epoch": 2.9356887653830883,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004148470885958066,
      "loss": 5.0637,
      "mean_token_accuracy": 0.17487114667892456,
      "num_tokens": 64801878.0,
      "step": 29580
    },
    {
      "entropy": 5.3883425235748295,
      "epoch": 2.936184994045256,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004148193305240727,
      "loss": 4.9923,
      "mean_token_accuracy": 0.17460866421461105,
      "num_tokens": 64813442.0,
      "step": 29585
    },
    {
      "entropy": 5.438806200027466,
      "epoch": 2.9366812227074237,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041479156898512076,
      "loss": 5.049,
      "mean_token_accuracy": 0.1737948939204216,
      "num_tokens": 64824095.0,
      "step": 29590
    },
    {
      "entropy": 5.486423063278198,
      "epoch": 2.937177451369591,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004147638039796391,
      "loss": 5.0844,
      "mean_token_accuracy": 0.1769652396440506,
      "num_tokens": 64833101.0,
      "step": 29595
    },
    {
      "entropy": 5.361846160888672,
      "epoch": 2.9376736800317587,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041473603550831636,
      "loss": 5.0129,
      "mean_token_accuracy": 0.17838729918003082,
      "num_tokens": 64844315.0,
      "step": 29600
    },
    {
      "entropy": 5.447410678863525,
      "epoch": 2.938169908693926,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041470826357184096,
      "loss": 5.0963,
      "mean_token_accuracy": 0.16919099241495134,
      "num_tokens": 64855090.0,
      "step": 29605
    },
    {
      "entropy": 5.400208282470703,
      "epoch": 2.9386661373560936,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004146804881709017,
      "loss": 5.0568,
      "mean_token_accuracy": 0.17264586985111235,
      "num_tokens": 64865093.0,
      "step": 29610
    },
    {
      "entropy": 5.536527681350708,
      "epoch": 2.939162366018261,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004146527093061873,
      "loss": 5.1727,
      "mean_token_accuracy": 0.163956618309021,
      "num_tokens": 64876250.0,
      "step": 29615
    },
    {
      "entropy": 5.497835493087768,
      "epoch": 2.939658594680429,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004146249269783867,
      "loss": 5.0911,
      "mean_token_accuracy": 0.17123040407896042,
      "num_tokens": 64886339.0,
      "step": 29620
    },
    {
      "entropy": 5.482382297515869,
      "epoch": 2.9401548233425965,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004145971411881886,
      "loss": 5.2057,
      "mean_token_accuracy": 0.16451116651296616,
      "num_tokens": 64899473.0,
      "step": 29625
    },
    {
      "entropy": 5.492380285263062,
      "epoch": 2.940651052004764,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041456935193628236,
      "loss": 5.103,
      "mean_token_accuracy": 0.1716645449399948,
      "num_tokens": 64909919.0,
      "step": 29630
    },
    {
      "entropy": 5.497350692749023,
      "epoch": 2.9411472806669314,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004145415592233569,
      "loss": 5.124,
      "mean_token_accuracy": 0.1718156635761261,
      "num_tokens": 64921069.0,
      "step": 29635
    },
    {
      "entropy": 5.37377896308899,
      "epoch": 2.941643509329099,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041451376305010144,
      "loss": 5.0618,
      "mean_token_accuracy": 0.17994342893362045,
      "num_tokens": 64932173.0,
      "step": 29640
    },
    {
      "entropy": 5.4337443828582765,
      "epoch": 2.9421397379912664,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004144859634172053,
      "loss": 4.9982,
      "mean_token_accuracy": 0.1798928827047348,
      "num_tokens": 64941650.0,
      "step": 29645
    },
    {
      "entropy": 5.404904747009278,
      "epoch": 2.942635966653434,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041445816032535784,
      "loss": 5.0432,
      "mean_token_accuracy": 0.1781666785478592,
      "num_tokens": 64952520.0,
      "step": 29650
    },
    {
      "entropy": 5.443080472946167,
      "epoch": 2.9431321953156013,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041443035377524844,
      "loss": 5.2044,
      "mean_token_accuracy": 0.16727815121412276,
      "num_tokens": 64965030.0,
      "step": 29655
    },
    {
      "entropy": 5.41821436882019,
      "epoch": 2.943628423977769,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004144025437675668,
      "loss": 5.0372,
      "mean_token_accuracy": 0.17888158410787583,
      "num_tokens": 64976260.0,
      "step": 29660
    },
    {
      "entropy": 5.520702934265136,
      "epoch": 2.9441246526399363,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004143747303030025,
      "loss": 5.1807,
      "mean_token_accuracy": 0.1747586354613304,
      "num_tokens": 64988049.0,
      "step": 29665
    },
    {
      "entropy": 5.438289976119995,
      "epoch": 2.9446208813021038,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041434691338224506,
      "loss": 5.0589,
      "mean_token_accuracy": 0.17795372754335403,
      "num_tokens": 64998506.0,
      "step": 29670
    },
    {
      "entropy": 5.390826416015625,
      "epoch": 2.9451171099642717,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004143190930059846,
      "loss": 5.0567,
      "mean_token_accuracy": 0.17872685939073563,
      "num_tokens": 65010533.0,
      "step": 29675
    },
    {
      "entropy": 5.517217826843262,
      "epoch": 2.945613338626439,
      "grad_norm": 1.125,
      "learning_rate": 0.0004142912691749107,
      "loss": 5.1196,
      "mean_token_accuracy": 0.17312434762716294,
      "num_tokens": 65021238.0,
      "step": 29680
    },
    {
      "entropy": 5.52191915512085,
      "epoch": 2.9461095672886066,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004142634418897135,
      "loss": 5.1833,
      "mean_token_accuracy": 0.1650900050997734,
      "num_tokens": 65033251.0,
      "step": 29685
    },
    {
      "entropy": 5.49473762512207,
      "epoch": 2.946605795950774,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004142356111510831,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17030481696128846,
      "num_tokens": 65043845.0,
      "step": 29690
    },
    {
      "entropy": 5.40798511505127,
      "epoch": 2.9471020246129416,
      "grad_norm": 1.390625,
      "learning_rate": 0.00041420777695970945,
      "loss": 5.0713,
      "mean_token_accuracy": 0.18089466094970702,
      "num_tokens": 65055395.0,
      "step": 29695
    },
    {
      "entropy": 5.46071720123291,
      "epoch": 2.947598253275109,
      "grad_norm": 1.046875,
      "learning_rate": 0.000414179939316283,
      "loss": 5.104,
      "mean_token_accuracy": 0.17241654694080352,
      "num_tokens": 65064931.0,
      "step": 29700
    },
    {
      "entropy": 5.454668951034546,
      "epoch": 2.948094481937277,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041415209822149397,
      "loss": 5.07,
      "mean_token_accuracy": 0.1773312970995903,
      "num_tokens": 65074678.0,
      "step": 29705
    },
    {
      "entropy": 5.449082088470459,
      "epoch": 2.9485907105994444,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004141242536760327,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17905500531196594,
      "num_tokens": 65084562.0,
      "step": 29710
    },
    {
      "entropy": 5.478402519226075,
      "epoch": 2.949086939261612,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004140964056805897,
      "loss": 5.0913,
      "mean_token_accuracy": 0.17543769180774688,
      "num_tokens": 65095924.0,
      "step": 29715
    },
    {
      "entropy": 5.489739990234375,
      "epoch": 2.9495831679237794,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004140685542358556,
      "loss": 5.1187,
      "mean_token_accuracy": 0.18353429287672043,
      "num_tokens": 65106802.0,
      "step": 29720
    },
    {
      "entropy": 5.450521326065063,
      "epoch": 2.950079396585947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004140406993425211,
      "loss": 5.1439,
      "mean_token_accuracy": 0.17194535136222838,
      "num_tokens": 65118196.0,
      "step": 29725
    },
    {
      "entropy": 5.447009325027466,
      "epoch": 2.9505756252481143,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004140128410012767,
      "loss": 5.0723,
      "mean_token_accuracy": 0.18142233192920684,
      "num_tokens": 65129924.0,
      "step": 29730
    },
    {
      "entropy": 5.312336492538452,
      "epoch": 2.951071853910282,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041398497921281354,
      "loss": 4.9387,
      "mean_token_accuracy": 0.18559879064559937,
      "num_tokens": 65140994.0,
      "step": 29735
    },
    {
      "entropy": 5.464554262161255,
      "epoch": 2.9515680825724493,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041395711397782236,
      "loss": 5.1665,
      "mean_token_accuracy": 0.16994375884532928,
      "num_tokens": 65153261.0,
      "step": 29740
    },
    {
      "entropy": 5.584180879592895,
      "epoch": 2.9520643112346168,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004139292452969941,
      "loss": 5.2153,
      "mean_token_accuracy": 0.16975643634796142,
      "num_tokens": 65164127.0,
      "step": 29745
    },
    {
      "entropy": 5.4658278465271,
      "epoch": 2.9525605398967842,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041390137317102,
      "loss": 5.0585,
      "mean_token_accuracy": 0.17148589193820954,
      "num_tokens": 65175603.0,
      "step": 29750
    },
    {
      "entropy": 5.4588062286376955,
      "epoch": 2.9530567685589517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004138734976005911,
      "loss": 5.0231,
      "mean_token_accuracy": 0.17165874242782592,
      "num_tokens": 65186780.0,
      "step": 29755
    },
    {
      "entropy": 5.413276672363281,
      "epoch": 2.9535529972211196,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004138456185863986,
      "loss": 4.9929,
      "mean_token_accuracy": 0.17962402403354644,
      "num_tokens": 65198883.0,
      "step": 29760
    },
    {
      "entropy": 5.407801008224487,
      "epoch": 2.954049225883287,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041381773612913413,
      "loss": 5.1028,
      "mean_token_accuracy": 0.17353330552577972,
      "num_tokens": 65209088.0,
      "step": 29765
    },
    {
      "entropy": 5.378012943267822,
      "epoch": 2.9545454545454546,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004137898502294889,
      "loss": 5.0771,
      "mean_token_accuracy": 0.17282765358686447,
      "num_tokens": 65220816.0,
      "step": 29770
    },
    {
      "entropy": 5.4618510723114015,
      "epoch": 2.955041683207622,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004137619608881544,
      "loss": 5.058,
      "mean_token_accuracy": 0.1741151601076126,
      "num_tokens": 65232549.0,
      "step": 29775
    },
    {
      "entropy": 5.5052385330200195,
      "epoch": 2.9555379118697895,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004137340681058223,
      "loss": 5.1079,
      "mean_token_accuracy": 0.1699843868613243,
      "num_tokens": 65243691.0,
      "step": 29780
    },
    {
      "entropy": 5.437232923507691,
      "epoch": 2.956034140531957,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041370617188318424,
      "loss": 5.1076,
      "mean_token_accuracy": 0.1765761137008667,
      "num_tokens": 65254651.0,
      "step": 29785
    },
    {
      "entropy": 5.47172179222107,
      "epoch": 2.956530369194125,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041367827222093195,
      "loss": 5.064,
      "mean_token_accuracy": 0.17713183015584946,
      "num_tokens": 65264669.0,
      "step": 29790
    },
    {
      "entropy": 5.457728672027588,
      "epoch": 2.9570265978562924,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004136503691197574,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17027458995580674,
      "num_tokens": 65274663.0,
      "step": 29795
    },
    {
      "entropy": 5.449551963806153,
      "epoch": 2.95752282651846,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041362246258035254,
      "loss": 5.1172,
      "mean_token_accuracy": 0.1694058820605278,
      "num_tokens": 65286383.0,
      "step": 29800
    },
    {
      "entropy": 5.406468248367309,
      "epoch": 2.9580190551806274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004135945526034091,
      "loss": 4.9437,
      "mean_token_accuracy": 0.1776731640100479,
      "num_tokens": 65297156.0,
      "step": 29805
    },
    {
      "entropy": 5.461227178573608,
      "epoch": 2.958515283842795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041356663918961956,
      "loss": 5.1176,
      "mean_token_accuracy": 0.17388241738080978,
      "num_tokens": 65306836.0,
      "step": 29810
    },
    {
      "entropy": 5.463908863067627,
      "epoch": 2.9590115125049623,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041353872233967595,
      "loss": 5.1506,
      "mean_token_accuracy": 0.16931123584508895,
      "num_tokens": 65318355.0,
      "step": 29815
    },
    {
      "entropy": 5.4678754806518555,
      "epoch": 2.95950774116713,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004135108020542705,
      "loss": 5.1299,
      "mean_token_accuracy": 0.17256377935409545,
      "num_tokens": 65330039.0,
      "step": 29820
    },
    {
      "entropy": 5.429415988922119,
      "epoch": 2.9600039698292973,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041348287833409557,
      "loss": 5.0444,
      "mean_token_accuracy": 0.17689217031002044,
      "num_tokens": 65339771.0,
      "step": 29825
    },
    {
      "entropy": 5.450910472869873,
      "epoch": 2.9605001984914647,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004134549511798438,
      "loss": 5.1124,
      "mean_token_accuracy": 0.1738789275288582,
      "num_tokens": 65352108.0,
      "step": 29830
    },
    {
      "entropy": 5.443082046508789,
      "epoch": 2.960996427153632,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004134270205922075,
      "loss": 5.0384,
      "mean_token_accuracy": 0.1710464060306549,
      "num_tokens": 65364117.0,
      "step": 29835
    },
    {
      "entropy": 5.405335140228272,
      "epoch": 2.9614926558157997,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041339908657187933,
      "loss": 5.0208,
      "mean_token_accuracy": 0.17415360659360885,
      "num_tokens": 65374168.0,
      "step": 29840
    },
    {
      "entropy": 5.521789598464966,
      "epoch": 2.9619888844779676,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004133711491195522,
      "loss": 5.233,
      "mean_token_accuracy": 0.16555067598819734,
      "num_tokens": 65385243.0,
      "step": 29845
    },
    {
      "entropy": 5.48348822593689,
      "epoch": 2.962485113140135,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041334320823591857,
      "loss": 5.0902,
      "mean_token_accuracy": 0.1688772439956665,
      "num_tokens": 65396674.0,
      "step": 29850
    },
    {
      "entropy": 5.49470477104187,
      "epoch": 2.9629813418023025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041331526392167155,
      "loss": 5.1903,
      "mean_token_accuracy": 0.16137987822294236,
      "num_tokens": 65408121.0,
      "step": 29855
    },
    {
      "entropy": 5.503620195388794,
      "epoch": 2.96347757046447,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000413287316177504,
      "loss": 5.1235,
      "mean_token_accuracy": 0.17305453717708588,
      "num_tokens": 65420093.0,
      "step": 29860
    },
    {
      "entropy": 5.386522340774536,
      "epoch": 2.9639737991266375,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041325936500410905,
      "loss": 5.0322,
      "mean_token_accuracy": 0.1784329816699028,
      "num_tokens": 65431338.0,
      "step": 29865
    },
    {
      "entropy": 5.442981290817261,
      "epoch": 2.964470027788805,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041323141040217976,
      "loss": 5.0653,
      "mean_token_accuracy": 0.17338355332612992,
      "num_tokens": 65442305.0,
      "step": 29870
    },
    {
      "entropy": 5.396025943756103,
      "epoch": 2.9649662564509724,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041320345237240923,
      "loss": 5.04,
      "mean_token_accuracy": 0.17256354689598083,
      "num_tokens": 65453028.0,
      "step": 29875
    },
    {
      "entropy": 5.423275661468506,
      "epoch": 2.9654624851131404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000413175490915491,
      "loss": 5.1398,
      "mean_token_accuracy": 0.1747940346598625,
      "num_tokens": 65463374.0,
      "step": 29880
    },
    {
      "entropy": 5.469547319412231,
      "epoch": 2.965958713775308,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004131475260321183,
      "loss": 5.0459,
      "mean_token_accuracy": 0.17314018458127975,
      "num_tokens": 65474144.0,
      "step": 29885
    },
    {
      "entropy": 5.446628093719482,
      "epoch": 2.9664549424374753,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041311955772298464,
      "loss": 5.0503,
      "mean_token_accuracy": 0.19020700454711914,
      "num_tokens": 65485453.0,
      "step": 29890
    },
    {
      "entropy": 5.451904773712158,
      "epoch": 2.966951171099643,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004130915859887837,
      "loss": 5.0414,
      "mean_token_accuracy": 0.18085677474737166,
      "num_tokens": 65496400.0,
      "step": 29895
    },
    {
      "entropy": 5.427404451370239,
      "epoch": 2.9674473997618103,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004130636108302088,
      "loss": 5.1171,
      "mean_token_accuracy": 0.16878614574670792,
      "num_tokens": 65508524.0,
      "step": 29900
    },
    {
      "entropy": 5.283773612976074,
      "epoch": 2.9679436284239777,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041303563224795393,
      "loss": 4.9792,
      "mean_token_accuracy": 0.18349905014038087,
      "num_tokens": 65519264.0,
      "step": 29905
    },
    {
      "entropy": 5.388592052459717,
      "epoch": 2.968439857086145,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004130076502427128,
      "loss": 4.9874,
      "mean_token_accuracy": 0.17640840262174606,
      "num_tokens": 65531049.0,
      "step": 29910
    },
    {
      "entropy": 5.4656439304351805,
      "epoch": 2.9689360857483127,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041297966481517937,
      "loss": 5.1334,
      "mean_token_accuracy": 0.16964331418275833,
      "num_tokens": 65543126.0,
      "step": 29915
    },
    {
      "entropy": 5.522463417053222,
      "epoch": 2.96943231441048,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004129516759660475,
      "loss": 5.1702,
      "mean_token_accuracy": 0.1620056912302971,
      "num_tokens": 65555050.0,
      "step": 29920
    },
    {
      "entropy": 5.494954204559326,
      "epoch": 2.9699285430726476,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041292368369601136,
      "loss": 5.156,
      "mean_token_accuracy": 0.18270967453718184,
      "num_tokens": 65564156.0,
      "step": 29925
    },
    {
      "entropy": 5.487244319915772,
      "epoch": 2.9704247717348156,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000412895688005765,
      "loss": 5.164,
      "mean_token_accuracy": 0.16936672627925872,
      "num_tokens": 65574639.0,
      "step": 29930
    },
    {
      "entropy": 5.447089672088623,
      "epoch": 2.970921000396983,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004128676888960028,
      "loss": 5.0453,
      "mean_token_accuracy": 0.1738000690937042,
      "num_tokens": 65585628.0,
      "step": 29935
    },
    {
      "entropy": 5.394157218933105,
      "epoch": 2.9714172290591505,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041283968636741903,
      "loss": 5.0615,
      "mean_token_accuracy": 0.17981644868850707,
      "num_tokens": 65595544.0,
      "step": 29940
    },
    {
      "entropy": 5.483736991882324,
      "epoch": 2.971913457721318,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004128116804207081,
      "loss": 5.1499,
      "mean_token_accuracy": 0.1739909753203392,
      "num_tokens": 65607963.0,
      "step": 29945
    },
    {
      "entropy": 5.498672008514404,
      "epoch": 2.9724096863834855,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004127836710565642,
      "loss": 5.058,
      "mean_token_accuracy": 0.18117828369140626,
      "num_tokens": 65619881.0,
      "step": 29950
    },
    {
      "entropy": 5.456424522399902,
      "epoch": 2.972905915045653,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041275565827568234,
      "loss": 5.1505,
      "mean_token_accuracy": 0.16778779178857803,
      "num_tokens": 65629967.0,
      "step": 29955
    },
    {
      "entropy": 5.446930313110352,
      "epoch": 2.9734021437078204,
      "grad_norm": 1.03125,
      "learning_rate": 0.000412727642078757,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1739317834377289,
      "num_tokens": 65640804.0,
      "step": 29960
    },
    {
      "entropy": 5.487183952331543,
      "epoch": 2.9738983723699883,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004126996224664829,
      "loss": 5.1684,
      "mean_token_accuracy": 0.1689542457461357,
      "num_tokens": 65652355.0,
      "step": 29965
    },
    {
      "entropy": 5.468720579147339,
      "epoch": 2.974394601032156,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041267159943955483,
      "loss": 5.115,
      "mean_token_accuracy": 0.17231829166412355,
      "num_tokens": 65664047.0,
      "step": 29970
    },
    {
      "entropy": 5.4554526805877686,
      "epoch": 2.9748908296943233,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004126435729986678,
      "loss": 5.0983,
      "mean_token_accuracy": 0.17191943526268005,
      "num_tokens": 65674942.0,
      "step": 29975
    },
    {
      "entropy": 5.446119976043701,
      "epoch": 2.9753870583564908,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041261554314451674,
      "loss": 5.0211,
      "mean_token_accuracy": 0.1778789281845093,
      "num_tokens": 65685285.0,
      "step": 29980
    },
    {
      "entropy": 5.38848729133606,
      "epoch": 2.9758832870186582,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004125875098777967,
      "loss": 5.0139,
      "mean_token_accuracy": 0.18188452571630478,
      "num_tokens": 65695967.0,
      "step": 29985
    },
    {
      "entropy": 5.427148151397705,
      "epoch": 2.9763795156808257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041255947319920297,
      "loss": 5.0996,
      "mean_token_accuracy": 0.17105559557676314,
      "num_tokens": 65707434.0,
      "step": 29990
    },
    {
      "entropy": 5.397427701950074,
      "epoch": 2.976875744342993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041253143310943064,
      "loss": 5.054,
      "mean_token_accuracy": 0.1824893146753311,
      "num_tokens": 65717836.0,
      "step": 29995
    },
    {
      "entropy": 5.449925279617309,
      "epoch": 2.9773719730051607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041250338960917515,
      "loss": 5.0912,
      "mean_token_accuracy": 0.17892757505178453,
      "num_tokens": 65728527.0,
      "step": 30000
    },
    {
      "epoch": 2.9773719730051607,
      "eval_entropy": 5.3078282235721765,
      "eval_loss": 5.222660541534424,
      "eval_mean_token_accuracy": 0.17336195248851133,
      "eval_num_tokens": 65728527.0,
      "eval_runtime": 27.2132,
      "eval_samples_per_second": 1237.231,
      "eval_steps_per_second": 154.668,
      "step": 30000
    },
    {
      "entropy": 5.431055116653442,
      "epoch": 2.977868201667328,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041247534269913197,
      "loss": 5.0469,
      "mean_token_accuracy": 0.17687505036592482,
      "num_tokens": 65739724.0,
      "step": 30005
    },
    {
      "entropy": 5.460644769668579,
      "epoch": 2.9783644303294956,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004124472923799964,
      "loss": 5.1331,
      "mean_token_accuracy": 0.17186431884765624,
      "num_tokens": 65750057.0,
      "step": 30010
    },
    {
      "entropy": 5.42467041015625,
      "epoch": 2.9788606589916635,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004124192386524642,
      "loss": 5.1112,
      "mean_token_accuracy": 0.17535593658685683,
      "num_tokens": 65760534.0,
      "step": 30015
    },
    {
      "entropy": 5.440285873413086,
      "epoch": 2.979356887653831,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000412391181517231,
      "loss": 5.138,
      "mean_token_accuracy": 0.17242445796728134,
      "num_tokens": 65772524.0,
      "step": 30020
    },
    {
      "entropy": 5.467345952987671,
      "epoch": 2.9798531163159985,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041236312097499257,
      "loss": 5.0788,
      "mean_token_accuracy": 0.16977330148220063,
      "num_tokens": 65783003.0,
      "step": 30025
    },
    {
      "entropy": 5.414723110198975,
      "epoch": 2.980349344978166,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004123350570264447,
      "loss": 5.0839,
      "mean_token_accuracy": 0.1688024714589119,
      "num_tokens": 65793219.0,
      "step": 30030
    },
    {
      "entropy": 5.48905234336853,
      "epoch": 2.9808455736403334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041230698967228334,
      "loss": 5.174,
      "mean_token_accuracy": 0.17548916041851043,
      "num_tokens": 65804105.0,
      "step": 30035
    },
    {
      "entropy": 5.620985078811645,
      "epoch": 2.981341802302501,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004122789189132045,
      "loss": 5.1739,
      "mean_token_accuracy": 0.17275945842266083,
      "num_tokens": 65814211.0,
      "step": 30040
    },
    {
      "entropy": 5.520527458190918,
      "epoch": 2.9818380309646684,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004122508447499043,
      "loss": 5.1783,
      "mean_token_accuracy": 0.16582731157541275,
      "num_tokens": 65825304.0,
      "step": 30045
    },
    {
      "entropy": 5.383825397491455,
      "epoch": 2.9823342596268363,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041222276718307896,
      "loss": 5.0193,
      "mean_token_accuracy": 0.17886130958795549,
      "num_tokens": 65835933.0,
      "step": 30050
    },
    {
      "entropy": 5.4451817035675045,
      "epoch": 2.9828304882890038,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004121946862134246,
      "loss": 5.1204,
      "mean_token_accuracy": 0.1724451944231987,
      "num_tokens": 65846479.0,
      "step": 30055
    },
    {
      "entropy": 5.398353624343872,
      "epoch": 2.9833267169511712,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041216660184163767,
      "loss": 5.0026,
      "mean_token_accuracy": 0.18158747702836991,
      "num_tokens": 65857033.0,
      "step": 30060
    },
    {
      "entropy": 5.495658302307129,
      "epoch": 2.9838229456133387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041213851406841456,
      "loss": 5.1838,
      "mean_token_accuracy": 0.16622278839349747,
      "num_tokens": 65867983.0,
      "step": 30065
    },
    {
      "entropy": 5.5114940166473385,
      "epoch": 2.984319174275506,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004121104228944518,
      "loss": 5.1253,
      "mean_token_accuracy": 0.1690904140472412,
      "num_tokens": 65878535.0,
      "step": 30070
    },
    {
      "entropy": 5.504496955871582,
      "epoch": 2.9848154029376737,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004120823283204461,
      "loss": 5.1115,
      "mean_token_accuracy": 0.173164501786232,
      "num_tokens": 65890484.0,
      "step": 30075
    },
    {
      "entropy": 5.36523585319519,
      "epoch": 2.985311631599841,
      "grad_norm": 0.96875,
      "learning_rate": 0.000412054230347094,
      "loss": 5.0307,
      "mean_token_accuracy": 0.17624232023954392,
      "num_tokens": 65901220.0,
      "step": 30080
    },
    {
      "entropy": 5.490283632278443,
      "epoch": 2.9858078602620086,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004120261289750923,
      "loss": 5.0981,
      "mean_token_accuracy": 0.17026471495628356,
      "num_tokens": 65911875.0,
      "step": 30085
    },
    {
      "entropy": 5.422423219680786,
      "epoch": 2.986304088924176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041199802420513785,
      "loss": 5.0593,
      "mean_token_accuracy": 0.17757266610860825,
      "num_tokens": 65921915.0,
      "step": 30090
    },
    {
      "entropy": 5.380258941650391,
      "epoch": 2.9868003175863436,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004119699160379276,
      "loss": 5.0271,
      "mean_token_accuracy": 0.17336097806692125,
      "num_tokens": 65932061.0,
      "step": 30095
    },
    {
      "entropy": 5.395411539077759,
      "epoch": 2.987296546248511,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041194180447415855,
      "loss": 5.0382,
      "mean_token_accuracy": 0.1787525326013565,
      "num_tokens": 65942818.0,
      "step": 30100
    },
    {
      "entropy": 5.419028806686401,
      "epoch": 2.987792774910679,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004119136895145279,
      "loss": 5.0639,
      "mean_token_accuracy": 0.17430567294359206,
      "num_tokens": 65953715.0,
      "step": 30105
    },
    {
      "entropy": 5.50084285736084,
      "epoch": 2.9882890035728464,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004118855711597327,
      "loss": 5.1286,
      "mean_token_accuracy": 0.17468033879995346,
      "num_tokens": 65964235.0,
      "step": 30110
    },
    {
      "entropy": 5.3932915210723875,
      "epoch": 2.988785232235014,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041185744941047033,
      "loss": 5.1106,
      "mean_token_accuracy": 0.16788383275270463,
      "num_tokens": 65976457.0,
      "step": 30115
    },
    {
      "entropy": 5.424220180511474,
      "epoch": 2.9892814608971814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004118293242674382,
      "loss": 5.0587,
      "mean_token_accuracy": 0.17586135268211364,
      "num_tokens": 65986810.0,
      "step": 30120
    },
    {
      "entropy": 5.462913227081299,
      "epoch": 2.989777689559349,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004118011957313336,
      "loss": 5.1856,
      "mean_token_accuracy": 0.16554687321186065,
      "num_tokens": 65998019.0,
      "step": 30125
    },
    {
      "entropy": 5.486494445800782,
      "epoch": 2.9902739182215163,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041177306380285405,
      "loss": 5.1146,
      "mean_token_accuracy": 0.17371783256530762,
      "num_tokens": 66009404.0,
      "step": 30130
    },
    {
      "entropy": 5.354526805877685,
      "epoch": 2.9907701468836843,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004117449284826974,
      "loss": 5.008,
      "mean_token_accuracy": 0.17844681739807128,
      "num_tokens": 66020610.0,
      "step": 30135
    },
    {
      "entropy": 5.438806676864624,
      "epoch": 2.9912663755458517,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041171678977156104,
      "loss": 5.1209,
      "mean_token_accuracy": 0.1669069766998291,
      "num_tokens": 66032576.0,
      "step": 30140
    },
    {
      "entropy": 5.439836645126343,
      "epoch": 2.991762604208019,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041168864767014295,
      "loss": 5.0866,
      "mean_token_accuracy": 0.18044020980596542,
      "num_tokens": 66043378.0,
      "step": 30145
    },
    {
      "entropy": 5.434825611114502,
      "epoch": 2.9922588328701867,
      "grad_norm": 1.078125,
      "learning_rate": 0.000411660502179141,
      "loss": 5.0246,
      "mean_token_accuracy": 0.17557888776063918,
      "num_tokens": 66053960.0,
      "step": 30150
    },
    {
      "entropy": 5.4603582382202145,
      "epoch": 2.992755061532354,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004116323532992529,
      "loss": 5.1068,
      "mean_token_accuracy": 0.1680501639842987,
      "num_tokens": 66066069.0,
      "step": 30155
    },
    {
      "entropy": 5.4068543910980225,
      "epoch": 2.9932512901945216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041160420103117704,
      "loss": 5.0474,
      "mean_token_accuracy": 0.17485309690237044,
      "num_tokens": 66076789.0,
      "step": 30160
    },
    {
      "entropy": 5.403859710693359,
      "epoch": 2.993747518856689,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041157604537561123,
      "loss": 5.0283,
      "mean_token_accuracy": 0.1753920316696167,
      "num_tokens": 66088267.0,
      "step": 30165
    },
    {
      "entropy": 5.335519027709961,
      "epoch": 2.9942437475188566,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041154788633325387,
      "loss": 5.0244,
      "mean_token_accuracy": 0.18166213482618332,
      "num_tokens": 66098789.0,
      "step": 30170
    },
    {
      "entropy": 5.392931938171387,
      "epoch": 2.994739976181024,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004115197239048031,
      "loss": 5.0813,
      "mean_token_accuracy": 0.17177371233701705,
      "num_tokens": 66110881.0,
      "step": 30175
    },
    {
      "entropy": 5.481053638458252,
      "epoch": 2.9952362048431915,
      "grad_norm": 1.265625,
      "learning_rate": 0.00041149155809095734,
      "loss": 5.0639,
      "mean_token_accuracy": 0.17071876525878907,
      "num_tokens": 66121728.0,
      "step": 30180
    },
    {
      "entropy": 5.403755712509155,
      "epoch": 2.995732433505359,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041146338889241506,
      "loss": 4.9736,
      "mean_token_accuracy": 0.17728696167469024,
      "num_tokens": 66131386.0,
      "step": 30185
    },
    {
      "entropy": 5.427803468704224,
      "epoch": 2.996228662167527,
      "grad_norm": 0.90625,
      "learning_rate": 0.00041143521630987483,
      "loss": 5.1529,
      "mean_token_accuracy": 0.17217678874731063,
      "num_tokens": 66142905.0,
      "step": 30190
    },
    {
      "entropy": 5.50144329071045,
      "epoch": 2.9967248908296944,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041140704034403527,
      "loss": 5.2076,
      "mean_token_accuracy": 0.15968669950962067,
      "num_tokens": 66152708.0,
      "step": 30195
    },
    {
      "entropy": 5.576226806640625,
      "epoch": 2.997221119491862,
      "grad_norm": 1.03125,
      "learning_rate": 0.000411378860995595,
      "loss": 5.2434,
      "mean_token_accuracy": 0.17085915356874465,
      "num_tokens": 66164351.0,
      "step": 30200
    },
    {
      "entropy": 5.467614841461182,
      "epoch": 2.9977173481540293,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004113506782652528,
      "loss": 5.1444,
      "mean_token_accuracy": 0.17272571176290513,
      "num_tokens": 66176641.0,
      "step": 30205
    },
    {
      "entropy": 5.561791133880615,
      "epoch": 2.998213576816197,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004113224921537076,
      "loss": 5.1832,
      "mean_token_accuracy": 0.1673513025045395,
      "num_tokens": 66187822.0,
      "step": 30210
    },
    {
      "entropy": 5.469231128692627,
      "epoch": 2.9987098054783643,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041129430266165836,
      "loss": 5.1088,
      "mean_token_accuracy": 0.17153813391923906,
      "num_tokens": 66199636.0,
      "step": 30215
    },
    {
      "entropy": 5.473243570327758,
      "epoch": 2.999206034140532,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004112661097898041,
      "loss": 5.0875,
      "mean_token_accuracy": 0.17134064584970474,
      "num_tokens": 66210649.0,
      "step": 30220
    },
    {
      "entropy": 5.436373662948609,
      "epoch": 2.9997022628026997,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041123791353884397,
      "loss": 5.0754,
      "mean_token_accuracy": 0.17414116114377975,
      "num_tokens": 66221806.0,
      "step": 30225
    },
    {
      "entropy": 5.406945085525512,
      "epoch": 3.000198491464867,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004112097139094771,
      "loss": 5.0443,
      "mean_token_accuracy": 0.1729095995426178,
      "num_tokens": 66232016.0,
      "step": 30230
    },
    {
      "entropy": 5.38259916305542,
      "epoch": 3.0006947201270346,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041118151090240284,
      "loss": 5.0069,
      "mean_token_accuracy": 0.1744341343641281,
      "num_tokens": 66242852.0,
      "step": 30235
    },
    {
      "entropy": 5.470015668869019,
      "epoch": 3.001190948789202,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041115330451832065,
      "loss": 5.0324,
      "mean_token_accuracy": 0.17806857824325562,
      "num_tokens": 66253254.0,
      "step": 30240
    },
    {
      "entropy": 5.465604162216186,
      "epoch": 3.0016871774513696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041112509475792983,
      "loss": 5.0111,
      "mean_token_accuracy": 0.17492075860500336,
      "num_tokens": 66262721.0,
      "step": 30245
    },
    {
      "entropy": 5.412879705429077,
      "epoch": 3.002183406113537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041109688162193,
      "loss": 4.9789,
      "mean_token_accuracy": 0.17760569900274276,
      "num_tokens": 66273718.0,
      "step": 30250
    },
    {
      "entropy": 5.435727500915528,
      "epoch": 3.0026796347757045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004110686651110207,
      "loss": 5.0628,
      "mean_token_accuracy": 0.17392727136611938,
      "num_tokens": 66285527.0,
      "step": 30255
    },
    {
      "entropy": 5.4301858901977536,
      "epoch": 3.003175863437872,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004110404452259017,
      "loss": 5.0112,
      "mean_token_accuracy": 0.17254581153392792,
      "num_tokens": 66295564.0,
      "step": 30260
    },
    {
      "entropy": 5.367746591567993,
      "epoch": 3.0036720921000395,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004110122219672729,
      "loss": 4.9839,
      "mean_token_accuracy": 0.1850528120994568,
      "num_tokens": 66306721.0,
      "step": 30265
    },
    {
      "entropy": 5.394903516769409,
      "epoch": 3.0041683207622074,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000410983995335834,
      "loss": 4.986,
      "mean_token_accuracy": 0.18661617189645768,
      "num_tokens": 66316990.0,
      "step": 30270
    },
    {
      "entropy": 5.404667615890503,
      "epoch": 3.004664549424375,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000410955765332285,
      "loss": 5.0232,
      "mean_token_accuracy": 0.17904876470565795,
      "num_tokens": 66327814.0,
      "step": 30275
    },
    {
      "entropy": 5.405667734146118,
      "epoch": 3.0051607780865424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041092753195732607,
      "loss": 4.9662,
      "mean_token_accuracy": 0.18010715395212173,
      "num_tokens": 66338373.0,
      "step": 30280
    },
    {
      "entropy": 5.415690612792969,
      "epoch": 3.00565700674871,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041089929521165725,
      "loss": 5.0487,
      "mean_token_accuracy": 0.17741331160068513,
      "num_tokens": 66349311.0,
      "step": 30285
    },
    {
      "entropy": 5.494445753097534,
      "epoch": 3.0061532354108773,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004108710550959787,
      "loss": 5.0546,
      "mean_token_accuracy": 0.1787821903824806,
      "num_tokens": 66360771.0,
      "step": 30290
    },
    {
      "entropy": 5.445593357086182,
      "epoch": 3.006649464073045,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041084281161099073,
      "loss": 5.0277,
      "mean_token_accuracy": 0.1742590457201004,
      "num_tokens": 66371348.0,
      "step": 30295
    },
    {
      "entropy": 5.461780405044555,
      "epoch": 3.0071456927352123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004108145647573938,
      "loss": 5.0854,
      "mean_token_accuracy": 0.1770724505186081,
      "num_tokens": 66382489.0,
      "step": 30300
    },
    {
      "entropy": 5.4144213676452635,
      "epoch": 3.0076419213973797,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041078631453588834,
      "loss": 4.9531,
      "mean_token_accuracy": 0.188298037648201,
      "num_tokens": 66392825.0,
      "step": 30305
    },
    {
      "entropy": 5.400065946578979,
      "epoch": 3.0081381500595477,
      "grad_norm": 1.0,
      "learning_rate": 0.0004107580609471748,
      "loss": 5.0238,
      "mean_token_accuracy": 0.1780552327632904,
      "num_tokens": 66403646.0,
      "step": 30310
    },
    {
      "entropy": 5.361766242980957,
      "epoch": 3.008634378721715,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041072980399195394,
      "loss": 4.9206,
      "mean_token_accuracy": 0.19169664829969407,
      "num_tokens": 66414199.0,
      "step": 30315
    },
    {
      "entropy": 5.392786502838135,
      "epoch": 3.0091306073838826,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004107015436709264,
      "loss": 4.9364,
      "mean_token_accuracy": 0.1786775454878807,
      "num_tokens": 66426331.0,
      "step": 30320
    },
    {
      "entropy": 5.367162799835205,
      "epoch": 3.00962683604605,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041067327998479296,
      "loss": 4.9546,
      "mean_token_accuracy": 0.18020122051239013,
      "num_tokens": 66436854.0,
      "step": 30325
    },
    {
      "entropy": 5.401478052139282,
      "epoch": 3.0101230647082176,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041064501293425453,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17798384726047517,
      "num_tokens": 66447757.0,
      "step": 30330
    },
    {
      "entropy": 5.443259906768799,
      "epoch": 3.010619293370385,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004106167425200121,
      "loss": 5.0121,
      "mean_token_accuracy": 0.18059840500354768,
      "num_tokens": 66458395.0,
      "step": 30335
    },
    {
      "entropy": 5.411854600906372,
      "epoch": 3.0111155220325525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004105884687427667,
      "loss": 4.8798,
      "mean_token_accuracy": 0.18946147561073304,
      "num_tokens": 66467728.0,
      "step": 30340
    },
    {
      "entropy": 5.384000301361084,
      "epoch": 3.01161175069472,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004105601916032194,
      "loss": 5.0067,
      "mean_token_accuracy": 0.18579656779766082,
      "num_tokens": 66478113.0,
      "step": 30345
    },
    {
      "entropy": 5.389457130432129,
      "epoch": 3.0121079793568875,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004105319111020714,
      "loss": 4.9887,
      "mean_token_accuracy": 0.17639948427677155,
      "num_tokens": 66489055.0,
      "step": 30350
    },
    {
      "entropy": 5.39278826713562,
      "epoch": 3.0126042080190554,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041050362724002413,
      "loss": 4.9887,
      "mean_token_accuracy": 0.18022744059562684,
      "num_tokens": 66499516.0,
      "step": 30355
    },
    {
      "entropy": 5.434782791137695,
      "epoch": 3.013100436681223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041047534001777884,
      "loss": 5.0576,
      "mean_token_accuracy": 0.17877386659383773,
      "num_tokens": 66510832.0,
      "step": 30360
    },
    {
      "entropy": 5.50333981513977,
      "epoch": 3.0135966653433903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004104470494360371,
      "loss": 5.0997,
      "mean_token_accuracy": 0.17304202169179916,
      "num_tokens": 66522457.0,
      "step": 30365
    },
    {
      "entropy": 5.515033149719239,
      "epoch": 3.014092894005558,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041041875549550043,
      "loss": 5.0463,
      "mean_token_accuracy": 0.177375890314579,
      "num_tokens": 66532527.0,
      "step": 30370
    },
    {
      "entropy": 5.468449640274048,
      "epoch": 3.0145891226677253,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004103904581968703,
      "loss": 5.0305,
      "mean_token_accuracy": 0.18048815429210663,
      "num_tokens": 66542437.0,
      "step": 30375
    },
    {
      "entropy": 5.350233554840088,
      "epoch": 3.0150853513298927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041036215754084864,
      "loss": 4.9995,
      "mean_token_accuracy": 0.17815047651529312,
      "num_tokens": 66552927.0,
      "step": 30380
    },
    {
      "entropy": 5.46967077255249,
      "epoch": 3.0155815799920602,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041033385352813714,
      "loss": 5.1138,
      "mean_token_accuracy": 0.17315081954002381,
      "num_tokens": 66563664.0,
      "step": 30385
    },
    {
      "entropy": 5.47079496383667,
      "epoch": 3.0160778086542277,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004103055461594377,
      "loss": 5.0133,
      "mean_token_accuracy": 0.1799133837223053,
      "num_tokens": 66573968.0,
      "step": 30390
    },
    {
      "entropy": 5.428218746185303,
      "epoch": 3.0165740373163956,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004102772354354522,
      "loss": 5.1024,
      "mean_token_accuracy": 0.17272367924451829,
      "num_tokens": 66585048.0,
      "step": 30395
    },
    {
      "entropy": 5.44521918296814,
      "epoch": 3.017070265978563,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041024892135688287,
      "loss": 5.0765,
      "mean_token_accuracy": 0.17857796102762222,
      "num_tokens": 66594813.0,
      "step": 30400
    },
    {
      "entropy": 5.495073652267456,
      "epoch": 3.0175664946407306,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004102206039244316,
      "loss": 5.0408,
      "mean_token_accuracy": 0.17655157595872878,
      "num_tokens": 66607096.0,
      "step": 30405
    },
    {
      "entropy": 5.454712200164795,
      "epoch": 3.018062723302898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004101922831388008,
      "loss": 5.04,
      "mean_token_accuracy": 0.16867363303899766,
      "num_tokens": 66617001.0,
      "step": 30410
    },
    {
      "entropy": 5.446730422973633,
      "epoch": 3.0185589519650655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041016395900069267,
      "loss": 5.0128,
      "mean_token_accuracy": 0.17389992326498033,
      "num_tokens": 66627036.0,
      "step": 30415
    },
    {
      "entropy": 5.464971208572388,
      "epoch": 3.019055180627233,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004101356315108097,
      "loss": 5.0415,
      "mean_token_accuracy": 0.17568958699703216,
      "num_tokens": 66637526.0,
      "step": 30420
    },
    {
      "entropy": 5.410581398010254,
      "epoch": 3.0195514092894005,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041010730066985414,
      "loss": 4.9788,
      "mean_token_accuracy": 0.18433357179164886,
      "num_tokens": 66650046.0,
      "step": 30425
    },
    {
      "entropy": 5.354899740219116,
      "epoch": 3.020047637951568,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004100789664785287,
      "loss": 4.9301,
      "mean_token_accuracy": 0.184501314163208,
      "num_tokens": 66660922.0,
      "step": 30430
    },
    {
      "entropy": 5.444672060012818,
      "epoch": 3.0205438666137354,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004100506289375359,
      "loss": 5.0934,
      "mean_token_accuracy": 0.17495789974927903,
      "num_tokens": 66671861.0,
      "step": 30435
    },
    {
      "entropy": 5.439648962020874,
      "epoch": 3.0210400952759033,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004100222880475786,
      "loss": 5.0323,
      "mean_token_accuracy": 0.18081325888633729,
      "num_tokens": 66682834.0,
      "step": 30440
    },
    {
      "entropy": 5.412413215637207,
      "epoch": 3.021536323938071,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040999394380935943,
      "loss": 5.0354,
      "mean_token_accuracy": 0.17805605679750441,
      "num_tokens": 66693692.0,
      "step": 30445
    },
    {
      "entropy": 5.453705310821533,
      "epoch": 3.0220325526002383,
      "grad_norm": 1.0,
      "learning_rate": 0.00040996559622358145,
      "loss": 5.031,
      "mean_token_accuracy": 0.17900633066892624,
      "num_tokens": 66705035.0,
      "step": 30450
    },
    {
      "entropy": 5.372064161300659,
      "epoch": 3.0225287812624058,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004099372452909474,
      "loss": 4.936,
      "mean_token_accuracy": 0.1832581043243408,
      "num_tokens": 66714713.0,
      "step": 30455
    },
    {
      "entropy": 5.409888553619385,
      "epoch": 3.0230250099245732,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004099088910121605,
      "loss": 5.0631,
      "mean_token_accuracy": 0.17434003651142121,
      "num_tokens": 66726454.0,
      "step": 30460
    },
    {
      "entropy": 5.448588562011719,
      "epoch": 3.0235212385867407,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040988053338792367,
      "loss": 5.0068,
      "mean_token_accuracy": 0.1763746052980423,
      "num_tokens": 66736103.0,
      "step": 30465
    },
    {
      "entropy": 5.425334548950195,
      "epoch": 3.024017467248908,
      "grad_norm": 1.0,
      "learning_rate": 0.0004098521724189404,
      "loss": 5.0419,
      "mean_token_accuracy": 0.17489339858293534,
      "num_tokens": 66746554.0,
      "step": 30470
    },
    {
      "entropy": 5.467946529388428,
      "epoch": 3.0245136959110757,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040982380810591386,
      "loss": 4.9921,
      "mean_token_accuracy": 0.18031537234783174,
      "num_tokens": 66756568.0,
      "step": 30475
    },
    {
      "entropy": 5.404170751571655,
      "epoch": 3.025009924573243,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004097954404495473,
      "loss": 5.0079,
      "mean_token_accuracy": 0.1775126963853836,
      "num_tokens": 66769036.0,
      "step": 30480
    },
    {
      "entropy": 5.382669925689697,
      "epoch": 3.025506153235411,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040976706945054434,
      "loss": 4.9101,
      "mean_token_accuracy": 0.20191882103681563,
      "num_tokens": 66779868.0,
      "step": 30485
    },
    {
      "entropy": 5.440544843673706,
      "epoch": 3.0260023818975785,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040973869510960837,
      "loss": 5.0308,
      "mean_token_accuracy": 0.17685392498970032,
      "num_tokens": 66791903.0,
      "step": 30490
    },
    {
      "entropy": 5.462393140792846,
      "epoch": 3.026498610559746,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040971031742744314,
      "loss": 5.0548,
      "mean_token_accuracy": 0.17840603590011597,
      "num_tokens": 66802118.0,
      "step": 30495
    },
    {
      "entropy": 5.499448442459107,
      "epoch": 3.0269948392219135,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040968193640475236,
      "loss": 5.0752,
      "mean_token_accuracy": 0.17789812833070756,
      "num_tokens": 66813653.0,
      "step": 30500
    },
    {
      "entropy": 5.515874719619751,
      "epoch": 3.027491067884081,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004096535520422397,
      "loss": 5.1349,
      "mean_token_accuracy": 0.16769630163908006,
      "num_tokens": 66824752.0,
      "step": 30505
    },
    {
      "entropy": 5.457335996627807,
      "epoch": 3.0279872965462484,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040962516434060915,
      "loss": 4.9894,
      "mean_token_accuracy": 0.1816416233778,
      "num_tokens": 66835483.0,
      "step": 30510
    },
    {
      "entropy": 5.455051517486572,
      "epoch": 3.028483525208416,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004095967733005646,
      "loss": 5.0544,
      "mean_token_accuracy": 0.1726195067167282,
      "num_tokens": 66846890.0,
      "step": 30515
    },
    {
      "entropy": 5.35471625328064,
      "epoch": 3.0289797538705834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040956837892281014,
      "loss": 4.9404,
      "mean_token_accuracy": 0.18291661143302917,
      "num_tokens": 66857326.0,
      "step": 30520
    },
    {
      "entropy": 5.459028863906861,
      "epoch": 3.0294759825327513,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040953998120804976,
      "loss": 5.0239,
      "mean_token_accuracy": 0.1809391885995865,
      "num_tokens": 66868266.0,
      "step": 30525
    },
    {
      "entropy": 5.46102204322815,
      "epoch": 3.0299722111949188,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004095115801569878,
      "loss": 5.0481,
      "mean_token_accuracy": 0.1683516949415207,
      "num_tokens": 66879121.0,
      "step": 30530
    },
    {
      "entropy": 5.412670469284057,
      "epoch": 3.0304684398570862,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004094831757703285,
      "loss": 4.9876,
      "mean_token_accuracy": 0.17840183824300765,
      "num_tokens": 66889804.0,
      "step": 30535
    },
    {
      "entropy": 5.443566608428955,
      "epoch": 3.0309646685192537,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040945476804877623,
      "loss": 5.0097,
      "mean_token_accuracy": 0.18038547933101653,
      "num_tokens": 66900421.0,
      "step": 30540
    },
    {
      "entropy": 5.448716402053833,
      "epoch": 3.031460897181421,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004094263569930355,
      "loss": 5.0929,
      "mean_token_accuracy": 0.17304910570383072,
      "num_tokens": 66910754.0,
      "step": 30545
    },
    {
      "entropy": 5.5185283660888675,
      "epoch": 3.0319571258435887,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040939794260381074,
      "loss": 5.0769,
      "mean_token_accuracy": 0.1769281193614006,
      "num_tokens": 66921107.0,
      "step": 30550
    },
    {
      "entropy": 5.382026338577271,
      "epoch": 3.032453354505756,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004093695248818065,
      "loss": 4.9817,
      "mean_token_accuracy": 0.17761482149362565,
      "num_tokens": 66932851.0,
      "step": 30555
    },
    {
      "entropy": 5.431115388870239,
      "epoch": 3.0329495831679236,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004093411038277277,
      "loss": 5.0326,
      "mean_token_accuracy": 0.17486691176891328,
      "num_tokens": 66943403.0,
      "step": 30560
    },
    {
      "entropy": 5.495821571350097,
      "epoch": 3.033445811830091,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000409312679442279,
      "loss": 5.154,
      "mean_token_accuracy": 0.17644871771335602,
      "num_tokens": 66955021.0,
      "step": 30565
    },
    {
      "entropy": 5.478147792816162,
      "epoch": 3.033942040492259,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040928425172616524,
      "loss": 5.06,
      "mean_token_accuracy": 0.17360526323318481,
      "num_tokens": 66966147.0,
      "step": 30570
    },
    {
      "entropy": 5.476765727996826,
      "epoch": 3.0344382691544265,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040925582068009145,
      "loss": 4.9759,
      "mean_token_accuracy": 0.17904510200023652,
      "num_tokens": 66977946.0,
      "step": 30575
    },
    {
      "entropy": 5.4458198070526125,
      "epoch": 3.034934497816594,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004092273863047625,
      "loss": 5.0889,
      "mean_token_accuracy": 0.1763350933790207,
      "num_tokens": 66988014.0,
      "step": 30580
    },
    {
      "entropy": 5.462210750579834,
      "epoch": 3.0354307264787614,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040919894860088366,
      "loss": 5.0516,
      "mean_token_accuracy": 0.17718992829322816,
      "num_tokens": 66999142.0,
      "step": 30585
    },
    {
      "entropy": 5.418029403686523,
      "epoch": 3.035926955140929,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004091705075691601,
      "loss": 5.0776,
      "mean_token_accuracy": 0.17167674452066423,
      "num_tokens": 67010849.0,
      "step": 30590
    },
    {
      "entropy": 5.423986291885376,
      "epoch": 3.0364231838030964,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000409142063210297,
      "loss": 4.9959,
      "mean_token_accuracy": 0.1839539110660553,
      "num_tokens": 67021681.0,
      "step": 30595
    },
    {
      "entropy": 5.446949434280396,
      "epoch": 3.036919412465264,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040911361552499974,
      "loss": 5.0652,
      "mean_token_accuracy": 0.16935237497091293,
      "num_tokens": 67034080.0,
      "step": 30600
    },
    {
      "entropy": 5.44123797416687,
      "epoch": 3.0374156411274313,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040908516451397395,
      "loss": 4.9713,
      "mean_token_accuracy": 0.1766346052289009,
      "num_tokens": 67045528.0,
      "step": 30605
    },
    {
      "entropy": 5.424309778213501,
      "epoch": 3.0379118697895993,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004090567101779249,
      "loss": 4.9853,
      "mean_token_accuracy": 0.1805747330188751,
      "num_tokens": 67057611.0,
      "step": 30610
    },
    {
      "entropy": 5.412502431869507,
      "epoch": 3.0384080984517667,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004090282525175583,
      "loss": 5.0015,
      "mean_token_accuracy": 0.18012103885412217,
      "num_tokens": 67067843.0,
      "step": 30615
    },
    {
      "entropy": 5.477461242675782,
      "epoch": 3.038904327113934,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004089997915335799,
      "loss": 5.1084,
      "mean_token_accuracy": 0.17726214230060577,
      "num_tokens": 67079125.0,
      "step": 30620
    },
    {
      "entropy": 5.377762317657471,
      "epoch": 3.0394005557761017,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004089713272266954,
      "loss": 5.0166,
      "mean_token_accuracy": 0.17489588856697083,
      "num_tokens": 67089750.0,
      "step": 30625
    },
    {
      "entropy": 5.465986680984497,
      "epoch": 3.039896784438269,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004089428595976106,
      "loss": 5.0825,
      "mean_token_accuracy": 0.1748105600476265,
      "num_tokens": 67101738.0,
      "step": 30630
    },
    {
      "entropy": 5.466230249404907,
      "epoch": 3.0403930131004366,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004089143886470316,
      "loss": 4.9804,
      "mean_token_accuracy": 0.18374812304973603,
      "num_tokens": 67111941.0,
      "step": 30635
    },
    {
      "entropy": 5.470699977874756,
      "epoch": 3.040889241762604,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004088859143756643,
      "loss": 5.1174,
      "mean_token_accuracy": 0.17443456053733825,
      "num_tokens": 67124755.0,
      "step": 30640
    },
    {
      "entropy": 5.510725498199463,
      "epoch": 3.0413854704247716,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004088574367842148,
      "loss": 5.2054,
      "mean_token_accuracy": 0.16928387731313704,
      "num_tokens": 67136597.0,
      "step": 30645
    },
    {
      "entropy": 5.436651229858398,
      "epoch": 3.041881699086939,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004088289558733893,
      "loss": 5.089,
      "mean_token_accuracy": 0.17774313390254975,
      "num_tokens": 67146790.0,
      "step": 30650
    },
    {
      "entropy": 5.492081260681152,
      "epoch": 3.042377927749107,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004088004716438941,
      "loss": 5.045,
      "mean_token_accuracy": 0.18249323219060898,
      "num_tokens": 67158054.0,
      "step": 30655
    },
    {
      "entropy": 5.421402502059936,
      "epoch": 3.0428741564112745,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004087719840964355,
      "loss": 4.9792,
      "mean_token_accuracy": 0.1804100379347801,
      "num_tokens": 67168244.0,
      "step": 30660
    },
    {
      "entropy": 5.389609527587891,
      "epoch": 3.043370385073442,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040874349323171994,
      "loss": 4.995,
      "mean_token_accuracy": 0.17884285748004913,
      "num_tokens": 67179165.0,
      "step": 30665
    },
    {
      "entropy": 5.408632373809814,
      "epoch": 3.0438666137356094,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004087149990504539,
      "loss": 4.9919,
      "mean_token_accuracy": 0.17868306189775468,
      "num_tokens": 67189653.0,
      "step": 30670
    },
    {
      "entropy": 5.449233675003052,
      "epoch": 3.044362842397777,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040868650155334406,
      "loss": 5.0674,
      "mean_token_accuracy": 0.17605403512716294,
      "num_tokens": 67203933.0,
      "step": 30675
    },
    {
      "entropy": 5.380284357070923,
      "epoch": 3.0448590710599444,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040865800074109713,
      "loss": 4.9811,
      "mean_token_accuracy": 0.17755133658647537,
      "num_tokens": 67215590.0,
      "step": 30680
    },
    {
      "entropy": 5.407812595367432,
      "epoch": 3.045355299722112,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004086294966144196,
      "loss": 4.9816,
      "mean_token_accuracy": 0.17878589481115342,
      "num_tokens": 67225416.0,
      "step": 30685
    },
    {
      "entropy": 5.414379692077636,
      "epoch": 3.0458515283842793,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004086009891740187,
      "loss": 4.9431,
      "mean_token_accuracy": 0.18239559829235077,
      "num_tokens": 67234492.0,
      "step": 30690
    },
    {
      "entropy": 5.392285108566284,
      "epoch": 3.046347757046447,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040857247842060105,
      "loss": 5.0101,
      "mean_token_accuracy": 0.17575391978025437,
      "num_tokens": 67245630.0,
      "step": 30695
    },
    {
      "entropy": 5.4431548595428465,
      "epoch": 3.0468439857086147,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004085439643548737,
      "loss": 5.1077,
      "mean_token_accuracy": 0.17180492281913756,
      "num_tokens": 67257014.0,
      "step": 30700
    },
    {
      "entropy": 5.328056955337525,
      "epoch": 3.047340214370782,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004085154469775439,
      "loss": 4.9149,
      "mean_token_accuracy": 0.18249691873788834,
      "num_tokens": 67267165.0,
      "step": 30705
    },
    {
      "entropy": 5.422935676574707,
      "epoch": 3.0478364430329496,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004084869262893187,
      "loss": 5.013,
      "mean_token_accuracy": 0.18479709327220917,
      "num_tokens": 67277077.0,
      "step": 30710
    },
    {
      "entropy": 5.4516092300415036,
      "epoch": 3.048332671695117,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004084584022909053,
      "loss": 5.0344,
      "mean_token_accuracy": 0.17888316214084626,
      "num_tokens": 67287941.0,
      "step": 30715
    },
    {
      "entropy": 5.321969652175904,
      "epoch": 3.0488289003572846,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040842987498301115,
      "loss": 4.9156,
      "mean_token_accuracy": 0.17825210988521575,
      "num_tokens": 67299764.0,
      "step": 30720
    },
    {
      "entropy": 5.413554430007935,
      "epoch": 3.049325129019452,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004084013443663437,
      "loss": 5.0032,
      "mean_token_accuracy": 0.17867373079061508,
      "num_tokens": 67310527.0,
      "step": 30725
    },
    {
      "entropy": 5.399587249755859,
      "epoch": 3.0498213576816195,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004083728104416102,
      "loss": 5.0134,
      "mean_token_accuracy": 0.18364480286836624,
      "num_tokens": 67321369.0,
      "step": 30730
    },
    {
      "entropy": 5.377016210556031,
      "epoch": 3.050317586343787,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004083442732095185,
      "loss": 4.9612,
      "mean_token_accuracy": 0.18847027868032457,
      "num_tokens": 67332215.0,
      "step": 30735
    },
    {
      "entropy": 5.483770799636841,
      "epoch": 3.050813815005955,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004083157326707762,
      "loss": 5.1435,
      "mean_token_accuracy": 0.1683424025774002,
      "num_tokens": 67344407.0,
      "step": 30740
    },
    {
      "entropy": 5.473863506317139,
      "epoch": 3.0513100436681224,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040828718882609085,
      "loss": 4.9922,
      "mean_token_accuracy": 0.17917542308568954,
      "num_tokens": 67355513.0,
      "step": 30745
    },
    {
      "entropy": 5.367768096923828,
      "epoch": 3.05180627233029,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040825864167617047,
      "loss": 5.0281,
      "mean_token_accuracy": 0.1760680839419365,
      "num_tokens": 67366515.0,
      "step": 30750
    },
    {
      "entropy": 5.44716215133667,
      "epoch": 3.0523025009924574,
      "grad_norm": 1.015625,
      "learning_rate": 0.000408230091221723,
      "loss": 5.0205,
      "mean_token_accuracy": 0.18233031779527664,
      "num_tokens": 67376568.0,
      "step": 30755
    },
    {
      "entropy": 5.453838634490967,
      "epoch": 3.052798729654625,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004082015374634563,
      "loss": 5.1047,
      "mean_token_accuracy": 0.1718516543507576,
      "num_tokens": 67388639.0,
      "step": 30760
    },
    {
      "entropy": 5.457580661773681,
      "epoch": 3.0532949583167923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004081729804020785,
      "loss": 5.0286,
      "mean_token_accuracy": 0.17119999527931212,
      "num_tokens": 67400330.0,
      "step": 30765
    },
    {
      "entropy": 5.393634510040283,
      "epoch": 3.05379118697896,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004081444200382978,
      "loss": 4.9982,
      "mean_token_accuracy": 0.18151001632213593,
      "num_tokens": 67411067.0,
      "step": 30770
    },
    {
      "entropy": 5.400415420532227,
      "epoch": 3.0542874156411273,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040811585637282234,
      "loss": 4.9588,
      "mean_token_accuracy": 0.18382303416728973,
      "num_tokens": 67421142.0,
      "step": 30775
    },
    {
      "entropy": 5.348321914672852,
      "epoch": 3.0547836443032947,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004080872894063605,
      "loss": 4.9803,
      "mean_token_accuracy": 0.18495043367147446,
      "num_tokens": 67433609.0,
      "step": 30780
    },
    {
      "entropy": 5.502224397659302,
      "epoch": 3.0552798729654627,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040805871913962056,
      "loss": 5.1711,
      "mean_token_accuracy": 0.16880275905132294,
      "num_tokens": 67445281.0,
      "step": 30785
    },
    {
      "entropy": 5.398635149002075,
      "epoch": 3.05577610162763,
      "grad_norm": 1.0,
      "learning_rate": 0.0004080301455733112,
      "loss": 5.0004,
      "mean_token_accuracy": 0.17634643465280533,
      "num_tokens": 67455831.0,
      "step": 30790
    },
    {
      "entropy": 5.443590211868286,
      "epoch": 3.0562723302897976,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040800156870814087,
      "loss": 5.106,
      "mean_token_accuracy": 0.1739383891224861,
      "num_tokens": 67467195.0,
      "step": 30795
    },
    {
      "entropy": 5.388328552246094,
      "epoch": 3.056768558951965,
      "grad_norm": 1.0,
      "learning_rate": 0.00040797298854481826,
      "loss": 4.9717,
      "mean_token_accuracy": 0.18096451908349992,
      "num_tokens": 67478332.0,
      "step": 30800
    },
    {
      "entropy": 5.380906248092652,
      "epoch": 3.0572647876141326,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00040794440508405195,
      "loss": 5.0121,
      "mean_token_accuracy": 0.1832285776734352,
      "num_tokens": 67489600.0,
      "step": 30805
    },
    {
      "entropy": 5.457529306411743,
      "epoch": 3.0577610162763,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000407915818326551,
      "loss": 5.1132,
      "mean_token_accuracy": 0.17064999788999557,
      "num_tokens": 67498837.0,
      "step": 30810
    },
    {
      "entropy": 5.513046789169311,
      "epoch": 3.0582572449384675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000407887228273024,
      "loss": 5.0701,
      "mean_token_accuracy": 0.1773009330034256,
      "num_tokens": 67509118.0,
      "step": 30815
    },
    {
      "entropy": 5.353437280654907,
      "epoch": 3.058753473600635,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004078586349241802,
      "loss": 4.9626,
      "mean_token_accuracy": 0.17451939284801482,
      "num_tokens": 67520577.0,
      "step": 30820
    },
    {
      "entropy": 5.401577615737915,
      "epoch": 3.059249702262803,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040783003828072855,
      "loss": 4.9536,
      "mean_token_accuracy": 0.18178744465112687,
      "num_tokens": 67532020.0,
      "step": 30825
    },
    {
      "entropy": 5.414557695388794,
      "epoch": 3.0597459309249704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004078014383433782,
      "loss": 5.002,
      "mean_token_accuracy": 0.17399215549230576,
      "num_tokens": 67540845.0,
      "step": 30830
    },
    {
      "entropy": 5.365915203094483,
      "epoch": 3.060242159587138,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004077728351128383,
      "loss": 5.0228,
      "mean_token_accuracy": 0.1810273200273514,
      "num_tokens": 67552698.0,
      "step": 30835
    },
    {
      "entropy": 5.319047594070435,
      "epoch": 3.0607383882493053,
      "grad_norm": 1.0,
      "learning_rate": 0.0004077442285898181,
      "loss": 4.9949,
      "mean_token_accuracy": 0.17299688458442689,
      "num_tokens": 67563772.0,
      "step": 30840
    },
    {
      "entropy": 5.438723945617676,
      "epoch": 3.061234616911473,
      "grad_norm": 0.890625,
      "learning_rate": 0.00040771561877502727,
      "loss": 5.0303,
      "mean_token_accuracy": 0.17954077422618867,
      "num_tokens": 67574955.0,
      "step": 30845
    },
    {
      "entropy": 5.4170726299285885,
      "epoch": 3.0617308455736403,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004076870056691749,
      "loss": 4.9874,
      "mean_token_accuracy": 0.1813470557332039,
      "num_tokens": 67586086.0,
      "step": 30850
    },
    {
      "entropy": 5.462694692611694,
      "epoch": 3.0622270742358078,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040765838927297077,
      "loss": 5.0705,
      "mean_token_accuracy": 0.18212185949087142,
      "num_tokens": 67597036.0,
      "step": 30855
    },
    {
      "entropy": 5.444452428817749,
      "epoch": 3.0627233028979752,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004076297695871244,
      "loss": 4.9374,
      "mean_token_accuracy": 0.19318423122167588,
      "num_tokens": 67606499.0,
      "step": 30860
    },
    {
      "entropy": 5.4502228736877445,
      "epoch": 3.0632195315601427,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040760114661234553,
      "loss": 5.0868,
      "mean_token_accuracy": 0.17460348606109619,
      "num_tokens": 67617820.0,
      "step": 30865
    },
    {
      "entropy": 5.40705509185791,
      "epoch": 3.0637157602223106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000407572520349344,
      "loss": 4.9399,
      "mean_token_accuracy": 0.1857955738902092,
      "num_tokens": 67628122.0,
      "step": 30870
    },
    {
      "entropy": 5.3909914016723635,
      "epoch": 3.064211988884478,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004075438907988296,
      "loss": 5.0286,
      "mean_token_accuracy": 0.1779693618416786,
      "num_tokens": 67639127.0,
      "step": 30875
    },
    {
      "entropy": 5.392962121963501,
      "epoch": 3.0647082175466456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004075152579615124,
      "loss": 4.9278,
      "mean_token_accuracy": 0.19110322296619414,
      "num_tokens": 67650122.0,
      "step": 30880
    },
    {
      "entropy": 5.410052108764648,
      "epoch": 3.065204446208813,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004074866218381023,
      "loss": 5.056,
      "mean_token_accuracy": 0.17470774352550505,
      "num_tokens": 67660427.0,
      "step": 30885
    },
    {
      "entropy": 5.397086191177368,
      "epoch": 3.0657006748709805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040745798242930947,
      "loss": 5.1112,
      "mean_token_accuracy": 0.18024676740169526,
      "num_tokens": 67671151.0,
      "step": 30890
    },
    {
      "entropy": 5.481261444091797,
      "epoch": 3.066196903533148,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000407429339735844,
      "loss": 5.0426,
      "mean_token_accuracy": 0.17409393787384034,
      "num_tokens": 67681338.0,
      "step": 30895
    },
    {
      "entropy": 5.372413110733032,
      "epoch": 3.0666931321953155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040740069375841637,
      "loss": 4.9533,
      "mean_token_accuracy": 0.18004682958126067,
      "num_tokens": 67691884.0,
      "step": 30900
    },
    {
      "entropy": 5.376897382736206,
      "epoch": 3.067189360857483,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004073720444977368,
      "loss": 5.0464,
      "mean_token_accuracy": 0.17751684039831161,
      "num_tokens": 67703011.0,
      "step": 30905
    },
    {
      "entropy": 5.51412034034729,
      "epoch": 3.067685589519651,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004073433919545157,
      "loss": 5.2025,
      "mean_token_accuracy": 0.16770657151937485,
      "num_tokens": 67714042.0,
      "step": 30910
    },
    {
      "entropy": 5.453482627868652,
      "epoch": 3.0681818181818183,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004073147361294638,
      "loss": 5.0559,
      "mean_token_accuracy": 0.17612412571907043,
      "num_tokens": 67725449.0,
      "step": 30915
    },
    {
      "entropy": 5.465670108795166,
      "epoch": 3.068678046843986,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004072860770232915,
      "loss": 5.1054,
      "mean_token_accuracy": 0.17757390737533568,
      "num_tokens": 67737334.0,
      "step": 30920
    },
    {
      "entropy": 5.437681436538696,
      "epoch": 3.0691742755061533,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004072574146367095,
      "loss": 5.0035,
      "mean_token_accuracy": 0.17973531633615494,
      "num_tokens": 67747777.0,
      "step": 30925
    },
    {
      "entropy": 5.439814043045044,
      "epoch": 3.0696705041683208,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040722874897042863,
      "loss": 4.9698,
      "mean_token_accuracy": 0.18482491821050645,
      "num_tokens": 67758584.0,
      "step": 30930
    },
    {
      "entropy": 5.364232110977173,
      "epoch": 3.0701667328304882,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004072000800251597,
      "loss": 4.9945,
      "mean_token_accuracy": 0.1809239059686661,
      "num_tokens": 67769168.0,
      "step": 30935
    },
    {
      "entropy": 5.421642160415649,
      "epoch": 3.0706629614926557,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040717140780161367,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18144869804382324,
      "num_tokens": 67779680.0,
      "step": 30940
    },
    {
      "entropy": 5.419221353530884,
      "epoch": 3.071159190154823,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040714273230050153,
      "loss": 5.0253,
      "mean_token_accuracy": 0.16884757429361344,
      "num_tokens": 67789678.0,
      "step": 30945
    },
    {
      "entropy": 5.441141510009766,
      "epoch": 3.0716554188169907,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004071140535225343,
      "loss": 5.0716,
      "mean_token_accuracy": 0.1794372022151947,
      "num_tokens": 67800991.0,
      "step": 30950
    },
    {
      "entropy": 5.5432517528533936,
      "epoch": 3.0721516474791586,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004070853714684233,
      "loss": 5.1511,
      "mean_token_accuracy": 0.17154100984334947,
      "num_tokens": 67811319.0,
      "step": 30955
    },
    {
      "entropy": 5.539687776565552,
      "epoch": 3.072647876141326,
      "grad_norm": 1.0,
      "learning_rate": 0.0004070566861388797,
      "loss": 5.1418,
      "mean_token_accuracy": 0.17393586039543152,
      "num_tokens": 67823704.0,
      "step": 30960
    },
    {
      "entropy": 5.476608228683472,
      "epoch": 3.0731441048034935,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040702799753461477,
      "loss": 5.0993,
      "mean_token_accuracy": 0.17272307723760605,
      "num_tokens": 67835460.0,
      "step": 30965
    },
    {
      "entropy": 5.403738117218017,
      "epoch": 3.073640333465661,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004069993056563401,
      "loss": 4.9823,
      "mean_token_accuracy": 0.18100702613592148,
      "num_tokens": 67845464.0,
      "step": 30970
    },
    {
      "entropy": 5.531709957122803,
      "epoch": 3.0741365621278285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000406970610504767,
      "loss": 5.1058,
      "mean_token_accuracy": 0.1651589274406433,
      "num_tokens": 67856718.0,
      "step": 30975
    },
    {
      "entropy": 5.430958938598633,
      "epoch": 3.074632790789996,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004069419120806072,
      "loss": 5.0004,
      "mean_token_accuracy": 0.18050578385591506,
      "num_tokens": 67867066.0,
      "step": 30980
    },
    {
      "entropy": 5.4299695014953615,
      "epoch": 3.0751290194521634,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040691321038457224,
      "loss": 5.049,
      "mean_token_accuracy": 0.17988150119781493,
      "num_tokens": 67878213.0,
      "step": 30985
    },
    {
      "entropy": 5.436560773849488,
      "epoch": 3.075625248114331,
      "grad_norm": 1.0,
      "learning_rate": 0.0004068845054173739,
      "loss": 5.05,
      "mean_token_accuracy": 0.1721322149038315,
      "num_tokens": 67888503.0,
      "step": 30990
    },
    {
      "entropy": 5.469586706161499,
      "epoch": 3.076121476776499,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004068557971797241,
      "loss": 5.0306,
      "mean_token_accuracy": 0.1706790342926979,
      "num_tokens": 67899035.0,
      "step": 30995
    },
    {
      "entropy": 5.370755290985107,
      "epoch": 3.0766177054386663,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040682708567233464,
      "loss": 4.9986,
      "mean_token_accuracy": 0.18321681022644043,
      "num_tokens": 67909137.0,
      "step": 31000
    },
    {
      "entropy": 5.453863430023193,
      "epoch": 3.077113934100834,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040679837089591743,
      "loss": 5.0222,
      "mean_token_accuracy": 0.18092633336782454,
      "num_tokens": 67919026.0,
      "step": 31005
    },
    {
      "entropy": 5.522797966003418,
      "epoch": 3.0776101627630013,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040676965285118465,
      "loss": 5.1186,
      "mean_token_accuracy": 0.17414498776197435,
      "num_tokens": 67930333.0,
      "step": 31010
    },
    {
      "entropy": 5.427002000808716,
      "epoch": 3.0781063914251687,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004067409315388484,
      "loss": 4.9494,
      "mean_token_accuracy": 0.18482450842857362,
      "num_tokens": 67940448.0,
      "step": 31015
    },
    {
      "entropy": 5.41689920425415,
      "epoch": 3.078602620087336,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040671220695962104,
      "loss": 4.9981,
      "mean_token_accuracy": 0.18075920194387435,
      "num_tokens": 67951398.0,
      "step": 31020
    },
    {
      "entropy": 5.453214120864868,
      "epoch": 3.0790988487495037,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004066834791142146,
      "loss": 5.0501,
      "mean_token_accuracy": 0.1764697477221489,
      "num_tokens": 67962741.0,
      "step": 31025
    },
    {
      "entropy": 5.482502412796021,
      "epoch": 3.079595077411671,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004066547480033417,
      "loss": 5.0358,
      "mean_token_accuracy": 0.17983899414539337,
      "num_tokens": 67975015.0,
      "step": 31030
    },
    {
      "entropy": 5.499275207519531,
      "epoch": 3.0800913060738386,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004066260136277148,
      "loss": 5.0885,
      "mean_token_accuracy": 0.17422713339328766,
      "num_tokens": 67985918.0,
      "step": 31035
    },
    {
      "entropy": 5.401401805877685,
      "epoch": 3.0805875347360065,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040659727598804635,
      "loss": 5.0529,
      "mean_token_accuracy": 0.176213638484478,
      "num_tokens": 67997419.0,
      "step": 31040
    },
    {
      "entropy": 5.410384559631348,
      "epoch": 3.081083763398174,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040656853508504897,
      "loss": 5.0054,
      "mean_token_accuracy": 0.17813187837600708,
      "num_tokens": 68008739.0,
      "step": 31045
    },
    {
      "entropy": 5.4253246784210205,
      "epoch": 3.0815799920603415,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040653979091943555,
      "loss": 5.0559,
      "mean_token_accuracy": 0.1768235057592392,
      "num_tokens": 68020707.0,
      "step": 31050
    },
    {
      "entropy": 5.485356569290161,
      "epoch": 3.082076220722509,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040651104349191864,
      "loss": 4.9755,
      "mean_token_accuracy": 0.1866676241159439,
      "num_tokens": 68030825.0,
      "step": 31055
    },
    {
      "entropy": 5.3900146484375,
      "epoch": 3.0825724493846764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040648229280321126,
      "loss": 5.0311,
      "mean_token_accuracy": 0.1782729595899582,
      "num_tokens": 68041385.0,
      "step": 31060
    },
    {
      "entropy": 5.477252721786499,
      "epoch": 3.083068678046844,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004064535388540264,
      "loss": 5.1256,
      "mean_token_accuracy": 0.17277114242315292,
      "num_tokens": 68053123.0,
      "step": 31065
    },
    {
      "entropy": 5.43350043296814,
      "epoch": 3.0835649067090114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004064247816450769,
      "loss": 4.997,
      "mean_token_accuracy": 0.1771653950214386,
      "num_tokens": 68064529.0,
      "step": 31070
    },
    {
      "entropy": 5.517982912063599,
      "epoch": 3.084061135371179,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004063960211770761,
      "loss": 5.1058,
      "mean_token_accuracy": 0.18003688454627992,
      "num_tokens": 68074091.0,
      "step": 31075
    },
    {
      "entropy": 5.420708847045899,
      "epoch": 3.084557364033347,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040636725745073714,
      "loss": 5.0278,
      "mean_token_accuracy": 0.1785854533314705,
      "num_tokens": 68085185.0,
      "step": 31080
    },
    {
      "entropy": 5.380991458892822,
      "epoch": 3.0850535926955143,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004063384904667732,
      "loss": 4.9284,
      "mean_token_accuracy": 0.18644952327013015,
      "num_tokens": 68095977.0,
      "step": 31085
    },
    {
      "entropy": 5.4147539138793945,
      "epoch": 3.0855498213576817,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040630972022589774,
      "loss": 5.04,
      "mean_token_accuracy": 0.17459308505058288,
      "num_tokens": 68107469.0,
      "step": 31090
    },
    {
      "entropy": 5.56690821647644,
      "epoch": 3.086046050019849,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040628094672882415,
      "loss": 5.2006,
      "mean_token_accuracy": 0.16469986438751222,
      "num_tokens": 68119153.0,
      "step": 31095
    },
    {
      "entropy": 5.413688230514526,
      "epoch": 3.0865422786820167,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000406252169976266,
      "loss": 4.9777,
      "mean_token_accuracy": 0.18006521612405776,
      "num_tokens": 68130336.0,
      "step": 31100
    },
    {
      "entropy": 5.464451742172241,
      "epoch": 3.087038507344184,
      "grad_norm": 1.0,
      "learning_rate": 0.0004062233899689368,
      "loss": 5.0248,
      "mean_token_accuracy": 0.17547304630279542,
      "num_tokens": 68141231.0,
      "step": 31105
    },
    {
      "entropy": 5.420977210998535,
      "epoch": 3.0875347360063516,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004061946067075503,
      "loss": 4.9937,
      "mean_token_accuracy": 0.1830208942294121,
      "num_tokens": 68151891.0,
      "step": 31110
    },
    {
      "entropy": 5.373433303833008,
      "epoch": 3.088030964668519,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004061658201928203,
      "loss": 4.9922,
      "mean_token_accuracy": 0.185211643576622,
      "num_tokens": 68162537.0,
      "step": 31115
    },
    {
      "entropy": 5.491012048721314,
      "epoch": 3.0885271933306866,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004061370304254606,
      "loss": 5.1962,
      "mean_token_accuracy": 0.17073767632246017,
      "num_tokens": 68173393.0,
      "step": 31120
    },
    {
      "entropy": 5.442473077774048,
      "epoch": 3.0890234219928545,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000406108237406185,
      "loss": 5.0715,
      "mean_token_accuracy": 0.17884996980428697,
      "num_tokens": 68184937.0,
      "step": 31125
    },
    {
      "entropy": 5.45105562210083,
      "epoch": 3.089519650655022,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040607944113570777,
      "loss": 5.0879,
      "mean_token_accuracy": 0.179939004778862,
      "num_tokens": 68196962.0,
      "step": 31130
    },
    {
      "entropy": 5.550146579742432,
      "epoch": 3.0900158793171895,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004060506416147428,
      "loss": 5.0695,
      "mean_token_accuracy": 0.17747127413749694,
      "num_tokens": 68208603.0,
      "step": 31135
    },
    {
      "entropy": 5.4403058052062985,
      "epoch": 3.090512107979357,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004060218388440043,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18280685544013978,
      "num_tokens": 68219424.0,
      "step": 31140
    },
    {
      "entropy": 5.38943772315979,
      "epoch": 3.0910083366415244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004059930328242065,
      "loss": 4.9813,
      "mean_token_accuracy": 0.18410756886005403,
      "num_tokens": 68229758.0,
      "step": 31145
    },
    {
      "entropy": 5.348029661178589,
      "epoch": 3.091504565303692,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040596422355606383,
      "loss": 4.9682,
      "mean_token_accuracy": 0.18955538421869278,
      "num_tokens": 68239146.0,
      "step": 31150
    },
    {
      "entropy": 5.428818273544311,
      "epoch": 3.0920007939658594,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004059354110402905,
      "loss": 5.0432,
      "mean_token_accuracy": 0.17175285071134566,
      "num_tokens": 68249339.0,
      "step": 31155
    },
    {
      "entropy": 5.460995483398437,
      "epoch": 3.092497022628027,
      "grad_norm": 1.0,
      "learning_rate": 0.00040590659527760123,
      "loss": 5.0254,
      "mean_token_accuracy": 0.18075771182775496,
      "num_tokens": 68260136.0,
      "step": 31160
    },
    {
      "entropy": 5.4306206703186035,
      "epoch": 3.0929932512901943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040587777626871046,
      "loss": 5.0283,
      "mean_token_accuracy": 0.1749975085258484,
      "num_tokens": 68270033.0,
      "step": 31165
    },
    {
      "entropy": 5.432028484344483,
      "epoch": 3.0934894799523622,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004058489540143328,
      "loss": 5.0902,
      "mean_token_accuracy": 0.171872878074646,
      "num_tokens": 68283344.0,
      "step": 31170
    },
    {
      "entropy": 5.438060283660889,
      "epoch": 3.0939857086145297,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004058201285151831,
      "loss": 4.9963,
      "mean_token_accuracy": 0.17799341827630996,
      "num_tokens": 68294604.0,
      "step": 31175
    },
    {
      "entropy": 5.426261854171753,
      "epoch": 3.094481937276697,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004057912997719761,
      "loss": 4.9815,
      "mean_token_accuracy": 0.19051601737737656,
      "num_tokens": 68304436.0,
      "step": 31180
    },
    {
      "entropy": 5.370155191421508,
      "epoch": 3.0949781659388647,
      "grad_norm": 1.0,
      "learning_rate": 0.0004057624677854266,
      "loss": 4.9559,
      "mean_token_accuracy": 0.1852072447538376,
      "num_tokens": 68315593.0,
      "step": 31185
    },
    {
      "entropy": 5.403580999374389,
      "epoch": 3.095474394601032,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040573363255624976,
      "loss": 5.0678,
      "mean_token_accuracy": 0.1801113799214363,
      "num_tokens": 68327540.0,
      "step": 31190
    },
    {
      "entropy": 5.541655969619751,
      "epoch": 3.0959706232631996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004057047940851605,
      "loss": 5.1155,
      "mean_token_accuracy": 0.17386146187782286,
      "num_tokens": 68339784.0,
      "step": 31195
    },
    {
      "entropy": 5.427757263183594,
      "epoch": 3.096466851925367,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004056759523728739,
      "loss": 5.0291,
      "mean_token_accuracy": 0.17516302913427353,
      "num_tokens": 68351182.0,
      "step": 31200
    },
    {
      "entropy": 5.400889348983765,
      "epoch": 3.0969630805875346,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040564710742010545,
      "loss": 5.076,
      "mean_token_accuracy": 0.17350465208292007,
      "num_tokens": 68361726.0,
      "step": 31205
    },
    {
      "entropy": 5.419814395904541,
      "epoch": 3.0974593092497025,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004056182592275702,
      "loss": 5.0527,
      "mean_token_accuracy": 0.17764192223548889,
      "num_tokens": 68373167.0,
      "step": 31210
    },
    {
      "entropy": 5.445407676696777,
      "epoch": 3.09795553791187,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004055894077959835,
      "loss": 5.1003,
      "mean_token_accuracy": 0.17580832690000534,
      "num_tokens": 68384646.0,
      "step": 31215
    },
    {
      "entropy": 5.4206513404846195,
      "epoch": 3.0984517665740374,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004055605531260609,
      "loss": 5.0457,
      "mean_token_accuracy": 0.17830387353897095,
      "num_tokens": 68395953.0,
      "step": 31220
    },
    {
      "entropy": 5.49874324798584,
      "epoch": 3.098947995236205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040553169521851794,
      "loss": 5.1056,
      "mean_token_accuracy": 0.17680739164352416,
      "num_tokens": 68407167.0,
      "step": 31225
    },
    {
      "entropy": 5.47255916595459,
      "epoch": 3.0994442238983724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040550283407407025,
      "loss": 5.1014,
      "mean_token_accuracy": 0.17586308866739273,
      "num_tokens": 68419530.0,
      "step": 31230
    },
    {
      "entropy": 5.428594017028809,
      "epoch": 3.09994045256054,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004054739696934334,
      "loss": 5.0413,
      "mean_token_accuracy": 0.1777152821421623,
      "num_tokens": 68430970.0,
      "step": 31235
    },
    {
      "entropy": 5.418001794815064,
      "epoch": 3.1004366812227073,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040544510207732337,
      "loss": 5.0137,
      "mean_token_accuracy": 0.17828303277492524,
      "num_tokens": 68441257.0,
      "step": 31240
    },
    {
      "entropy": 5.374426794052124,
      "epoch": 3.100932909884875,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040541623122645576,
      "loss": 5.0156,
      "mean_token_accuracy": 0.17922561764717101,
      "num_tokens": 68452493.0,
      "step": 31245
    },
    {
      "entropy": 5.4008856296539305,
      "epoch": 3.1014291385470423,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040538735714154675,
      "loss": 4.9628,
      "mean_token_accuracy": 0.1880912348628044,
      "num_tokens": 68464023.0,
      "step": 31250
    },
    {
      "entropy": 5.454248094558716,
      "epoch": 3.10192536720921,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004053584798233122,
      "loss": 5.0513,
      "mean_token_accuracy": 0.18702177256345748,
      "num_tokens": 68474519.0,
      "step": 31255
    },
    {
      "entropy": 5.48228178024292,
      "epoch": 3.1024215958713777,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040532959927246814,
      "loss": 5.0303,
      "mean_token_accuracy": 0.18305650353431702,
      "num_tokens": 68486239.0,
      "step": 31260
    },
    {
      "entropy": 5.492786645889282,
      "epoch": 3.102917824533545,
      "grad_norm": 1.03125,
      "learning_rate": 0.000405300715489731,
      "loss": 5.0871,
      "mean_token_accuracy": 0.17292223870754242,
      "num_tokens": 68495921.0,
      "step": 31265
    },
    {
      "entropy": 5.480477666854858,
      "epoch": 3.1034140531957126,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004052718284758169,
      "loss": 4.9821,
      "mean_token_accuracy": 0.18467807173728942,
      "num_tokens": 68505773.0,
      "step": 31270
    },
    {
      "entropy": 5.486834192276001,
      "epoch": 3.10391028185788,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000405242938231442,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17327849715948104,
      "num_tokens": 68516563.0,
      "step": 31275
    },
    {
      "entropy": 5.456728839874268,
      "epoch": 3.1044065105200476,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000405214044757323,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1819281682372093,
      "num_tokens": 68527952.0,
      "step": 31280
    },
    {
      "entropy": 5.451285934448242,
      "epoch": 3.104902739182215,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004051851480541763,
      "loss": 5.0406,
      "mean_token_accuracy": 0.1752521052956581,
      "num_tokens": 68538450.0,
      "step": 31285
    },
    {
      "entropy": 5.395168447494507,
      "epoch": 3.1053989678443825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004051562481227184,
      "loss": 4.9854,
      "mean_token_accuracy": 0.17712334394454957,
      "num_tokens": 68547941.0,
      "step": 31290
    },
    {
      "entropy": 5.40633602142334,
      "epoch": 3.10589519650655,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000405127344963666,
      "loss": 4.982,
      "mean_token_accuracy": 0.18072637021541596,
      "num_tokens": 68558525.0,
      "step": 31295
    },
    {
      "entropy": 5.385242080688476,
      "epoch": 3.106391425168718,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004050984385777359,
      "loss": 4.9876,
      "mean_token_accuracy": 0.17467813789844513,
      "num_tokens": 68569487.0,
      "step": 31300
    },
    {
      "entropy": 5.420099496841431,
      "epoch": 3.1068876538308854,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004050695289656448,
      "loss": 5.0241,
      "mean_token_accuracy": 0.1772749423980713,
      "num_tokens": 68580069.0,
      "step": 31305
    },
    {
      "entropy": 5.4363926410675045,
      "epoch": 3.107383882493053,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040504061612810964,
      "loss": 4.9891,
      "mean_token_accuracy": 0.18077816665172577,
      "num_tokens": 68591511.0,
      "step": 31310
    },
    {
      "entropy": 5.330124378204346,
      "epoch": 3.1078801111552203,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004050117000658475,
      "loss": 5.0262,
      "mean_token_accuracy": 0.18432814478874207,
      "num_tokens": 68602464.0,
      "step": 31315
    },
    {
      "entropy": 5.472047615051269,
      "epoch": 3.108376339817388,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040498278077957534,
      "loss": 4.9919,
      "mean_token_accuracy": 0.18034832179546356,
      "num_tokens": 68613492.0,
      "step": 31320
    },
    {
      "entropy": 5.39381594657898,
      "epoch": 3.1088725684795553,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004049538582700102,
      "loss": 4.9626,
      "mean_token_accuracy": 0.1854908734560013,
      "num_tokens": 68625259.0,
      "step": 31325
    },
    {
      "entropy": 5.45538215637207,
      "epoch": 3.1093687971417228,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004049249325378695,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1727551758289337,
      "num_tokens": 68635455.0,
      "step": 31330
    },
    {
      "entropy": 5.440607929229737,
      "epoch": 3.1098650258038902,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040489600358387036,
      "loss": 5.1097,
      "mean_token_accuracy": 0.17507832795381545,
      "num_tokens": 68648102.0,
      "step": 31335
    },
    {
      "entropy": 5.4727987289428714,
      "epoch": 3.110361254466058,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004048670714087303,
      "loss": 5.0552,
      "mean_token_accuracy": 0.18101482540369035,
      "num_tokens": 68660854.0,
      "step": 31340
    },
    {
      "entropy": 5.353068923950195,
      "epoch": 3.1108574831282256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004048381360131667,
      "loss": 4.9123,
      "mean_token_accuracy": 0.1916647046804428,
      "num_tokens": 68671839.0,
      "step": 31345
    },
    {
      "entropy": 5.4908326148986815,
      "epoch": 3.111353711790393,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004048091973978971,
      "loss": 5.0542,
      "mean_token_accuracy": 0.1811479449272156,
      "num_tokens": 68682535.0,
      "step": 31350
    },
    {
      "entropy": 5.421372556686402,
      "epoch": 3.1118499404525606,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004047802555636391,
      "loss": 5.0162,
      "mean_token_accuracy": 0.17997049838304519,
      "num_tokens": 68694283.0,
      "step": 31355
    },
    {
      "entropy": 5.455580329895019,
      "epoch": 3.112346169114728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004047513105111104,
      "loss": 5.0293,
      "mean_token_accuracy": 0.1796354442834854,
      "num_tokens": 68704268.0,
      "step": 31360
    },
    {
      "entropy": 5.392293214797974,
      "epoch": 3.1128423977768955,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040472236224102883,
      "loss": 4.9421,
      "mean_token_accuracy": 0.17860536724328996,
      "num_tokens": 68714762.0,
      "step": 31365
    },
    {
      "entropy": 5.435914182662964,
      "epoch": 3.113338626439063,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040469341075411217,
      "loss": 4.994,
      "mean_token_accuracy": 0.17347205728292464,
      "num_tokens": 68725302.0,
      "step": 31370
    },
    {
      "entropy": 5.47335786819458,
      "epoch": 3.1138348551012305,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004046644560510783,
      "loss": 5.0976,
      "mean_token_accuracy": 0.16375306844711304,
      "num_tokens": 68736893.0,
      "step": 31375
    },
    {
      "entropy": 5.370839548110962,
      "epoch": 3.114331083763398,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040463549813264557,
      "loss": 4.9904,
      "mean_token_accuracy": 0.17862183302640916,
      "num_tokens": 68748202.0,
      "step": 31380
    },
    {
      "entropy": 5.477337884902954,
      "epoch": 3.114827312425566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040460653699953166,
      "loss": 5.0113,
      "mean_token_accuracy": 0.18281923979520798,
      "num_tokens": 68759033.0,
      "step": 31385
    },
    {
      "entropy": 5.445584058761597,
      "epoch": 3.1153235410877333,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040457757265245485,
      "loss": 4.9864,
      "mean_token_accuracy": 0.18533600866794586,
      "num_tokens": 68770879.0,
      "step": 31390
    },
    {
      "entropy": 5.3746199131011965,
      "epoch": 3.115819769749901,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004045486050921336,
      "loss": 4.9606,
      "mean_token_accuracy": 0.18752655237913132,
      "num_tokens": 68782087.0,
      "step": 31395
    },
    {
      "entropy": 5.422916555404663,
      "epoch": 3.1163159984120683,
      "grad_norm": 0.96875,
      "learning_rate": 0.000404519634319286,
      "loss": 4.9944,
      "mean_token_accuracy": 0.1804192155599594,
      "num_tokens": 68792355.0,
      "step": 31400
    },
    {
      "entropy": 5.409803819656372,
      "epoch": 3.1168122270742358,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040449066033463053,
      "loss": 5.0417,
      "mean_token_accuracy": 0.17402148246765137,
      "num_tokens": 68804255.0,
      "step": 31405
    },
    {
      "entropy": 5.406555128097534,
      "epoch": 3.1173084557364032,
      "grad_norm": 1.0,
      "learning_rate": 0.0004044616831388858,
      "loss": 4.9905,
      "mean_token_accuracy": 0.1790971800684929,
      "num_tokens": 68816102.0,
      "step": 31410
    },
    {
      "entropy": 5.445473957061767,
      "epoch": 3.1178046843985707,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004044327027327703,
      "loss": 5.0266,
      "mean_token_accuracy": 0.1746950015425682,
      "num_tokens": 68827303.0,
      "step": 31415
    },
    {
      "entropy": 5.445245218276978,
      "epoch": 3.118300913060738,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004044037191170026,
      "loss": 5.0036,
      "mean_token_accuracy": 0.1824526757001877,
      "num_tokens": 68838418.0,
      "step": 31420
    },
    {
      "entropy": 5.511459827423096,
      "epoch": 3.118797141722906,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040437473229230155,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17356811761856078,
      "num_tokens": 68848431.0,
      "step": 31425
    },
    {
      "entropy": 5.387138843536377,
      "epoch": 3.1192933703850736,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004043457422593859,
      "loss": 4.9653,
      "mean_token_accuracy": 0.17693071216344833,
      "num_tokens": 68858983.0,
      "step": 31430
    },
    {
      "entropy": 5.4024042129516605,
      "epoch": 3.119789599047241,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004043167490189746,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18375889211893082,
      "num_tokens": 68870170.0,
      "step": 31435
    },
    {
      "entropy": 5.392703723907471,
      "epoch": 3.1202858277094085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040428775257178655,
      "loss": 5.0434,
      "mean_token_accuracy": 0.18488616049289702,
      "num_tokens": 68880242.0,
      "step": 31440
    },
    {
      "entropy": 5.485537052154541,
      "epoch": 3.120782056371576,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040425875291854083,
      "loss": 5.1103,
      "mean_token_accuracy": 0.17383488714694978,
      "num_tokens": 68892150.0,
      "step": 31445
    },
    {
      "entropy": 5.509271430969238,
      "epoch": 3.1212782850337435,
      "grad_norm": 1.0,
      "learning_rate": 0.00040422975005995656,
      "loss": 5.101,
      "mean_token_accuracy": 0.1760658785700798,
      "num_tokens": 68903164.0,
      "step": 31450
    },
    {
      "entropy": 5.413443183898925,
      "epoch": 3.121774513695911,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004042007439967529,
      "loss": 4.9903,
      "mean_token_accuracy": 0.17902772128582,
      "num_tokens": 68914514.0,
      "step": 31455
    },
    {
      "entropy": 5.425589847564697,
      "epoch": 3.1222707423580784,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004041717347296493,
      "loss": 5.0109,
      "mean_token_accuracy": 0.17838136702775956,
      "num_tokens": 68925366.0,
      "step": 31460
    },
    {
      "entropy": 5.40103931427002,
      "epoch": 3.122766971020246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004041427222593649,
      "loss": 4.9636,
      "mean_token_accuracy": 0.17823778390884398,
      "num_tokens": 68936019.0,
      "step": 31465
    },
    {
      "entropy": 5.32774920463562,
      "epoch": 3.123263199682414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040411370658661935,
      "loss": 4.9374,
      "mean_token_accuracy": 0.187249656021595,
      "num_tokens": 68947405.0,
      "step": 31470
    },
    {
      "entropy": 5.405989313125611,
      "epoch": 3.1237594283445813,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040408468771213206,
      "loss": 5.0222,
      "mean_token_accuracy": 0.17782043665647507,
      "num_tokens": 68958822.0,
      "step": 31475
    },
    {
      "entropy": 5.429111099243164,
      "epoch": 3.124255657006749,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040405566563662265,
      "loss": 5.0611,
      "mean_token_accuracy": 0.17853304594755173,
      "num_tokens": 68969992.0,
      "step": 31480
    },
    {
      "entropy": 5.431702041625977,
      "epoch": 3.1247518856689163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040402664036081086,
      "loss": 4.994,
      "mean_token_accuracy": 0.17607382386922837,
      "num_tokens": 68982081.0,
      "step": 31485
    },
    {
      "entropy": 5.487847137451172,
      "epoch": 3.1252481143310837,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004039976118854164,
      "loss": 5.0438,
      "mean_token_accuracy": 0.17383482456207275,
      "num_tokens": 68993430.0,
      "step": 31490
    },
    {
      "entropy": 5.526445198059082,
      "epoch": 3.125744342993251,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004039685802111591,
      "loss": 5.1077,
      "mean_token_accuracy": 0.17528558671474456,
      "num_tokens": 69004939.0,
      "step": 31495
    },
    {
      "entropy": 5.406407594680786,
      "epoch": 3.1262405716554187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040393954533875904,
      "loss": 5.0124,
      "mean_token_accuracy": 0.18180859684944153,
      "num_tokens": 69016293.0,
      "step": 31500
    },
    {
      "entropy": 5.482573318481445,
      "epoch": 3.126736800317586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040391050726893595,
      "loss": 5.139,
      "mean_token_accuracy": 0.17175114452838897,
      "num_tokens": 69029087.0,
      "step": 31505
    },
    {
      "entropy": 5.454280710220337,
      "epoch": 3.127233028979754,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004038814660024101,
      "loss": 5.0282,
      "mean_token_accuracy": 0.17422485053539277,
      "num_tokens": 69040937.0,
      "step": 31510
    },
    {
      "entropy": 5.461332607269287,
      "epoch": 3.1277292576419216,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004038524215399016,
      "loss": 5.1025,
      "mean_token_accuracy": 0.17604524046182632,
      "num_tokens": 69052212.0,
      "step": 31515
    },
    {
      "entropy": 5.457508611679077,
      "epoch": 3.128225486304089,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004038233738821308,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18894447833299638,
      "num_tokens": 69061951.0,
      "step": 31520
    },
    {
      "entropy": 5.470997524261475,
      "epoch": 3.1287217149662565,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004037943230298178,
      "loss": 4.9661,
      "mean_token_accuracy": 0.190520079433918,
      "num_tokens": 69072156.0,
      "step": 31525
    },
    {
      "entropy": 5.40261025428772,
      "epoch": 3.129217943628424,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040376526898368314,
      "loss": 5.0167,
      "mean_token_accuracy": 0.1790131464600563,
      "num_tokens": 69084032.0,
      "step": 31530
    },
    {
      "entropy": 5.45599684715271,
      "epoch": 3.1297141722905915,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040373621174444736,
      "loss": 5.0657,
      "mean_token_accuracy": 0.17781862616539001,
      "num_tokens": 69096436.0,
      "step": 31535
    },
    {
      "entropy": 5.457815027236938,
      "epoch": 3.130210400952759,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040370715131283076,
      "loss": 5.0038,
      "mean_token_accuracy": 0.17943631410598754,
      "num_tokens": 69107090.0,
      "step": 31540
    },
    {
      "entropy": 5.442255115509033,
      "epoch": 3.1307066296149264,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004036780876895543,
      "loss": 4.9512,
      "mean_token_accuracy": 0.18381458073854445,
      "num_tokens": 69117644.0,
      "step": 31545
    },
    {
      "entropy": 5.361119604110717,
      "epoch": 3.131202858277094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004036490208753385,
      "loss": 5.0355,
      "mean_token_accuracy": 0.17841470390558242,
      "num_tokens": 69129639.0,
      "step": 31550
    },
    {
      "entropy": 5.384012603759766,
      "epoch": 3.131699086939262,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040361995087090407,
      "loss": 4.976,
      "mean_token_accuracy": 0.18100745230913162,
      "num_tokens": 69139429.0,
      "step": 31555
    },
    {
      "entropy": 5.308837985992431,
      "epoch": 3.1321953156014293,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040359087767697223,
      "loss": 4.9152,
      "mean_token_accuracy": 0.19052889794111252,
      "num_tokens": 69150777.0,
      "step": 31560
    },
    {
      "entropy": 5.426954126358032,
      "epoch": 3.1326915442635968,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040356180129426356,
      "loss": 5.0016,
      "mean_token_accuracy": 0.17539652138948442,
      "num_tokens": 69161970.0,
      "step": 31565
    },
    {
      "entropy": 5.361167860031128,
      "epoch": 3.1331877729257642,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004035327217234993,
      "loss": 4.9899,
      "mean_token_accuracy": 0.18094566315412522,
      "num_tokens": 69172990.0,
      "step": 31570
    },
    {
      "entropy": 5.38245849609375,
      "epoch": 3.1336840015879317,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040350363896540054,
      "loss": 5.0263,
      "mean_token_accuracy": 0.1773155763745308,
      "num_tokens": 69184044.0,
      "step": 31575
    },
    {
      "entropy": 5.344258165359497,
      "epoch": 3.134180230250099,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040347455302068835,
      "loss": 4.9755,
      "mean_token_accuracy": 0.17793227434158326,
      "num_tokens": 69194841.0,
      "step": 31580
    },
    {
      "entropy": 5.416613531112671,
      "epoch": 3.1346764589122667,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040344546389008406,
      "loss": 5.0449,
      "mean_token_accuracy": 0.17797410786151885,
      "num_tokens": 69205465.0,
      "step": 31585
    },
    {
      "entropy": 5.445279264450074,
      "epoch": 3.135172687574434,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040341637157430916,
      "loss": 5.0746,
      "mean_token_accuracy": 0.17440278083086014,
      "num_tokens": 69215831.0,
      "step": 31590
    },
    {
      "entropy": 5.444326591491699,
      "epoch": 3.135668916236602,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004033872760740849,
      "loss": 5.0061,
      "mean_token_accuracy": 0.18213892579078675,
      "num_tokens": 69225156.0,
      "step": 31595
    },
    {
      "entropy": 5.420374774932862,
      "epoch": 3.1361651448987695,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004033581773901328,
      "loss": 5.0525,
      "mean_token_accuracy": 0.17196638137102127,
      "num_tokens": 69237590.0,
      "step": 31600
    },
    {
      "entropy": 5.446770763397216,
      "epoch": 3.136661373560937,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040332907552317453,
      "loss": 5.0051,
      "mean_token_accuracy": 0.18124110400676727,
      "num_tokens": 69246904.0,
      "step": 31605
    },
    {
      "entropy": 5.545190238952637,
      "epoch": 3.1371576022231045,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004032999704739317,
      "loss": 5.1255,
      "mean_token_accuracy": 0.1691249877214432,
      "num_tokens": 69258289.0,
      "step": 31610
    },
    {
      "entropy": 5.4198554992675785,
      "epoch": 3.137653830885272,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004032708622431261,
      "loss": 5.0025,
      "mean_token_accuracy": 0.179735866189003,
      "num_tokens": 69268232.0,
      "step": 31615
    },
    {
      "entropy": 5.433409357070923,
      "epoch": 3.1381500595474394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004032417508314795,
      "loss": 5.005,
      "mean_token_accuracy": 0.1709439292550087,
      "num_tokens": 69278596.0,
      "step": 31620
    },
    {
      "entropy": 5.413076400756836,
      "epoch": 3.138646288209607,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040321263623971376,
      "loss": 5.1028,
      "mean_token_accuracy": 0.17598633468151093,
      "num_tokens": 69290675.0,
      "step": 31625
    },
    {
      "entropy": 5.488792610168457,
      "epoch": 3.1391425168717744,
      "grad_norm": 0.859375,
      "learning_rate": 0.000403183518468551,
      "loss": 5.1078,
      "mean_token_accuracy": 0.1807928964495659,
      "num_tokens": 69303600.0,
      "step": 31630
    },
    {
      "entropy": 5.505779075622558,
      "epoch": 3.139638745533942,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004031543975187132,
      "loss": 5.123,
      "mean_token_accuracy": 0.1722521185874939,
      "num_tokens": 69314672.0,
      "step": 31635
    },
    {
      "entropy": 5.403490304946899,
      "epoch": 3.1401349741961098,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040312527339092237,
      "loss": 4.9833,
      "mean_token_accuracy": 0.1851476475596428,
      "num_tokens": 69325574.0,
      "step": 31640
    },
    {
      "entropy": 5.386563491821289,
      "epoch": 3.1406312028582772,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040309614608590095,
      "loss": 4.9662,
      "mean_token_accuracy": 0.17717470079660416,
      "num_tokens": 69336430.0,
      "step": 31645
    },
    {
      "entropy": 5.377104806900024,
      "epoch": 3.1411274315204447,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004030670156043711,
      "loss": 4.8978,
      "mean_token_accuracy": 0.19291851073503494,
      "num_tokens": 69347007.0,
      "step": 31650
    },
    {
      "entropy": 5.446631288528442,
      "epoch": 3.141623660182612,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040303788194705523,
      "loss": 5.0296,
      "mean_token_accuracy": 0.17975423187017442,
      "num_tokens": 69357796.0,
      "step": 31655
    },
    {
      "entropy": 5.463075780868531,
      "epoch": 3.1421198888447797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040300874511467574,
      "loss": 5.0554,
      "mean_token_accuracy": 0.17883034348487853,
      "num_tokens": 69370098.0,
      "step": 31660
    },
    {
      "entropy": 5.431987476348877,
      "epoch": 3.142616117506947,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040297960510795526,
      "loss": 5.0289,
      "mean_token_accuracy": 0.17190966606140137,
      "num_tokens": 69380991.0,
      "step": 31665
    },
    {
      "entropy": 5.4747590065002445,
      "epoch": 3.1431123461691146,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004029504619276163,
      "loss": 5.0196,
      "mean_token_accuracy": 0.177481572329998,
      "num_tokens": 69391748.0,
      "step": 31670
    },
    {
      "entropy": 5.508114004135132,
      "epoch": 3.143608574831282,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040292131557438166,
      "loss": 5.0513,
      "mean_token_accuracy": 0.17033583223819732,
      "num_tokens": 69404964.0,
      "step": 31675
    },
    {
      "entropy": 5.355388069152832,
      "epoch": 3.14410480349345,
      "grad_norm": 1.046875,
      "learning_rate": 0.000402892166048974,
      "loss": 4.9929,
      "mean_token_accuracy": 0.1849071741104126,
      "num_tokens": 69416045.0,
      "step": 31680
    },
    {
      "entropy": 5.38374547958374,
      "epoch": 3.1446010321556175,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004028630133521162,
      "loss": 4.9843,
      "mean_token_accuracy": 0.17467437833547592,
      "num_tokens": 69427489.0,
      "step": 31685
    },
    {
      "entropy": 5.417068195343018,
      "epoch": 3.145097260817785,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004028338574845312,
      "loss": 4.9741,
      "mean_token_accuracy": 0.18272719979286195,
      "num_tokens": 69437929.0,
      "step": 31690
    },
    {
      "entropy": 5.396613502502442,
      "epoch": 3.1455934894799524,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040280469844694203,
      "loss": 5.0563,
      "mean_token_accuracy": 0.18056145012378694,
      "num_tokens": 69449449.0,
      "step": 31695
    },
    {
      "entropy": 5.397697830200196,
      "epoch": 3.14608971814212,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004027755362400716,
      "loss": 5.0658,
      "mean_token_accuracy": 0.17444247752428055,
      "num_tokens": 69459451.0,
      "step": 31700
    },
    {
      "entropy": 5.533738088607788,
      "epoch": 3.1465859468042874,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004027463708646434,
      "loss": 5.0703,
      "mean_token_accuracy": 0.17075889110565184,
      "num_tokens": 69470615.0,
      "step": 31705
    },
    {
      "entropy": 5.489894390106201,
      "epoch": 3.147082175466455,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004027172023213803,
      "loss": 5.0073,
      "mean_token_accuracy": 0.18486927449703217,
      "num_tokens": 69481123.0,
      "step": 31710
    },
    {
      "entropy": 5.442805337905884,
      "epoch": 3.1475784041286223,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004026880306110058,
      "loss": 5.0532,
      "mean_token_accuracy": 0.1711467832326889,
      "num_tokens": 69491590.0,
      "step": 31715
    },
    {
      "entropy": 5.3626072883605955,
      "epoch": 3.14807463279079,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040265885573424336,
      "loss": 5.0093,
      "mean_token_accuracy": 0.1803894191980362,
      "num_tokens": 69501883.0,
      "step": 31720
    },
    {
      "entropy": 5.3907075881958,
      "epoch": 3.1485708614529577,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040262967769181634,
      "loss": 4.97,
      "mean_token_accuracy": 0.17727602124214173,
      "num_tokens": 69514556.0,
      "step": 31725
    },
    {
      "entropy": 5.486307573318482,
      "epoch": 3.149067090115125,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040260049648444836,
      "loss": 5.0953,
      "mean_token_accuracy": 0.17640545815229416,
      "num_tokens": 69525483.0,
      "step": 31730
    },
    {
      "entropy": 5.46252121925354,
      "epoch": 3.1495633187772927,
      "grad_norm": 1.125,
      "learning_rate": 0.00040257131211286307,
      "loss": 5.0544,
      "mean_token_accuracy": 0.17977770417928696,
      "num_tokens": 69535681.0,
      "step": 31735
    },
    {
      "entropy": 5.475524663925171,
      "epoch": 3.15005954743946,
      "grad_norm": 0.90625,
      "learning_rate": 0.000402542124577784,
      "loss": 5.1637,
      "mean_token_accuracy": 0.16045575886964797,
      "num_tokens": 69547825.0,
      "step": 31740
    },
    {
      "entropy": 5.394737482070923,
      "epoch": 3.1505557761016276,
      "grad_norm": 1.0,
      "learning_rate": 0.00040251293387993513,
      "loss": 5.0088,
      "mean_token_accuracy": 0.17769081741571427,
      "num_tokens": 69557922.0,
      "step": 31745
    },
    {
      "entropy": 5.476569032669067,
      "epoch": 3.151052004763795,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004024837400200403,
      "loss": 5.0735,
      "mean_token_accuracy": 0.17336517125368117,
      "num_tokens": 69568165.0,
      "step": 31750
    },
    {
      "entropy": 5.3822966575622555,
      "epoch": 3.1515482334259626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040245454299882343,
      "loss": 4.872,
      "mean_token_accuracy": 0.1883975476026535,
      "num_tokens": 69578828.0,
      "step": 31755
    },
    {
      "entropy": 5.444305372238159,
      "epoch": 3.15204446208813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040242534281700843,
      "loss": 5.0771,
      "mean_token_accuracy": 0.17286327630281448,
      "num_tokens": 69590527.0,
      "step": 31760
    },
    {
      "entropy": 5.434608411788941,
      "epoch": 3.152540690750298,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040239613947531965,
      "loss": 5.0465,
      "mean_token_accuracy": 0.17288795113563538,
      "num_tokens": 69602032.0,
      "step": 31765
    },
    {
      "entropy": 5.39622483253479,
      "epoch": 3.1530369194124654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040236693297448107,
      "loss": 5.0308,
      "mean_token_accuracy": 0.17270180135965346,
      "num_tokens": 69613490.0,
      "step": 31770
    },
    {
      "entropy": 5.495949840545654,
      "epoch": 3.153533148074633,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004023377233152169,
      "loss": 5.1186,
      "mean_token_accuracy": 0.1704629182815552,
      "num_tokens": 69624623.0,
      "step": 31775
    },
    {
      "entropy": 5.455958890914917,
      "epoch": 3.1540293767368004,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040230851049825173,
      "loss": 5.0825,
      "mean_token_accuracy": 0.17882893681526185,
      "num_tokens": 69634956.0,
      "step": 31780
    },
    {
      "entropy": 5.451047420501709,
      "epoch": 3.154525605398968,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040227929452430974,
      "loss": 4.9999,
      "mean_token_accuracy": 0.18247926086187363,
      "num_tokens": 69645370.0,
      "step": 31785
    },
    {
      "entropy": 5.475924158096314,
      "epoch": 3.1550218340611353,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040225007539411553,
      "loss": 5.0683,
      "mean_token_accuracy": 0.16892875730991364,
      "num_tokens": 69656328.0,
      "step": 31790
    },
    {
      "entropy": 5.39500322341919,
      "epoch": 3.155518062723303,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004022208531083936,
      "loss": 5.0315,
      "mean_token_accuracy": 0.17788039594888688,
      "num_tokens": 69668584.0,
      "step": 31795
    },
    {
      "entropy": 5.404466676712036,
      "epoch": 3.1560142913854703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004021916276678688,
      "loss": 5.0519,
      "mean_token_accuracy": 0.1716272383928299,
      "num_tokens": 69677977.0,
      "step": 31800
    },
    {
      "entropy": 5.412043762207031,
      "epoch": 3.1565105200476378,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040216239907326555,
      "loss": 5.0328,
      "mean_token_accuracy": 0.18126794844865798,
      "num_tokens": 69688979.0,
      "step": 31805
    },
    {
      "entropy": 5.417147302627564,
      "epoch": 3.1570067487098052,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004021331673253089,
      "loss": 5.0598,
      "mean_token_accuracy": 0.19029378294944763,
      "num_tokens": 69698567.0,
      "step": 31810
    },
    {
      "entropy": 5.423541927337647,
      "epoch": 3.157502977371973,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004021039324247236,
      "loss": 4.9651,
      "mean_token_accuracy": 0.17654325664043427,
      "num_tokens": 69708623.0,
      "step": 31815
    },
    {
      "entropy": 5.362532138824463,
      "epoch": 3.1579992060341406,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040207469437223466,
      "loss": 5.0512,
      "mean_token_accuracy": 0.17220204323530197,
      "num_tokens": 69719613.0,
      "step": 31820
    },
    {
      "entropy": 5.418300771713257,
      "epoch": 3.158495434696308,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040204545316856714,
      "loss": 5.0414,
      "mean_token_accuracy": 0.17625727951526643,
      "num_tokens": 69731407.0,
      "step": 31825
    },
    {
      "entropy": 5.388593912124634,
      "epoch": 3.1589916633584756,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004020162088144461,
      "loss": 4.9756,
      "mean_token_accuracy": 0.17868475914001464,
      "num_tokens": 69742363.0,
      "step": 31830
    },
    {
      "entropy": 5.385921573638916,
      "epoch": 3.159487892020643,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00040198696131059684,
      "loss": 4.9904,
      "mean_token_accuracy": 0.18311012238264085,
      "num_tokens": 69754878.0,
      "step": 31835
    },
    {
      "entropy": 5.379011631011963,
      "epoch": 3.1599841206828105,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040195771065774455,
      "loss": 4.9582,
      "mean_token_accuracy": 0.18658592104911803,
      "num_tokens": 69764986.0,
      "step": 31840
    },
    {
      "entropy": 5.391807699203492,
      "epoch": 3.160480349344978,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004019284568566146,
      "loss": 5.0036,
      "mean_token_accuracy": 0.18214359134435654,
      "num_tokens": 69775762.0,
      "step": 31845
    },
    {
      "entropy": 5.405081176757813,
      "epoch": 3.1609765780071455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040189919990793243,
      "loss": 5.0597,
      "mean_token_accuracy": 0.17691610753536224,
      "num_tokens": 69786411.0,
      "step": 31850
    },
    {
      "entropy": 5.467455101013184,
      "epoch": 3.1614728066693134,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040186993981242353,
      "loss": 5.0164,
      "mean_token_accuracy": 0.1785281240940094,
      "num_tokens": 69797140.0,
      "step": 31855
    },
    {
      "entropy": 5.362284183502197,
      "epoch": 3.161969035331481,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004018406765708135,
      "loss": 4.9225,
      "mean_token_accuracy": 0.18303183019161223,
      "num_tokens": 69807452.0,
      "step": 31860
    },
    {
      "entropy": 5.362602806091308,
      "epoch": 3.1624652639936484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040181141018382806,
      "loss": 5.0151,
      "mean_token_accuracy": 0.17935281991958618,
      "num_tokens": 69819650.0,
      "step": 31865
    },
    {
      "entropy": 5.441627454757691,
      "epoch": 3.162961492655816,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040178214065219285,
      "loss": 4.9828,
      "mean_token_accuracy": 0.17717088907957076,
      "num_tokens": 69830411.0,
      "step": 31870
    },
    {
      "entropy": 5.409550523757934,
      "epoch": 3.1634577213179833,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004017528679766337,
      "loss": 4.9775,
      "mean_token_accuracy": 0.18706733137369155,
      "num_tokens": 69841795.0,
      "step": 31875
    },
    {
      "entropy": 5.3529839515686035,
      "epoch": 3.163953949980151,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004017235921578767,
      "loss": 4.99,
      "mean_token_accuracy": 0.17909646332263945,
      "num_tokens": 69851633.0,
      "step": 31880
    },
    {
      "entropy": 5.393439865112304,
      "epoch": 3.1644501786423183,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040169431319664756,
      "loss": 4.925,
      "mean_token_accuracy": 0.18644967973232268,
      "num_tokens": 69862398.0,
      "step": 31885
    },
    {
      "entropy": 5.4112310886383055,
      "epoch": 3.1649464073044857,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004016650310936725,
      "loss": 5.0048,
      "mean_token_accuracy": 0.18150992393493653,
      "num_tokens": 69874397.0,
      "step": 31890
    },
    {
      "entropy": 5.4167201042175295,
      "epoch": 3.165442635966653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004016357458496776,
      "loss": 4.9676,
      "mean_token_accuracy": 0.1908063694834709,
      "num_tokens": 69883950.0,
      "step": 31895
    },
    {
      "entropy": 5.392321825027466,
      "epoch": 3.165938864628821,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000401606457465389,
      "loss": 4.9434,
      "mean_token_accuracy": 0.18436367958784103,
      "num_tokens": 69895586.0,
      "step": 31900
    },
    {
      "entropy": 5.415063762664795,
      "epoch": 3.1664350932909886,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004015771659415332,
      "loss": 4.9995,
      "mean_token_accuracy": 0.18355436325073243,
      "num_tokens": 69906093.0,
      "step": 31905
    },
    {
      "entropy": 5.3522865772247314,
      "epoch": 3.166931321953156,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004015478712788364,
      "loss": 4.9488,
      "mean_token_accuracy": 0.18301405906677246,
      "num_tokens": 69916381.0,
      "step": 31910
    },
    {
      "entropy": 5.373984003067017,
      "epoch": 3.1674275506153236,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040151857347802506,
      "loss": 4.9816,
      "mean_token_accuracy": 0.18506899029016494,
      "num_tokens": 69926059.0,
      "step": 31915
    },
    {
      "entropy": 5.388499307632446,
      "epoch": 3.167923779277491,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040148927253982577,
      "loss": 5.021,
      "mean_token_accuracy": 0.17680176347494125,
      "num_tokens": 69937677.0,
      "step": 31920
    },
    {
      "entropy": 5.4069664478302,
      "epoch": 3.1684200079396585,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040145996846496497,
      "loss": 4.9898,
      "mean_token_accuracy": 0.17966058999300002,
      "num_tokens": 69949399.0,
      "step": 31925
    },
    {
      "entropy": 5.429215955734253,
      "epoch": 3.168916236601826,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004014306612541695,
      "loss": 5.0513,
      "mean_token_accuracy": 0.1864097982645035,
      "num_tokens": 69960595.0,
      "step": 31930
    },
    {
      "entropy": 5.430311012268066,
      "epoch": 3.1694124652639935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004014013509081661,
      "loss": 5.0103,
      "mean_token_accuracy": 0.18216459304094315,
      "num_tokens": 69971018.0,
      "step": 31935
    },
    {
      "entropy": 5.420721340179443,
      "epoch": 3.1699086939261614,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040137203742768155,
      "loss": 5.0099,
      "mean_token_accuracy": 0.1758624568581581,
      "num_tokens": 69982507.0,
      "step": 31940
    },
    {
      "entropy": 5.412484121322632,
      "epoch": 3.170404922588329,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004013427208134427,
      "loss": 4.9822,
      "mean_token_accuracy": 0.190234537422657,
      "num_tokens": 69994190.0,
      "step": 31945
    },
    {
      "entropy": 5.4235954761505125,
      "epoch": 3.1709011512504963,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040131340106617673,
      "loss": 5.169,
      "mean_token_accuracy": 0.16903722882270814,
      "num_tokens": 70005522.0,
      "step": 31950
    },
    {
      "entropy": 5.540057229995727,
      "epoch": 3.171397379912664,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004012840781866106,
      "loss": 5.0809,
      "mean_token_accuracy": 0.18258402049541472,
      "num_tokens": 70016706.0,
      "step": 31955
    },
    {
      "entropy": 5.423424434661865,
      "epoch": 3.1718936085748313,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040125475217547135,
      "loss": 4.9743,
      "mean_token_accuracy": 0.18194901645183564,
      "num_tokens": 70028577.0,
      "step": 31960
    },
    {
      "entropy": 5.370511531829834,
      "epoch": 3.1723898372369987,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040122542303348634,
      "loss": 4.9935,
      "mean_token_accuracy": 0.18199673146009446,
      "num_tokens": 70040258.0,
      "step": 31965
    },
    {
      "entropy": 5.4275102615356445,
      "epoch": 3.172886065899166,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004011960907613828,
      "loss": 5.1142,
      "mean_token_accuracy": 0.17480211704969406,
      "num_tokens": 70052105.0,
      "step": 31970
    },
    {
      "entropy": 5.428532934188842,
      "epoch": 3.1733822945613337,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040116675535988825,
      "loss": 4.9837,
      "mean_token_accuracy": 0.18139038234949112,
      "num_tokens": 70062272.0,
      "step": 31975
    },
    {
      "entropy": 5.402163410186768,
      "epoch": 3.173878523223501,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040113741682972994,
      "loss": 5.0123,
      "mean_token_accuracy": 0.18312013149261475,
      "num_tokens": 70073614.0,
      "step": 31980
    },
    {
      "entropy": 5.352928161621094,
      "epoch": 3.174374751885669,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004011080751716356,
      "loss": 5.0201,
      "mean_token_accuracy": 0.17611438781023026,
      "num_tokens": 70085191.0,
      "step": 31985
    },
    {
      "entropy": 5.44111180305481,
      "epoch": 3.1748709805478366,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040107873038633255,
      "loss": 5.0169,
      "mean_token_accuracy": 0.1800113797187805,
      "num_tokens": 70097610.0,
      "step": 31990
    },
    {
      "entropy": 5.366000461578369,
      "epoch": 3.175367209210004,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004010493824745487,
      "loss": 4.9581,
      "mean_token_accuracy": 0.17839191257953643,
      "num_tokens": 70109040.0,
      "step": 31995
    },
    {
      "entropy": 5.407608413696289,
      "epoch": 3.1758634378721715,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040102003143701194,
      "loss": 5.0598,
      "mean_token_accuracy": 0.17232741564512252,
      "num_tokens": 70120472.0,
      "step": 32000
    },
    {
      "entropy": 5.399544954299927,
      "epoch": 3.176359666534339,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040099067727444984,
      "loss": 4.912,
      "mean_token_accuracy": 0.1904263585805893,
      "num_tokens": 70130481.0,
      "step": 32005
    },
    {
      "entropy": 5.349321031570435,
      "epoch": 3.1768558951965065,
      "grad_norm": 1.0,
      "learning_rate": 0.0004009613199875904,
      "loss": 4.9485,
      "mean_token_accuracy": 0.18073015809059143,
      "num_tokens": 70140783.0,
      "step": 32010
    },
    {
      "entropy": 5.3844434261322025,
      "epoch": 3.177352123858674,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004009319595771617,
      "loss": 5.0456,
      "mean_token_accuracy": 0.179297037422657,
      "num_tokens": 70151624.0,
      "step": 32015
    },
    {
      "entropy": 5.457970762252808,
      "epoch": 3.1778483525208414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004009025960438918,
      "loss": 5.031,
      "mean_token_accuracy": 0.17713291347026824,
      "num_tokens": 70162600.0,
      "step": 32020
    },
    {
      "entropy": 5.490862035751343,
      "epoch": 3.1783445811830093,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040087322938850867,
      "loss": 5.064,
      "mean_token_accuracy": 0.17708141058683396,
      "num_tokens": 70175563.0,
      "step": 32025
    },
    {
      "entropy": 5.443698596954346,
      "epoch": 3.178840809845177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004008438596117408,
      "loss": 4.9722,
      "mean_token_accuracy": 0.17647278010845185,
      "num_tokens": 70186358.0,
      "step": 32030
    },
    {
      "entropy": 5.426778268814087,
      "epoch": 3.1793370385073443,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040081448671431635,
      "loss": 5.0072,
      "mean_token_accuracy": 0.18045843094587327,
      "num_tokens": 70196872.0,
      "step": 32035
    },
    {
      "entropy": 5.357103443145752,
      "epoch": 3.1798332671695118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040078511069696377,
      "loss": 4.9947,
      "mean_token_accuracy": 0.17262820899486542,
      "num_tokens": 70208136.0,
      "step": 32040
    },
    {
      "entropy": 5.488703298568725,
      "epoch": 3.1803294958316792,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040075573156041146,
      "loss": 5.2527,
      "mean_token_accuracy": 0.16451481282711028,
      "num_tokens": 70219716.0,
      "step": 32045
    },
    {
      "entropy": 5.468333721160889,
      "epoch": 3.1808257244938467,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040072634930538796,
      "loss": 5.0508,
      "mean_token_accuracy": 0.17436484545469283,
      "num_tokens": 70230502.0,
      "step": 32050
    },
    {
      "entropy": 5.477762460708618,
      "epoch": 3.181321953156014,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040069696393262196,
      "loss": 5.0614,
      "mean_token_accuracy": 0.17510006874799727,
      "num_tokens": 70241820.0,
      "step": 32055
    },
    {
      "entropy": 5.406827402114868,
      "epoch": 3.1818181818181817,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004006675754428421,
      "loss": 4.9714,
      "mean_token_accuracy": 0.18063822239637375,
      "num_tokens": 70252243.0,
      "step": 32060
    },
    {
      "entropy": 5.365631151199341,
      "epoch": 3.182314410480349,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040063818383677716,
      "loss": 4.8993,
      "mean_token_accuracy": 0.18310720473527908,
      "num_tokens": 70261835.0,
      "step": 32065
    },
    {
      "entropy": 5.431239652633667,
      "epoch": 3.182810639142517,
      "grad_norm": 1.046875,
      "learning_rate": 0.000400608789115156,
      "loss": 5.0796,
      "mean_token_accuracy": 0.17688443809747695,
      "num_tokens": 70273210.0,
      "step": 32070
    },
    {
      "entropy": 5.413159799575806,
      "epoch": 3.1833068678046845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040057939127870754,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1741419330239296,
      "num_tokens": 70283904.0,
      "step": 32075
    },
    {
      "entropy": 5.445113897323608,
      "epoch": 3.183803096466852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004005499903281607,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18484007716178893,
      "num_tokens": 70295199.0,
      "step": 32080
    },
    {
      "entropy": 5.342087650299073,
      "epoch": 3.1842993251290195,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040052058626424464,
      "loss": 4.9062,
      "mean_token_accuracy": 0.18452021926641465,
      "num_tokens": 70305418.0,
      "step": 32085
    },
    {
      "entropy": 5.398749113082886,
      "epoch": 3.184795553791187,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004004911790876886,
      "loss": 5.0926,
      "mean_token_accuracy": 0.17645997405052186,
      "num_tokens": 70316131.0,
      "step": 32090
    },
    {
      "entropy": 5.4219855785369875,
      "epoch": 3.1852917824533544,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040046176879922156,
      "loss": 4.9605,
      "mean_token_accuracy": 0.17629352807998658,
      "num_tokens": 70326804.0,
      "step": 32095
    },
    {
      "entropy": 5.5513200759887695,
      "epoch": 3.185788011115522,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040043235539957315,
      "loss": 5.1644,
      "mean_token_accuracy": 0.1659463405609131,
      "num_tokens": 70338125.0,
      "step": 32100
    },
    {
      "entropy": 5.423504638671875,
      "epoch": 3.1862842397776894,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004004029388894726,
      "loss": 4.9962,
      "mean_token_accuracy": 0.17750122547149658,
      "num_tokens": 70348614.0,
      "step": 32105
    },
    {
      "entropy": 5.398408079147339,
      "epoch": 3.1867804684398573,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004003735192696494,
      "loss": 5.0595,
      "mean_token_accuracy": 0.17332563251256944,
      "num_tokens": 70358735.0,
      "step": 32110
    },
    {
      "entropy": 5.372394466400147,
      "epoch": 3.1872766971020248,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000400344096540833,
      "loss": 4.9133,
      "mean_token_accuracy": 0.18947056978940963,
      "num_tokens": 70369448.0,
      "step": 32115
    },
    {
      "entropy": 5.39379096031189,
      "epoch": 3.1877729257641922,
      "grad_norm": 1.0,
      "learning_rate": 0.00040031467070375315,
      "loss": 4.9881,
      "mean_token_accuracy": 0.1824806198477745,
      "num_tokens": 70379462.0,
      "step": 32120
    },
    {
      "entropy": 5.3970374584198,
      "epoch": 3.1882691544263597,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004002852417591395,
      "loss": 5.0127,
      "mean_token_accuracy": 0.1762759119272232,
      "num_tokens": 70389283.0,
      "step": 32125
    },
    {
      "entropy": 5.421643304824829,
      "epoch": 3.188765383088527,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004002558097077218,
      "loss": 5.0627,
      "mean_token_accuracy": 0.17599455267190933,
      "num_tokens": 70400058.0,
      "step": 32130
    },
    {
      "entropy": 5.375017261505127,
      "epoch": 3.1892616117506947,
      "grad_norm": 1.0,
      "learning_rate": 0.0004002263745502299,
      "loss": 4.9277,
      "mean_token_accuracy": 0.18394435495138167,
      "num_tokens": 70411614.0,
      "step": 32135
    },
    {
      "entropy": 5.388391399383545,
      "epoch": 3.189757840412862,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040019693628739377,
      "loss": 5.0022,
      "mean_token_accuracy": 0.17903356701135636,
      "num_tokens": 70422404.0,
      "step": 32140
    },
    {
      "entropy": 5.495574474334717,
      "epoch": 3.1902540690750296,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004001674949199434,
      "loss": 5.0733,
      "mean_token_accuracy": 0.1792507991194725,
      "num_tokens": 70432219.0,
      "step": 32145
    },
    {
      "entropy": 5.42336950302124,
      "epoch": 3.190750297737197,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004001380504486089,
      "loss": 5.0302,
      "mean_token_accuracy": 0.17795901000499725,
      "num_tokens": 70443129.0,
      "step": 32150
    },
    {
      "entropy": 5.406414651870728,
      "epoch": 3.191246526399365,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004001086028741204,
      "loss": 5.0557,
      "mean_token_accuracy": 0.177509169280529,
      "num_tokens": 70453225.0,
      "step": 32155
    },
    {
      "entropy": 5.385157632827759,
      "epoch": 3.1917427550615325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004000791521972081,
      "loss": 5.0265,
      "mean_token_accuracy": 0.18294699043035506,
      "num_tokens": 70464909.0,
      "step": 32160
    },
    {
      "entropy": 5.4320008754730225,
      "epoch": 3.1922389837237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040004969841860233,
      "loss": 5.0349,
      "mean_token_accuracy": 0.1776182934641838,
      "num_tokens": 70475211.0,
      "step": 32165
    },
    {
      "entropy": 5.51686806678772,
      "epoch": 3.1927352123858674,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040002024153903347,
      "loss": 5.0282,
      "mean_token_accuracy": 0.1791270449757576,
      "num_tokens": 70485603.0,
      "step": 32170
    },
    {
      "entropy": 5.382646608352661,
      "epoch": 3.193231441048035,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003999907815592321,
      "loss": 4.9576,
      "mean_token_accuracy": 0.18493513464927674,
      "num_tokens": 70497163.0,
      "step": 32175
    },
    {
      "entropy": 5.370192575454712,
      "epoch": 3.1937276697102024,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039996131847992863,
      "loss": 4.9807,
      "mean_token_accuracy": 0.18408907055854798,
      "num_tokens": 70509323.0,
      "step": 32180
    },
    {
      "entropy": 5.443582534790039,
      "epoch": 3.19422389837237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003999318523018537,
      "loss": 5.106,
      "mean_token_accuracy": 0.17083721309900285,
      "num_tokens": 70520767.0,
      "step": 32185
    },
    {
      "entropy": 5.378158807754517,
      "epoch": 3.1947201270345373,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003999023830257382,
      "loss": 4.9507,
      "mean_token_accuracy": 0.17971781343221666,
      "num_tokens": 70531671.0,
      "step": 32190
    },
    {
      "entropy": 5.395445871353149,
      "epoch": 3.1952163556967053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003998729106523126,
      "loss": 4.9872,
      "mean_token_accuracy": 0.1768506348133087,
      "num_tokens": 70542309.0,
      "step": 32195
    },
    {
      "entropy": 5.41291913986206,
      "epoch": 3.1957125843588727,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003998434351823078,
      "loss": 5.0839,
      "mean_token_accuracy": 0.17524780184030533,
      "num_tokens": 70553003.0,
      "step": 32200
    },
    {
      "entropy": 5.366220331192016,
      "epoch": 3.19620881302104,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000399813956616455,
      "loss": 5.0205,
      "mean_token_accuracy": 0.17572832256555557,
      "num_tokens": 70564161.0,
      "step": 32205
    },
    {
      "entropy": 5.44270076751709,
      "epoch": 3.1967050416832077,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039978447495548497,
      "loss": 4.9902,
      "mean_token_accuracy": 0.18107685893774034,
      "num_tokens": 70575891.0,
      "step": 32210
    },
    {
      "entropy": 5.4980189323425295,
      "epoch": 3.197201270345375,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039975499020012886,
      "loss": 5.14,
      "mean_token_accuracy": 0.17667128294706344,
      "num_tokens": 70586691.0,
      "step": 32215
    },
    {
      "entropy": 5.392534875869751,
      "epoch": 3.1976974990075426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039972550235111776,
      "loss": 4.9418,
      "mean_token_accuracy": 0.1798628479242325,
      "num_tokens": 70597423.0,
      "step": 32220
    },
    {
      "entropy": 5.40168571472168,
      "epoch": 3.19819372766971,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000399696011409183,
      "loss": 5.0401,
      "mean_token_accuracy": 0.1761089563369751,
      "num_tokens": 70609828.0,
      "step": 32225
    },
    {
      "entropy": 5.338017368316651,
      "epoch": 3.1986899563318776,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039966651737505585,
      "loss": 4.9422,
      "mean_token_accuracy": 0.1808035209774971,
      "num_tokens": 70620480.0,
      "step": 32230
    },
    {
      "entropy": 5.398492670059204,
      "epoch": 3.199186184994045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003996370202494677,
      "loss": 5.0028,
      "mean_token_accuracy": 0.18386178016662597,
      "num_tokens": 70630045.0,
      "step": 32235
    },
    {
      "entropy": 5.353194284439087,
      "epoch": 3.199682413656213,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039960752003315,
      "loss": 4.9196,
      "mean_token_accuracy": 0.18758370578289033,
      "num_tokens": 70640142.0,
      "step": 32240
    },
    {
      "entropy": 5.353458833694458,
      "epoch": 3.2001786423183805,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003995780167268343,
      "loss": 5.0009,
      "mean_token_accuracy": 0.1824328377842903,
      "num_tokens": 70652062.0,
      "step": 32245
    },
    {
      "entropy": 5.3969214916229244,
      "epoch": 3.200674870980548,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003995485103312523,
      "loss": 5.037,
      "mean_token_accuracy": 0.17980303913354873,
      "num_tokens": 70666187.0,
      "step": 32250
    },
    {
      "entropy": 5.475476217269898,
      "epoch": 3.2011710996427154,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039951900084713555,
      "loss": 5.0952,
      "mean_token_accuracy": 0.177191823720932,
      "num_tokens": 70677432.0,
      "step": 32255
    },
    {
      "entropy": 5.403021812438965,
      "epoch": 3.201667328304883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003994894882752159,
      "loss": 4.9664,
      "mean_token_accuracy": 0.18055194467306138,
      "num_tokens": 70686963.0,
      "step": 32260
    },
    {
      "entropy": 5.472529315948487,
      "epoch": 3.2021635569670504,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039945997261622526,
      "loss": 5.1717,
      "mean_token_accuracy": 0.1675131604075432,
      "num_tokens": 70698345.0,
      "step": 32265
    },
    {
      "entropy": 5.393300771713257,
      "epoch": 3.202659785629218,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003994304538708954,
      "loss": 4.9978,
      "mean_token_accuracy": 0.1824351280927658,
      "num_tokens": 70709695.0,
      "step": 32270
    },
    {
      "entropy": 5.37525634765625,
      "epoch": 3.2031560142913853,
      "grad_norm": 1.125,
      "learning_rate": 0.0003994009320399584,
      "loss": 4.9339,
      "mean_token_accuracy": 0.1871471583843231,
      "num_tokens": 70720843.0,
      "step": 32275
    },
    {
      "entropy": 5.393763303756714,
      "epoch": 3.203652242953553,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039937140712414644,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1765233024954796,
      "num_tokens": 70731886.0,
      "step": 32280
    },
    {
      "entropy": 5.425012063980103,
      "epoch": 3.2041484716157207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003993418791241915,
      "loss": 5.0189,
      "mean_token_accuracy": 0.1831313744187355,
      "num_tokens": 70741838.0,
      "step": 32285
    },
    {
      "entropy": 5.395780801773071,
      "epoch": 3.204644700277888,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003993123480408259,
      "loss": 4.9429,
      "mean_token_accuracy": 0.18804978728294372,
      "num_tokens": 70751454.0,
      "step": 32290
    },
    {
      "entropy": 5.411750602722168,
      "epoch": 3.2051409289400556,
      "grad_norm": 1.125,
      "learning_rate": 0.00039928281387478207,
      "loss": 5.0334,
      "mean_token_accuracy": 0.17442502677440644,
      "num_tokens": 70762294.0,
      "step": 32295
    },
    {
      "entropy": 5.369723653793335,
      "epoch": 3.205637157602223,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003992532766267921,
      "loss": 4.9855,
      "mean_token_accuracy": 0.18350532948970794,
      "num_tokens": 70773186.0,
      "step": 32300
    },
    {
      "entropy": 5.423339414596557,
      "epoch": 3.2061333862643906,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003992237362975887,
      "loss": 5.0199,
      "mean_token_accuracy": 0.17795643359422683,
      "num_tokens": 70783984.0,
      "step": 32305
    },
    {
      "entropy": 5.408421754837036,
      "epoch": 3.206629614926558,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039919419288790423,
      "loss": 5.0047,
      "mean_token_accuracy": 0.17688938230276108,
      "num_tokens": 70795779.0,
      "step": 32310
    },
    {
      "entropy": 5.41849455833435,
      "epoch": 3.2071258435887255,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039916464639847145,
      "loss": 4.9865,
      "mean_token_accuracy": 0.17756694853305816,
      "num_tokens": 70807247.0,
      "step": 32315
    },
    {
      "entropy": 5.420690679550171,
      "epoch": 3.207622072250893,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003991350968300229,
      "loss": 5.0249,
      "mean_token_accuracy": 0.17634512931108476,
      "num_tokens": 70818507.0,
      "step": 32320
    },
    {
      "entropy": 5.400649976730347,
      "epoch": 3.208118300913061,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0003991055441832916,
      "loss": 5.0534,
      "mean_token_accuracy": 0.17606579810380935,
      "num_tokens": 70830055.0,
      "step": 32325
    },
    {
      "entropy": 5.485328340530396,
      "epoch": 3.2086145295752284,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039907598845901016,
      "loss": 5.119,
      "mean_token_accuracy": 0.17350534349679947,
      "num_tokens": 70841052.0,
      "step": 32330
    },
    {
      "entropy": 5.501479434967041,
      "epoch": 3.209110758237396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003990464296579114,
      "loss": 5.0641,
      "mean_token_accuracy": 0.180710169672966,
      "num_tokens": 70852148.0,
      "step": 32335
    },
    {
      "entropy": 5.401936101913452,
      "epoch": 3.2096069868995634,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003990168677807287,
      "loss": 5.0468,
      "mean_token_accuracy": 0.17433507442474366,
      "num_tokens": 70863998.0,
      "step": 32340
    },
    {
      "entropy": 5.419267177581787,
      "epoch": 3.210103215561731,
      "grad_norm": 1.0,
      "learning_rate": 0.0003989873028281948,
      "loss": 5.1022,
      "mean_token_accuracy": 0.17264826148748397,
      "num_tokens": 70874013.0,
      "step": 32345
    },
    {
      "entropy": 5.409296083450317,
      "epoch": 3.2105994442238983,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000398957734801043,
      "loss": 4.917,
      "mean_token_accuracy": 0.1909039333462715,
      "num_tokens": 70885600.0,
      "step": 32350
    },
    {
      "entropy": 5.315614175796509,
      "epoch": 3.211095672886066,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039892816370000657,
      "loss": 4.9081,
      "mean_token_accuracy": 0.1923968330025673,
      "num_tokens": 70896447.0,
      "step": 32355
    },
    {
      "entropy": 5.474783325195313,
      "epoch": 3.2115919015482333,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003988985895258186,
      "loss": 5.0789,
      "mean_token_accuracy": 0.1744958370923996,
      "num_tokens": 70907558.0,
      "step": 32360
    },
    {
      "entropy": 5.430703258514404,
      "epoch": 3.212088130210401,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039886901227921263,
      "loss": 5.0281,
      "mean_token_accuracy": 0.1773388147354126,
      "num_tokens": 70919687.0,
      "step": 32365
    },
    {
      "entropy": 5.382816362380981,
      "epoch": 3.2125843588725687,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003988394319609221,
      "loss": 4.9818,
      "mean_token_accuracy": 0.18256630152463912,
      "num_tokens": 70930417.0,
      "step": 32370
    },
    {
      "entropy": 5.389739036560059,
      "epoch": 3.213080587534736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003988098485716804,
      "loss": 4.9721,
      "mean_token_accuracy": 0.18268530666828156,
      "num_tokens": 70941721.0,
      "step": 32375
    },
    {
      "entropy": 5.390892362594604,
      "epoch": 3.2135768161969036,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039878026211222136,
      "loss": 4.9671,
      "mean_token_accuracy": 0.18564027100801467,
      "num_tokens": 70952616.0,
      "step": 32380
    },
    {
      "entropy": 5.348203134536743,
      "epoch": 3.214073044859071,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039875067258327853,
      "loss": 4.964,
      "mean_token_accuracy": 0.18085323721170427,
      "num_tokens": 70963236.0,
      "step": 32385
    },
    {
      "entropy": 5.426494216918945,
      "epoch": 3.2145692735212386,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003987210799855856,
      "loss": 5.0244,
      "mean_token_accuracy": 0.17496349066495895,
      "num_tokens": 70973376.0,
      "step": 32390
    },
    {
      "entropy": 5.403327941894531,
      "epoch": 3.215065502183406,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039869148431987653,
      "loss": 5.0336,
      "mean_token_accuracy": 0.17697494477033615,
      "num_tokens": 70984937.0,
      "step": 32395
    },
    {
      "entropy": 5.504658031463623,
      "epoch": 3.2155617308455735,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003986618855868852,
      "loss": 5.0996,
      "mean_token_accuracy": 0.17890945822000504,
      "num_tokens": 70997023.0,
      "step": 32400
    },
    {
      "entropy": 5.414682865142822,
      "epoch": 3.216057959507741,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003986322837873455,
      "loss": 4.9658,
      "mean_token_accuracy": 0.18440572768449784,
      "num_tokens": 71008304.0,
      "step": 32405
    },
    {
      "entropy": 5.362898540496826,
      "epoch": 3.216554188169909,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039860267892199167,
      "loss": 5.0315,
      "mean_token_accuracy": 0.17978009730577468,
      "num_tokens": 71020160.0,
      "step": 32410
    },
    {
      "entropy": 5.4450788497924805,
      "epoch": 3.2170504168320764,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039857307099155773,
      "loss": 5.0594,
      "mean_token_accuracy": 0.1816249817609787,
      "num_tokens": 71031701.0,
      "step": 32415
    },
    {
      "entropy": 5.447153186798095,
      "epoch": 3.217546645494244,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039854345999677777,
      "loss": 5.094,
      "mean_token_accuracy": 0.18044985681772233,
      "num_tokens": 71041955.0,
      "step": 32420
    },
    {
      "entropy": 5.425708818435669,
      "epoch": 3.2180428741564113,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039851384593838637,
      "loss": 5.1111,
      "mean_token_accuracy": 0.1764673709869385,
      "num_tokens": 71052884.0,
      "step": 32425
    },
    {
      "entropy": 5.436284732818604,
      "epoch": 3.218539102818579,
      "grad_norm": 1.0,
      "learning_rate": 0.00039848422881711765,
      "loss": 5.0,
      "mean_token_accuracy": 0.17708517462015153,
      "num_tokens": 71065091.0,
      "step": 32430
    },
    {
      "entropy": 5.492131185531616,
      "epoch": 3.2190353314807463,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003984546086337061,
      "loss": 5.0804,
      "mean_token_accuracy": 0.17566454410552979,
      "num_tokens": 71075823.0,
      "step": 32435
    },
    {
      "entropy": 5.440209817886353,
      "epoch": 3.2195315601429138,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039842498538888646,
      "loss": 5.0189,
      "mean_token_accuracy": 0.18146552592515947,
      "num_tokens": 71087408.0,
      "step": 32440
    },
    {
      "entropy": 5.4201624393463135,
      "epoch": 3.2200277888050812,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039839535908339297,
      "loss": 5.0006,
      "mean_token_accuracy": 0.18047993928194045,
      "num_tokens": 71097474.0,
      "step": 32445
    },
    {
      "entropy": 5.358831548690796,
      "epoch": 3.2205240174672487,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039836572971796045,
      "loss": 4.9514,
      "mean_token_accuracy": 0.1876937761902809,
      "num_tokens": 71107064.0,
      "step": 32450
    },
    {
      "entropy": 5.40272855758667,
      "epoch": 3.2210202461294166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003983360972933239,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18150490522384644,
      "num_tokens": 71118233.0,
      "step": 32455
    },
    {
      "entropy": 5.39283766746521,
      "epoch": 3.221516474791584,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003983064618102177,
      "loss": 4.9326,
      "mean_token_accuracy": 0.18819442987442017,
      "num_tokens": 71129922.0,
      "step": 32460
    },
    {
      "entropy": 5.368580436706543,
      "epoch": 3.2220127034537516,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003982768232693769,
      "loss": 5.0238,
      "mean_token_accuracy": 0.18090435415506362,
      "num_tokens": 71139073.0,
      "step": 32465
    },
    {
      "entropy": 5.379867792129517,
      "epoch": 3.222508932115919,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039824718167153666,
      "loss": 4.9433,
      "mean_token_accuracy": 0.18141741156578065,
      "num_tokens": 71152240.0,
      "step": 32470
    },
    {
      "entropy": 5.430435848236084,
      "epoch": 3.2230051607780865,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003982175370174319,
      "loss": 5.0118,
      "mean_token_accuracy": 0.17868751883506775,
      "num_tokens": 71163001.0,
      "step": 32475
    },
    {
      "entropy": 5.397028970718384,
      "epoch": 3.223501389440254,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003981878893077976,
      "loss": 4.9867,
      "mean_token_accuracy": 0.179961197078228,
      "num_tokens": 71172811.0,
      "step": 32480
    },
    {
      "entropy": 5.387010288238526,
      "epoch": 3.2239976181024215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039815823854336916,
      "loss": 4.9081,
      "mean_token_accuracy": 0.1899524837732315,
      "num_tokens": 71182276.0,
      "step": 32485
    },
    {
      "entropy": 5.378119468688965,
      "epoch": 3.224493846764589,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003981285847248818,
      "loss": 4.918,
      "mean_token_accuracy": 0.19269874691963196,
      "num_tokens": 71193154.0,
      "step": 32490
    },
    {
      "entropy": 5.4178032875061035,
      "epoch": 3.2249900754267564,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003980989278530709,
      "loss": 5.07,
      "mean_token_accuracy": 0.171643428504467,
      "num_tokens": 71204394.0,
      "step": 32495
    },
    {
      "entropy": 5.396539163589478,
      "epoch": 3.2254863040889243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003980692679286718,
      "loss": 4.8968,
      "mean_token_accuracy": 0.1838608831167221,
      "num_tokens": 71215311.0,
      "step": 32500
    },
    {
      "entropy": 5.418135261535644,
      "epoch": 3.225982532751092,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00039803960495241996,
      "loss": 5.0185,
      "mean_token_accuracy": 0.18511952310800553,
      "num_tokens": 71225800.0,
      "step": 32505
    },
    {
      "entropy": 5.388052272796631,
      "epoch": 3.2264787614132593,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003980099389250511,
      "loss": 4.9753,
      "mean_token_accuracy": 0.18176034092903137,
      "num_tokens": 71237846.0,
      "step": 32510
    },
    {
      "entropy": 5.418475484848022,
      "epoch": 3.2269749900754268,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039798026984730084,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18646654039621352,
      "num_tokens": 71248221.0,
      "step": 32515
    },
    {
      "entropy": 5.425748205184936,
      "epoch": 3.2274712187375942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003979505977199049,
      "loss": 5.023,
      "mean_token_accuracy": 0.1777804061770439,
      "num_tokens": 71259950.0,
      "step": 32520
    },
    {
      "entropy": 5.420040941238403,
      "epoch": 3.2279674473997617,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039792092254359903,
      "loss": 5.042,
      "mean_token_accuracy": 0.17674288898706436,
      "num_tokens": 71270541.0,
      "step": 32525
    },
    {
      "entropy": 5.433341407775879,
      "epoch": 3.228463676061929,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039789124431911923,
      "loss": 5.0205,
      "mean_token_accuracy": 0.18025438338518143,
      "num_tokens": 71280545.0,
      "step": 32530
    },
    {
      "entropy": 5.497116994857788,
      "epoch": 3.2289599047240967,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039786156304720126,
      "loss": 5.1021,
      "mean_token_accuracy": 0.17427823841571807,
      "num_tokens": 71292368.0,
      "step": 32535
    },
    {
      "entropy": 5.461621379852295,
      "epoch": 3.2294561333862646,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003978318787285814,
      "loss": 5.0002,
      "mean_token_accuracy": 0.17779071182012557,
      "num_tokens": 71302570.0,
      "step": 32540
    },
    {
      "entropy": 5.402691125869751,
      "epoch": 3.229952362048432,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039780219136399556,
      "loss": 5.0191,
      "mean_token_accuracy": 0.18352761417627333,
      "num_tokens": 71312629.0,
      "step": 32545
    },
    {
      "entropy": 5.448590087890625,
      "epoch": 3.2304485907105995,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039777250095418,
      "loss": 5.0809,
      "mean_token_accuracy": 0.17357377707958221,
      "num_tokens": 71324313.0,
      "step": 32550
    },
    {
      "entropy": 5.4552085399627686,
      "epoch": 3.230944819372767,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003977428074998709,
      "loss": 5.0875,
      "mean_token_accuracy": 0.1766663283109665,
      "num_tokens": 71336136.0,
      "step": 32555
    },
    {
      "entropy": 5.381837463378906,
      "epoch": 3.2314410480349345,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039771311100180476,
      "loss": 4.9691,
      "mean_token_accuracy": 0.1842123046517372,
      "num_tokens": 71347267.0,
      "step": 32560
    },
    {
      "entropy": 5.354067850112915,
      "epoch": 3.231937276697102,
      "grad_norm": 1.125,
      "learning_rate": 0.00039768341146071784,
      "loss": 4.9699,
      "mean_token_accuracy": 0.18698361217975618,
      "num_tokens": 71357268.0,
      "step": 32565
    },
    {
      "entropy": 5.5095538139343265,
      "epoch": 3.2324335053592694,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003976537088773467,
      "loss": 5.0999,
      "mean_token_accuracy": 0.1730469807982445,
      "num_tokens": 71368365.0,
      "step": 32570
    },
    {
      "entropy": 5.486351060867309,
      "epoch": 3.232929734021437,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039762400325242785,
      "loss": 5.1023,
      "mean_token_accuracy": 0.17571324855089188,
      "num_tokens": 71379623.0,
      "step": 32575
    },
    {
      "entropy": 5.385748243331909,
      "epoch": 3.2334259626836044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039759429458669805,
      "loss": 4.9984,
      "mean_token_accuracy": 0.18406662940979004,
      "num_tokens": 71388522.0,
      "step": 32580
    },
    {
      "entropy": 5.387484884262085,
      "epoch": 3.2339221913457723,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039756458288089377,
      "loss": 4.9722,
      "mean_token_accuracy": 0.1872837781906128,
      "num_tokens": 71399127.0,
      "step": 32585
    },
    {
      "entropy": 5.511310386657715,
      "epoch": 3.2344184200079398,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000397534868135752,
      "loss": 5.1163,
      "mean_token_accuracy": 0.17562416642904283,
      "num_tokens": 71410780.0,
      "step": 32590
    },
    {
      "entropy": 5.494064712524414,
      "epoch": 3.2349146486701073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039750515035200947,
      "loss": 5.1256,
      "mean_token_accuracy": 0.16729025691747665,
      "num_tokens": 71422887.0,
      "step": 32595
    },
    {
      "entropy": 5.3802577495574955,
      "epoch": 3.2354108773322747,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039747542953040327,
      "loss": 5.0289,
      "mean_token_accuracy": 0.18115823715925217,
      "num_tokens": 71432525.0,
      "step": 32600
    },
    {
      "entropy": 5.380505084991455,
      "epoch": 3.235907105994442,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003974457056716703,
      "loss": 5.0561,
      "mean_token_accuracy": 0.18175048381090164,
      "num_tokens": 71444802.0,
      "step": 32605
    },
    {
      "entropy": 5.380873394012451,
      "epoch": 3.2364033346566097,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003974159787765477,
      "loss": 4.9314,
      "mean_token_accuracy": 0.1828133851289749,
      "num_tokens": 71455978.0,
      "step": 32610
    },
    {
      "entropy": 5.340788459777832,
      "epoch": 3.236899563318777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003973862488457726,
      "loss": 4.8965,
      "mean_token_accuracy": 0.18329519629478455,
      "num_tokens": 71466981.0,
      "step": 32615
    },
    {
      "entropy": 5.429397773742676,
      "epoch": 3.2373957919809446,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003973565158800822,
      "loss": 5.0479,
      "mean_token_accuracy": 0.17249325960874556,
      "num_tokens": 71478856.0,
      "step": 32620
    },
    {
      "entropy": 5.483404111862183,
      "epoch": 3.2378920206431125,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039732677988021394,
      "loss": 5.0647,
      "mean_token_accuracy": 0.17882877141237258,
      "num_tokens": 71489449.0,
      "step": 32625
    },
    {
      "entropy": 5.421832180023193,
      "epoch": 3.23838824930528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003972970408469051,
      "loss": 5.0599,
      "mean_token_accuracy": 0.17220931202173234,
      "num_tokens": 71499687.0,
      "step": 32630
    },
    {
      "entropy": 5.473094749450683,
      "epoch": 3.2388844779674475,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003972672987808932,
      "loss": 5.011,
      "mean_token_accuracy": 0.17936757653951646,
      "num_tokens": 71510087.0,
      "step": 32635
    },
    {
      "entropy": 5.415521764755249,
      "epoch": 3.239380706629615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039723755368291564,
      "loss": 4.9805,
      "mean_token_accuracy": 0.18157838732004167,
      "num_tokens": 71520504.0,
      "step": 32640
    },
    {
      "entropy": 5.449507236480713,
      "epoch": 3.2398769352917824,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003972078055537102,
      "loss": 5.056,
      "mean_token_accuracy": 0.17360638231039047,
      "num_tokens": 71532342.0,
      "step": 32645
    },
    {
      "entropy": 5.485736751556397,
      "epoch": 3.24037316395395,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003971780543940146,
      "loss": 5.0339,
      "mean_token_accuracy": 0.1740538537502289,
      "num_tokens": 71544523.0,
      "step": 32650
    },
    {
      "entropy": 5.426730108261109,
      "epoch": 3.2408693926161174,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003971483002045665,
      "loss": 5.0179,
      "mean_token_accuracy": 0.1811053767800331,
      "num_tokens": 71555162.0,
      "step": 32655
    },
    {
      "entropy": 5.402161312103272,
      "epoch": 3.241365621278285,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003971185429861037,
      "loss": 4.9923,
      "mean_token_accuracy": 0.18514541387557984,
      "num_tokens": 71566407.0,
      "step": 32660
    },
    {
      "entropy": 5.45770583152771,
      "epoch": 3.2418618499404523,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039708878273936424,
      "loss": 5.0363,
      "mean_token_accuracy": 0.17909489572048187,
      "num_tokens": 71577328.0,
      "step": 32665
    },
    {
      "entropy": 5.459929466247559,
      "epoch": 3.2423580786026203,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003970590194650859,
      "loss": 5.1474,
      "mean_token_accuracy": 0.17624386996030808,
      "num_tokens": 71587716.0,
      "step": 32670
    },
    {
      "entropy": 5.384181547164917,
      "epoch": 3.2428543072647877,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003970292531640071,
      "loss": 4.9881,
      "mean_token_accuracy": 0.17942506521940232,
      "num_tokens": 71599575.0,
      "step": 32675
    },
    {
      "entropy": 5.45197319984436,
      "epoch": 3.243350535926955,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00039699948383686563,
      "loss": 5.0495,
      "mean_token_accuracy": 0.17880223244428634,
      "num_tokens": 71610925.0,
      "step": 32680
    },
    {
      "entropy": 5.4288942337036135,
      "epoch": 3.2438467645891227,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003969697114843998,
      "loss": 4.9895,
      "mean_token_accuracy": 0.18304127156734468,
      "num_tokens": 71620250.0,
      "step": 32685
    },
    {
      "entropy": 5.4707855701446535,
      "epoch": 3.24434299325129,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039693993610734807,
      "loss": 5.1207,
      "mean_token_accuracy": 0.17330981343984603,
      "num_tokens": 71632633.0,
      "step": 32690
    },
    {
      "entropy": 5.4469499588012695,
      "epoch": 3.2448392219134576,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039691015770644875,
      "loss": 5.0863,
      "mean_token_accuracy": 0.17386470586061478,
      "num_tokens": 71644031.0,
      "step": 32695
    },
    {
      "entropy": 5.456206321716309,
      "epoch": 3.245335450575625,
      "grad_norm": 1.125,
      "learning_rate": 0.00039688037628244,
      "loss": 5.0693,
      "mean_token_accuracy": 0.18834338933229447,
      "num_tokens": 71654785.0,
      "step": 32700
    },
    {
      "entropy": 5.464949750900269,
      "epoch": 3.2458316792377926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039685059183606065,
      "loss": 5.0237,
      "mean_token_accuracy": 0.18025397062301635,
      "num_tokens": 71667014.0,
      "step": 32705
    },
    {
      "entropy": 5.435647010803223,
      "epoch": 3.2463279078999605,
      "grad_norm": 1.125,
      "learning_rate": 0.00039682080436804926,
      "loss": 5.0407,
      "mean_token_accuracy": 0.1881081208586693,
      "num_tokens": 71677157.0,
      "step": 32710
    },
    {
      "entropy": 5.478702592849731,
      "epoch": 3.246824136562128,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003967910138791442,
      "loss": 5.0669,
      "mean_token_accuracy": 0.16991938203573226,
      "num_tokens": 71688542.0,
      "step": 32715
    },
    {
      "entropy": 5.369105863571167,
      "epoch": 3.2473203652242955,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039676122037008473,
      "loss": 5.0138,
      "mean_token_accuracy": 0.1811184450984001,
      "num_tokens": 71699364.0,
      "step": 32720
    },
    {
      "entropy": 5.429362869262695,
      "epoch": 3.247816593886463,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003967314238416092,
      "loss": 5.053,
      "mean_token_accuracy": 0.17490830421447753,
      "num_tokens": 71710248.0,
      "step": 32725
    },
    {
      "entropy": 5.457511901855469,
      "epoch": 3.2483128225486304,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003967016242944567,
      "loss": 5.0378,
      "mean_token_accuracy": 0.17371099889278413,
      "num_tokens": 71720616.0,
      "step": 32730
    },
    {
      "entropy": 5.510946893692017,
      "epoch": 3.248809051210798,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003966718217293661,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1812607854604721,
      "num_tokens": 71731536.0,
      "step": 32735
    },
    {
      "entropy": 5.39995641708374,
      "epoch": 3.2493052798729654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039664201614707657,
      "loss": 5.0594,
      "mean_token_accuracy": 0.18152732104063035,
      "num_tokens": 71742459.0,
      "step": 32740
    },
    {
      "entropy": 5.33941068649292,
      "epoch": 3.249801508535133,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003966122075483271,
      "loss": 5.0269,
      "mean_token_accuracy": 0.17553159296512605,
      "num_tokens": 71753611.0,
      "step": 32745
    },
    {
      "entropy": 5.422653675079346,
      "epoch": 3.2502977371973003,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039658239593385697,
      "loss": 4.9781,
      "mean_token_accuracy": 0.18298117220401763,
      "num_tokens": 71764924.0,
      "step": 32750
    },
    {
      "entropy": 5.4115385055542,
      "epoch": 3.2507939658594682,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039655258130440543,
      "loss": 5.0643,
      "mean_token_accuracy": 0.1713058680295944,
      "num_tokens": 71776078.0,
      "step": 32755
    },
    {
      "entropy": 5.456792163848877,
      "epoch": 3.2512901945216357,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003965227636607117,
      "loss": 5.0029,
      "mean_token_accuracy": 0.17686382681131363,
      "num_tokens": 71785988.0,
      "step": 32760
    },
    {
      "entropy": 5.382821798324585,
      "epoch": 3.251786423183803,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039649294300351547,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18250346481800078,
      "num_tokens": 71796382.0,
      "step": 32765
    },
    {
      "entropy": 5.374724912643432,
      "epoch": 3.2522826518459707,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003964631193335559,
      "loss": 4.9677,
      "mean_token_accuracy": 0.179609814286232,
      "num_tokens": 71806533.0,
      "step": 32770
    },
    {
      "entropy": 5.389789772033692,
      "epoch": 3.252778880508138,
      "grad_norm": 1.0,
      "learning_rate": 0.00039643329265157275,
      "loss": 5.0042,
      "mean_token_accuracy": 0.18482969105243682,
      "num_tokens": 71816893.0,
      "step": 32775
    },
    {
      "entropy": 5.48032579421997,
      "epoch": 3.2532751091703056,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039640346295830557,
      "loss": 5.0129,
      "mean_token_accuracy": 0.17910777926445007,
      "num_tokens": 71827366.0,
      "step": 32780
    },
    {
      "entropy": 5.472847414016724,
      "epoch": 3.253771337832473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003963736302544942,
      "loss": 5.0432,
      "mean_token_accuracy": 0.17682681679725648,
      "num_tokens": 71837896.0,
      "step": 32785
    },
    {
      "entropy": 5.3697723865509035,
      "epoch": 3.2542675664946406,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039634379454087814,
      "loss": 5.0401,
      "mean_token_accuracy": 0.1721576675772667,
      "num_tokens": 71850583.0,
      "step": 32790
    },
    {
      "entropy": 5.407357311248779,
      "epoch": 3.2547637951568085,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003963139558181977,
      "loss": 5.0469,
      "mean_token_accuracy": 0.1802523121237755,
      "num_tokens": 71860392.0,
      "step": 32795
    },
    {
      "entropy": 5.409395980834961,
      "epoch": 3.255260023818976,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003962841140871924,
      "loss": 5.0276,
      "mean_token_accuracy": 0.18002101331949233,
      "num_tokens": 71871323.0,
      "step": 32800
    },
    {
      "entropy": 5.3725879192352295,
      "epoch": 3.2557562524811434,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003962542693486024,
      "loss": 4.9615,
      "mean_token_accuracy": 0.18960412591695786,
      "num_tokens": 71881612.0,
      "step": 32805
    },
    {
      "entropy": 5.500239276885987,
      "epoch": 3.256252481143311,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003962244216031679,
      "loss": 5.1673,
      "mean_token_accuracy": 0.1730325251817703,
      "num_tokens": 71893266.0,
      "step": 32810
    },
    {
      "entropy": 5.402393198013305,
      "epoch": 3.2567487098054784,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000396194570851629,
      "loss": 4.9863,
      "mean_token_accuracy": 0.18895207345485687,
      "num_tokens": 71904978.0,
      "step": 32815
    },
    {
      "entropy": 5.50586428642273,
      "epoch": 3.257244938467646,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039616471709472573,
      "loss": 5.0712,
      "mean_token_accuracy": 0.176172935962677,
      "num_tokens": 71915097.0,
      "step": 32820
    },
    {
      "entropy": 5.43486008644104,
      "epoch": 3.2577411671298133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003961348603331987,
      "loss": 4.9996,
      "mean_token_accuracy": 0.1735418125987053,
      "num_tokens": 71924672.0,
      "step": 32825
    },
    {
      "entropy": 5.295911169052124,
      "epoch": 3.258237395791981,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00039610500056778817,
      "loss": 4.9261,
      "mean_token_accuracy": 0.1802048549056053,
      "num_tokens": 71936122.0,
      "step": 32830
    },
    {
      "entropy": 5.358137416839599,
      "epoch": 3.2587336244541483,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003960751377992344,
      "loss": 4.9333,
      "mean_token_accuracy": 0.1882056102156639,
      "num_tokens": 71946601.0,
      "step": 32835
    },
    {
      "entropy": 5.451126527786255,
      "epoch": 3.259229853116316,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039604527202827825,
      "loss": 5.0696,
      "mean_token_accuracy": 0.17898847609758378,
      "num_tokens": 71958892.0,
      "step": 32840
    },
    {
      "entropy": 5.3926409721374515,
      "epoch": 3.2597260817784837,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039601540325566014,
      "loss": 4.9859,
      "mean_token_accuracy": 0.17851637303829193,
      "num_tokens": 71969183.0,
      "step": 32845
    },
    {
      "entropy": 5.441771507263184,
      "epoch": 3.260222310440651,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003959855314821208,
      "loss": 5.0086,
      "mean_token_accuracy": 0.18106933683156967,
      "num_tokens": 71981115.0,
      "step": 32850
    },
    {
      "entropy": 5.438884019851685,
      "epoch": 3.2607185391028186,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000395955656708401,
      "loss": 5.0549,
      "mean_token_accuracy": 0.17858253717422484,
      "num_tokens": 71991677.0,
      "step": 32855
    },
    {
      "entropy": 5.404206705093384,
      "epoch": 3.261214767764986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039592577893524155,
      "loss": 5.0412,
      "mean_token_accuracy": 0.18588811606168748,
      "num_tokens": 72003995.0,
      "step": 32860
    },
    {
      "entropy": 5.451698160171508,
      "epoch": 3.2617109964271536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039589589816338335,
      "loss": 5.0287,
      "mean_token_accuracy": 0.18239855468273164,
      "num_tokens": 72014052.0,
      "step": 32865
    },
    {
      "entropy": 5.42232232093811,
      "epoch": 3.262207225089321,
      "grad_norm": 1.0,
      "learning_rate": 0.0003958660143935674,
      "loss": 4.9589,
      "mean_token_accuracy": 0.17814176976680757,
      "num_tokens": 72024756.0,
      "step": 32870
    },
    {
      "entropy": 5.3650003433227536,
      "epoch": 3.2627034537514885,
      "grad_norm": 1.0,
      "learning_rate": 0.0003958361276265347,
      "loss": 4.9774,
      "mean_token_accuracy": 0.1801770880818367,
      "num_tokens": 72035883.0,
      "step": 32875
    },
    {
      "entropy": 5.428834772109985,
      "epoch": 3.2631996824136564,
      "grad_norm": 1.0,
      "learning_rate": 0.00039580623786302643,
      "loss": 4.9863,
      "mean_token_accuracy": 0.187019644677639,
      "num_tokens": 72047296.0,
      "step": 32880
    },
    {
      "entropy": 5.410967016220093,
      "epoch": 3.263695911075824,
      "grad_norm": 1.125,
      "learning_rate": 0.00039577634510378374,
      "loss": 5.0037,
      "mean_token_accuracy": 0.17330461442470552,
      "num_tokens": 72060163.0,
      "step": 32885
    },
    {
      "entropy": 5.4396766185760494,
      "epoch": 3.2641921397379914,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000395746449349548,
      "loss": 5.059,
      "mean_token_accuracy": 0.1731042355298996,
      "num_tokens": 72070384.0,
      "step": 32890
    },
    {
      "entropy": 5.420939874649048,
      "epoch": 3.264688368400159,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003957165506010605,
      "loss": 4.9646,
      "mean_token_accuracy": 0.18138243705034257,
      "num_tokens": 72080997.0,
      "step": 32895
    },
    {
      "entropy": 5.462356472015381,
      "epoch": 3.2651845970623263,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003956866488590626,
      "loss": 5.0176,
      "mean_token_accuracy": 0.1835634797811508,
      "num_tokens": 72091161.0,
      "step": 32900
    },
    {
      "entropy": 5.3758077144622805,
      "epoch": 3.265680825724494,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003956567441242959,
      "loss": 5.0047,
      "mean_token_accuracy": 0.1792465165257454,
      "num_tokens": 72101807.0,
      "step": 32905
    },
    {
      "entropy": 5.464919662475586,
      "epoch": 3.2661770543866613,
      "grad_norm": 0.953125,
      "learning_rate": 0.000395626836397502,
      "loss": 5.0719,
      "mean_token_accuracy": 0.1741136610507965,
      "num_tokens": 72112800.0,
      "step": 32910
    },
    {
      "entropy": 5.390661573410034,
      "epoch": 3.2666732830488288,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003955969256794224,
      "loss": 4.9876,
      "mean_token_accuracy": 0.17803874760866165,
      "num_tokens": 72122391.0,
      "step": 32915
    },
    {
      "entropy": 5.41625189781189,
      "epoch": 3.2671695117109962,
      "grad_norm": 1.0,
      "learning_rate": 0.00039556701197079896,
      "loss": 5.0317,
      "mean_token_accuracy": 0.17980241030454636,
      "num_tokens": 72133145.0,
      "step": 32920
    },
    {
      "entropy": 5.523226594924926,
      "epoch": 3.2676657403731637,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00039553709527237344,
      "loss": 5.0709,
      "mean_token_accuracy": 0.1697484239935875,
      "num_tokens": 72146126.0,
      "step": 32925
    },
    {
      "entropy": 5.387266159057617,
      "epoch": 3.2681619690353316,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003955071755848876,
      "loss": 4.9641,
      "mean_token_accuracy": 0.1819107487797737,
      "num_tokens": 72157325.0,
      "step": 32930
    },
    {
      "entropy": 5.302313232421875,
      "epoch": 3.268658197697499,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039547725290908347,
      "loss": 4.8853,
      "mean_token_accuracy": 0.18750304281711577,
      "num_tokens": 72168601.0,
      "step": 32935
    },
    {
      "entropy": 5.371862411499023,
      "epoch": 3.2691544263596666,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00039544732724570313,
      "loss": 4.985,
      "mean_token_accuracy": 0.18479963392019272,
      "num_tokens": 72179845.0,
      "step": 32940
    },
    {
      "entropy": 5.414711666107178,
      "epoch": 3.269650655021834,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003954173985954887,
      "loss": 4.9219,
      "mean_token_accuracy": 0.1784079447388649,
      "num_tokens": 72191135.0,
      "step": 32945
    },
    {
      "entropy": 5.40353798866272,
      "epoch": 3.2701468836840015,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003953874669591822,
      "loss": 4.9927,
      "mean_token_accuracy": 0.1816827267408371,
      "num_tokens": 72201551.0,
      "step": 32950
    },
    {
      "entropy": 5.383181476593018,
      "epoch": 3.270643112346169,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039535753233752583,
      "loss": 4.9943,
      "mean_token_accuracy": 0.18491510003805162,
      "num_tokens": 72211983.0,
      "step": 32955
    },
    {
      "entropy": 5.450747728347778,
      "epoch": 3.2711393410083365,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003953275947312621,
      "loss": 5.0397,
      "mean_token_accuracy": 0.17290544509887695,
      "num_tokens": 72223374.0,
      "step": 32960
    },
    {
      "entropy": 5.346310186386108,
      "epoch": 3.2716355696705044,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003952976541411332,
      "loss": 4.9913,
      "mean_token_accuracy": 0.1879296973347664,
      "num_tokens": 72233878.0,
      "step": 32965
    },
    {
      "entropy": 5.412832641601563,
      "epoch": 3.272131798332672,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039526771056788186,
      "loss": 5.0745,
      "mean_token_accuracy": 0.1838723286986351,
      "num_tokens": 72246157.0,
      "step": 32970
    },
    {
      "entropy": 5.435351705551147,
      "epoch": 3.2726280269948393,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039523776401225043,
      "loss": 4.9893,
      "mean_token_accuracy": 0.18433370888233186,
      "num_tokens": 72256805.0,
      "step": 32975
    },
    {
      "entropy": 5.439369058609008,
      "epoch": 3.273124255657007,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039520781447498145,
      "loss": 5.0197,
      "mean_token_accuracy": 0.17671131789684297,
      "num_tokens": 72267026.0,
      "step": 32980
    },
    {
      "entropy": 5.403189849853516,
      "epoch": 3.2736204843191743,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003951778619568177,
      "loss": 5.0128,
      "mean_token_accuracy": 0.18271225541830063,
      "num_tokens": 72277989.0,
      "step": 32985
    },
    {
      "entropy": 5.3562780857086185,
      "epoch": 3.2741167129813418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000395147906458502,
      "loss": 4.936,
      "mean_token_accuracy": 0.18185906559228898,
      "num_tokens": 72287989.0,
      "step": 32990
    },
    {
      "entropy": 5.35245680809021,
      "epoch": 3.2746129416435092,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039511794798077705,
      "loss": 4.9861,
      "mean_token_accuracy": 0.1780322313308716,
      "num_tokens": 72298718.0,
      "step": 32995
    },
    {
      "entropy": 5.388212108612061,
      "epoch": 3.2751091703056767,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003950879865243858,
      "loss": 4.9758,
      "mean_token_accuracy": 0.18596740961074829,
      "num_tokens": 72309217.0,
      "step": 33000
    },
    {
      "epoch": 3.2751091703056767,
      "eval_entropy": 5.263041518141804,
      "eval_loss": 5.185599327087402,
      "eval_mean_token_accuracy": 0.17721146509054864,
      "eval_num_tokens": 72309217.0,
      "eval_runtime": 26.5328,
      "eval_samples_per_second": 1268.958,
      "eval_steps_per_second": 158.634,
      "step": 33000
    },
    {
      "entropy": 5.450331926345825,
      "epoch": 3.275605398967844,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039505802209007136,
      "loss": 5.0252,
      "mean_token_accuracy": 0.1795649141073227,
      "num_tokens": 72319809.0,
      "step": 33005
    },
    {
      "entropy": 5.4205358028411865,
      "epoch": 3.2761016276300117,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003950280546785766,
      "loss": 5.054,
      "mean_token_accuracy": 0.17829269617795945,
      "num_tokens": 72330616.0,
      "step": 33010
    },
    {
      "entropy": 5.464061641693116,
      "epoch": 3.2765978562921796,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003949980842906447,
      "loss": 5.0306,
      "mean_token_accuracy": 0.17963909953832627,
      "num_tokens": 72339749.0,
      "step": 33015
    },
    {
      "entropy": 5.400103950500489,
      "epoch": 3.277094084954347,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000394968110927019,
      "loss": 4.9924,
      "mean_token_accuracy": 0.18314148336648942,
      "num_tokens": 72350653.0,
      "step": 33020
    },
    {
      "entropy": 5.399258995056153,
      "epoch": 3.2775903136165145,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039493813458844256,
      "loss": 5.0016,
      "mean_token_accuracy": 0.18214128762483597,
      "num_tokens": 72360261.0,
      "step": 33025
    },
    {
      "entropy": 5.359872245788575,
      "epoch": 3.278086542278682,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00039490815527565883,
      "loss": 4.9987,
      "mean_token_accuracy": 0.18002621531486512,
      "num_tokens": 72372180.0,
      "step": 33030
    },
    {
      "entropy": 5.354433584213257,
      "epoch": 3.2785827709408495,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039487817298941124,
      "loss": 4.9856,
      "mean_token_accuracy": 0.17671619802713395,
      "num_tokens": 72383324.0,
      "step": 33035
    },
    {
      "entropy": 5.367968559265137,
      "epoch": 3.279078999603017,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003948481877304433,
      "loss": 5.0043,
      "mean_token_accuracy": 0.177391354739666,
      "num_tokens": 72395014.0,
      "step": 33040
    },
    {
      "entropy": 5.482691144943237,
      "epoch": 3.2795752282651844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003948181994994985,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17773682326078416,
      "num_tokens": 72406806.0,
      "step": 33045
    },
    {
      "entropy": 5.430766201019287,
      "epoch": 3.2800714569273524,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003947882082973206,
      "loss": 4.9895,
      "mean_token_accuracy": 0.17428090721368789,
      "num_tokens": 72417002.0,
      "step": 33050
    },
    {
      "entropy": 5.4009765625,
      "epoch": 3.28056768558952,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039475821412465317,
      "loss": 5.0478,
      "mean_token_accuracy": 0.17725879251956939,
      "num_tokens": 72427107.0,
      "step": 33055
    },
    {
      "entropy": 5.34635100364685,
      "epoch": 3.2810639142516873,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039472821698224023,
      "loss": 4.9245,
      "mean_token_accuracy": 0.1925074815750122,
      "num_tokens": 72438737.0,
      "step": 33060
    },
    {
      "entropy": 5.323901224136352,
      "epoch": 3.281560142913855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003946982168708254,
      "loss": 4.9256,
      "mean_token_accuracy": 0.18361760526895524,
      "num_tokens": 72449070.0,
      "step": 33065
    },
    {
      "entropy": 5.401172637939453,
      "epoch": 3.2820563715760223,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003946682137911527,
      "loss": 5.0398,
      "mean_token_accuracy": 0.17972489148378373,
      "num_tokens": 72459437.0,
      "step": 33070
    },
    {
      "entropy": 5.471559953689575,
      "epoch": 3.2825526002381897,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039463820774396623,
      "loss": 5.0275,
      "mean_token_accuracy": 0.1796439617872238,
      "num_tokens": 72470804.0,
      "step": 33075
    },
    {
      "entropy": 5.403841876983643,
      "epoch": 3.283048828900357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039460819873000995,
      "loss": 4.9999,
      "mean_token_accuracy": 0.1749626025557518,
      "num_tokens": 72482400.0,
      "step": 33080
    },
    {
      "entropy": 5.328671884536743,
      "epoch": 3.2835450575625247,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039457818675002807,
      "loss": 4.9073,
      "mean_token_accuracy": 0.18688919246196747,
      "num_tokens": 72492697.0,
      "step": 33085
    },
    {
      "entropy": 5.3884107112884525,
      "epoch": 3.284041286224692,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003945481718047648,
      "loss": 4.9918,
      "mean_token_accuracy": 0.18195040971040727,
      "num_tokens": 72503913.0,
      "step": 33090
    },
    {
      "entropy": 5.4592526912689205,
      "epoch": 3.2845375148868596,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003945181538949646,
      "loss": 5.0162,
      "mean_token_accuracy": 0.18387335389852524,
      "num_tokens": 72515018.0,
      "step": 33095
    },
    {
      "entropy": 5.334270811080932,
      "epoch": 3.2850337435490276,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003944881330213716,
      "loss": 5.0255,
      "mean_token_accuracy": 0.17837542444467544,
      "num_tokens": 72525256.0,
      "step": 33100
    },
    {
      "entropy": 5.336289930343628,
      "epoch": 3.285529972211195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003944581091847304,
      "loss": 4.9363,
      "mean_token_accuracy": 0.18619169145822526,
      "num_tokens": 72535839.0,
      "step": 33105
    },
    {
      "entropy": 5.498153972625732,
      "epoch": 3.2860262008733625,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039442808238578553,
      "loss": 5.0646,
      "mean_token_accuracy": 0.18600267469882964,
      "num_tokens": 72546808.0,
      "step": 33110
    },
    {
      "entropy": 5.512692308425903,
      "epoch": 3.28652242953553,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039439805262528153,
      "loss": 5.0701,
      "mean_token_accuracy": 0.1778644785284996,
      "num_tokens": 72559268.0,
      "step": 33115
    },
    {
      "entropy": 5.454471206665039,
      "epoch": 3.2870186581976975,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003943680199039631,
      "loss": 5.0499,
      "mean_token_accuracy": 0.16840217113494874,
      "num_tokens": 72570466.0,
      "step": 33120
    },
    {
      "entropy": 5.422110080718994,
      "epoch": 3.287514886859865,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039433798422257505,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18483609557151795,
      "num_tokens": 72581373.0,
      "step": 33125
    },
    {
      "entropy": 5.485711145401001,
      "epoch": 3.2880111155220324,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003943079455818621,
      "loss": 5.085,
      "mean_token_accuracy": 0.17438378036022187,
      "num_tokens": 72592661.0,
      "step": 33130
    },
    {
      "entropy": 5.309210920333863,
      "epoch": 3.2885073441842003,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039427790398256925,
      "loss": 4.9361,
      "mean_token_accuracy": 0.19165662974119185,
      "num_tokens": 72603387.0,
      "step": 33135
    },
    {
      "entropy": 5.3597334861755375,
      "epoch": 3.289003572846368,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039424785942544126,
      "loss": 4.9896,
      "mean_token_accuracy": 0.18493088632822036,
      "num_tokens": 72614006.0,
      "step": 33140
    },
    {
      "entropy": 5.373026752471924,
      "epoch": 3.2894998015085353,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003942178119112234,
      "loss": 4.9399,
      "mean_token_accuracy": 0.18318322747945787,
      "num_tokens": 72625447.0,
      "step": 33145
    },
    {
      "entropy": 5.388800048828125,
      "epoch": 3.2899960301707027,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003941877614406607,
      "loss": 4.9899,
      "mean_token_accuracy": 0.18338817358016968,
      "num_tokens": 72636965.0,
      "step": 33150
    },
    {
      "entropy": 5.377870845794678,
      "epoch": 3.29049225883287,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003941577080144984,
      "loss": 4.9477,
      "mean_token_accuracy": 0.18589526265859604,
      "num_tokens": 72648161.0,
      "step": 33155
    },
    {
      "entropy": 5.386142063140869,
      "epoch": 3.2909884874950377,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003941276516334816,
      "loss": 4.9323,
      "mean_token_accuracy": 0.18461330831050873,
      "num_tokens": 72658253.0,
      "step": 33160
    },
    {
      "entropy": 5.364346361160278,
      "epoch": 3.291484716157205,
      "grad_norm": 1.0,
      "learning_rate": 0.00039409759229835586,
      "loss": 4.9592,
      "mean_token_accuracy": 0.1861182391643524,
      "num_tokens": 72669583.0,
      "step": 33165
    },
    {
      "entropy": 5.326032495498657,
      "epoch": 3.2919809448193726,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003940675300098663,
      "loss": 4.9174,
      "mean_token_accuracy": 0.18233090341091157,
      "num_tokens": 72680367.0,
      "step": 33170
    },
    {
      "entropy": 5.389262294769287,
      "epoch": 3.29247717348154,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003940374647687586,
      "loss": 5.0064,
      "mean_token_accuracy": 0.17794880121946335,
      "num_tokens": 72692437.0,
      "step": 33175
    },
    {
      "entropy": 5.377025461196899,
      "epoch": 3.2929734021437076,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003940073965757784,
      "loss": 5.0285,
      "mean_token_accuracy": 0.17572114765644073,
      "num_tokens": 72704286.0,
      "step": 33180
    },
    {
      "entropy": 5.453970813751221,
      "epoch": 3.2934696308058755,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003939773254316711,
      "loss": 5.0646,
      "mean_token_accuracy": 0.17416924238204956,
      "num_tokens": 72714767.0,
      "step": 33185
    },
    {
      "entropy": 5.444849824905395,
      "epoch": 3.293965859468043,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003939472513371826,
      "loss": 5.064,
      "mean_token_accuracy": 0.1791225790977478,
      "num_tokens": 72725110.0,
      "step": 33190
    },
    {
      "entropy": 5.429011821746826,
      "epoch": 3.2944620881302105,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003939171742930584,
      "loss": 5.0158,
      "mean_token_accuracy": 0.18272936791181565,
      "num_tokens": 72737150.0,
      "step": 33195
    },
    {
      "entropy": 5.43204140663147,
      "epoch": 3.294958316792378,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039388709430004465,
      "loss": 4.961,
      "mean_token_accuracy": 0.187857785820961,
      "num_tokens": 72747523.0,
      "step": 33200
    },
    {
      "entropy": 5.442828702926636,
      "epoch": 3.2954545454545454,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039385701135888715,
      "loss": 5.0544,
      "mean_token_accuracy": 0.18068304806947708,
      "num_tokens": 72760163.0,
      "step": 33205
    },
    {
      "entropy": 5.473714065551758,
      "epoch": 3.295950774116713,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003938269254703318,
      "loss": 5.0636,
      "mean_token_accuracy": 0.1756466507911682,
      "num_tokens": 72770031.0,
      "step": 33210
    },
    {
      "entropy": 5.427866983413696,
      "epoch": 3.2964470027788804,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039379683663512475,
      "loss": 4.9165,
      "mean_token_accuracy": 0.18576344102621078,
      "num_tokens": 72780187.0,
      "step": 33215
    },
    {
      "entropy": 5.38816647529602,
      "epoch": 3.2969432314410483,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003937667448540123,
      "loss": 5.0288,
      "mean_token_accuracy": 0.17717836648225785,
      "num_tokens": 72790486.0,
      "step": 33220
    },
    {
      "entropy": 5.353249263763428,
      "epoch": 3.2974394601032158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039373665012774027,
      "loss": 4.8827,
      "mean_token_accuracy": 0.18695154041051865,
      "num_tokens": 72800086.0,
      "step": 33225
    },
    {
      "entropy": 5.449350833892822,
      "epoch": 3.2979356887653832,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039370655245705527,
      "loss": 5.0933,
      "mean_token_accuracy": 0.17352961897850036,
      "num_tokens": 72811018.0,
      "step": 33230
    },
    {
      "entropy": 5.406621646881104,
      "epoch": 3.2984319174275507,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039367645184270364,
      "loss": 5.0069,
      "mean_token_accuracy": 0.18051015138626098,
      "num_tokens": 72821792.0,
      "step": 33235
    },
    {
      "entropy": 5.417872285842895,
      "epoch": 3.298928146089718,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003936463482854317,
      "loss": 5.1009,
      "mean_token_accuracy": 0.18401477187871934,
      "num_tokens": 72833782.0,
      "step": 33240
    },
    {
      "entropy": 5.442956256866455,
      "epoch": 3.2994243747518857,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003936162417859858,
      "loss": 5.0073,
      "mean_token_accuracy": 0.18474665135145188,
      "num_tokens": 72844480.0,
      "step": 33245
    },
    {
      "entropy": 5.47118911743164,
      "epoch": 3.299920603414053,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039358613234511286,
      "loss": 5.1515,
      "mean_token_accuracy": 0.17396641820669173,
      "num_tokens": 72855855.0,
      "step": 33250
    },
    {
      "entropy": 5.40775842666626,
      "epoch": 3.3004168320762206,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039355601996355944,
      "loss": 4.9634,
      "mean_token_accuracy": 0.17653957456350328,
      "num_tokens": 72866444.0,
      "step": 33255
    },
    {
      "entropy": 5.376863718032837,
      "epoch": 3.300913060738388,
      "grad_norm": 1.0625,
      "learning_rate": 0.000393525904642072,
      "loss": 5.0047,
      "mean_token_accuracy": 0.17708057463169097,
      "num_tokens": 72877083.0,
      "step": 33260
    },
    {
      "entropy": 5.343069839477539,
      "epoch": 3.3014092894005556,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039349578638139766,
      "loss": 4.9717,
      "mean_token_accuracy": 0.18173816800117493,
      "num_tokens": 72888183.0,
      "step": 33265
    },
    {
      "entropy": 5.526460599899292,
      "epoch": 3.3019055180627235,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003934656651822832,
      "loss": 5.1283,
      "mean_token_accuracy": 0.174381884932518,
      "num_tokens": 72899171.0,
      "step": 33270
    },
    {
      "entropy": 5.451638460159302,
      "epoch": 3.302401746724891,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003934355410454754,
      "loss": 5.0183,
      "mean_token_accuracy": 0.17286368310451508,
      "num_tokens": 72909362.0,
      "step": 33275
    },
    {
      "entropy": 5.438425970077515,
      "epoch": 3.3028979753870584,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039340541397172137,
      "loss": 4.9854,
      "mean_token_accuracy": 0.181201334297657,
      "num_tokens": 72920110.0,
      "step": 33280
    },
    {
      "entropy": 5.423817205429077,
      "epoch": 3.303394204049226,
      "grad_norm": 1.0,
      "learning_rate": 0.0003933752839617682,
      "loss": 5.0299,
      "mean_token_accuracy": 0.17466076612472534,
      "num_tokens": 72931006.0,
      "step": 33285
    },
    {
      "entropy": 5.454513454437256,
      "epoch": 3.3038904327113934,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003933451510163632,
      "loss": 5.1349,
      "mean_token_accuracy": 0.1680639386177063,
      "num_tokens": 72941371.0,
      "step": 33290
    },
    {
      "entropy": 5.396855592727661,
      "epoch": 3.304386661373561,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003933150151362534,
      "loss": 4.9776,
      "mean_token_accuracy": 0.18428311198949815,
      "num_tokens": 72953044.0,
      "step": 33295
    },
    {
      "entropy": 5.403037261962891,
      "epoch": 3.3048828900357283,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039328487632218613,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18429085612297058,
      "num_tokens": 72963581.0,
      "step": 33300
    },
    {
      "entropy": 5.349432563781738,
      "epoch": 3.305379118697896,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003932547345749088,
      "loss": 4.9487,
      "mean_token_accuracy": 0.187881538271904,
      "num_tokens": 72974153.0,
      "step": 33305
    },
    {
      "entropy": 5.427065753936768,
      "epoch": 3.3058753473600637,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039322458989516897,
      "loss": 5.0145,
      "mean_token_accuracy": 0.1790195509791374,
      "num_tokens": 72985071.0,
      "step": 33310
    },
    {
      "entropy": 5.410664796829224,
      "epoch": 3.306371576022231,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039319444228371397,
      "loss": 5.0028,
      "mean_token_accuracy": 0.18638053238391877,
      "num_tokens": 72995592.0,
      "step": 33315
    },
    {
      "entropy": 5.375752305984497,
      "epoch": 3.3068678046843987,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003931642917412916,
      "loss": 4.9386,
      "mean_token_accuracy": 0.18614509850740432,
      "num_tokens": 73005431.0,
      "step": 33320
    },
    {
      "entropy": 5.312470722198486,
      "epoch": 3.307364033346566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003931341382686493,
      "loss": 4.8817,
      "mean_token_accuracy": 0.19595985263586044,
      "num_tokens": 73015212.0,
      "step": 33325
    },
    {
      "entropy": 5.36554217338562,
      "epoch": 3.3078602620087336,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039310398186653497,
      "loss": 4.9607,
      "mean_token_accuracy": 0.1743878036737442,
      "num_tokens": 73026585.0,
      "step": 33330
    },
    {
      "entropy": 5.405345153808594,
      "epoch": 3.308356490670901,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039307382253569637,
      "loss": 5.0502,
      "mean_token_accuracy": 0.17939137071371078,
      "num_tokens": 73036844.0,
      "step": 33335
    },
    {
      "entropy": 5.42270302772522,
      "epoch": 3.3088527193330686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003930436602768814,
      "loss": 4.955,
      "mean_token_accuracy": 0.18507084995508194,
      "num_tokens": 73047881.0,
      "step": 33340
    },
    {
      "entropy": 5.393201589584351,
      "epoch": 3.309348947995236,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000393013495090838,
      "loss": 4.9825,
      "mean_token_accuracy": 0.1848638743162155,
      "num_tokens": 73058541.0,
      "step": 33345
    },
    {
      "entropy": 5.3081012725830075,
      "epoch": 3.3098451766574035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003929833269783143,
      "loss": 4.8658,
      "mean_token_accuracy": 0.19571426659822463,
      "num_tokens": 73069208.0,
      "step": 33350
    },
    {
      "entropy": 5.330741214752197,
      "epoch": 3.3103414053195714,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003929531559400582,
      "loss": 4.9491,
      "mean_token_accuracy": 0.1828716591000557,
      "num_tokens": 73079581.0,
      "step": 33355
    },
    {
      "entropy": 5.37287745475769,
      "epoch": 3.310837633981739,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039292298197681803,
      "loss": 4.9855,
      "mean_token_accuracy": 0.18520210832357406,
      "num_tokens": 73090126.0,
      "step": 33360
    },
    {
      "entropy": 5.379519271850586,
      "epoch": 3.3113338626439064,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039289280508934203,
      "loss": 4.9624,
      "mean_token_accuracy": 0.17986005544662476,
      "num_tokens": 73101421.0,
      "step": 33365
    },
    {
      "entropy": 5.42810754776001,
      "epoch": 3.311830091306074,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039286262527837856,
      "loss": 4.9663,
      "mean_token_accuracy": 0.18442622125148772,
      "num_tokens": 73111826.0,
      "step": 33370
    },
    {
      "entropy": 5.447700881958008,
      "epoch": 3.3123263199682413,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003928324425446759,
      "loss": 4.9899,
      "mean_token_accuracy": 0.18941664695739746,
      "num_tokens": 73124195.0,
      "step": 33375
    },
    {
      "entropy": 5.3517457962036135,
      "epoch": 3.312822548630409,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039280225688898256,
      "loss": 4.9753,
      "mean_token_accuracy": 0.1802009329199791,
      "num_tokens": 73135590.0,
      "step": 33380
    },
    {
      "entropy": 5.3775876522064205,
      "epoch": 3.3133187772925763,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039277206831204715,
      "loss": 4.918,
      "mean_token_accuracy": 0.18268030136823654,
      "num_tokens": 73145577.0,
      "step": 33385
    },
    {
      "entropy": 5.301714086532593,
      "epoch": 3.3138150059547438,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003927418768146182,
      "loss": 4.9479,
      "mean_token_accuracy": 0.19294449537992478,
      "num_tokens": 73156717.0,
      "step": 33390
    },
    {
      "entropy": 5.403112554550171,
      "epoch": 3.3143112346169117,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039271168239744434,
      "loss": 4.9609,
      "mean_token_accuracy": 0.1796252354979515,
      "num_tokens": 73166534.0,
      "step": 33395
    },
    {
      "entropy": 5.413969326019287,
      "epoch": 3.314807463279079,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003926814850612745,
      "loss": 4.9333,
      "mean_token_accuracy": 0.18709259927272798,
      "num_tokens": 73176182.0,
      "step": 33400
    },
    {
      "entropy": 5.4144996166229244,
      "epoch": 3.3153036919412466,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003926512848068573,
      "loss": 5.0343,
      "mean_token_accuracy": 0.17481852620840072,
      "num_tokens": 73186837.0,
      "step": 33405
    },
    {
      "entropy": 5.378273820877075,
      "epoch": 3.315799920603414,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003926210816349418,
      "loss": 4.9608,
      "mean_token_accuracy": 0.18584224283695222,
      "num_tokens": 73197786.0,
      "step": 33410
    },
    {
      "entropy": 5.337881755828858,
      "epoch": 3.3162961492655816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000392590875546277,
      "loss": 4.9082,
      "mean_token_accuracy": 0.186443467438221,
      "num_tokens": 73208408.0,
      "step": 33415
    },
    {
      "entropy": 5.383711910247802,
      "epoch": 3.316792377927749,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039256066654161175,
      "loss": 5.0339,
      "mean_token_accuracy": 0.18462282419204712,
      "num_tokens": 73219596.0,
      "step": 33420
    },
    {
      "entropy": 5.418648529052734,
      "epoch": 3.3172886065899165,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003925304546216954,
      "loss": 5.0509,
      "mean_token_accuracy": 0.1767071083188057,
      "num_tokens": 73232301.0,
      "step": 33425
    },
    {
      "entropy": 5.495206356048584,
      "epoch": 3.317784835252084,
      "grad_norm": 1.078125,
      "learning_rate": 0.000392500239787277,
      "loss": 5.049,
      "mean_token_accuracy": 0.1741191029548645,
      "num_tokens": 73243447.0,
      "step": 33430
    },
    {
      "entropy": 5.449351501464844,
      "epoch": 3.3182810639142515,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003924700220391058,
      "loss": 5.0624,
      "mean_token_accuracy": 0.18096065670251846,
      "num_tokens": 73254336.0,
      "step": 33435
    },
    {
      "entropy": 5.455349254608154,
      "epoch": 3.3187772925764194,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003924398013779312,
      "loss": 5.0396,
      "mean_token_accuracy": 0.18347631096839906,
      "num_tokens": 73266098.0,
      "step": 33440
    },
    {
      "entropy": 5.390202379226684,
      "epoch": 3.319273521238587,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003924095778045026,
      "loss": 5.0129,
      "mean_token_accuracy": 0.1864832878112793,
      "num_tokens": 73276026.0,
      "step": 33445
    },
    {
      "entropy": 5.378487539291382,
      "epoch": 3.3197697499007544,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003923793513195695,
      "loss": 4.9741,
      "mean_token_accuracy": 0.186916521191597,
      "num_tokens": 73286151.0,
      "step": 33450
    },
    {
      "entropy": 5.471568202972412,
      "epoch": 3.320265978562922,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039234912192388137,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1745286762714386,
      "num_tokens": 73298086.0,
      "step": 33455
    },
    {
      "entropy": 5.49872350692749,
      "epoch": 3.3207622072250893,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003923188896181879,
      "loss": 5.1541,
      "mean_token_accuracy": 0.17017113119363786,
      "num_tokens": 73308732.0,
      "step": 33460
    },
    {
      "entropy": 5.416680812835693,
      "epoch": 3.3212584358872568,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003922886544032388,
      "loss": 5.0043,
      "mean_token_accuracy": 0.18372486978769303,
      "num_tokens": 73318979.0,
      "step": 33465
    },
    {
      "entropy": 5.477697849273682,
      "epoch": 3.3217546645494243,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039225841627978374,
      "loss": 5.0945,
      "mean_token_accuracy": 0.17771629691123964,
      "num_tokens": 73329724.0,
      "step": 33470
    },
    {
      "entropy": 5.364883041381836,
      "epoch": 3.3222508932115917,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003922281752485727,
      "loss": 4.9967,
      "mean_token_accuracy": 0.17633552551269532,
      "num_tokens": 73340463.0,
      "step": 33475
    },
    {
      "entropy": 5.323131084442139,
      "epoch": 3.3227471218737596,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039219793131035544,
      "loss": 4.882,
      "mean_token_accuracy": 0.19695789515972137,
      "num_tokens": 73351089.0,
      "step": 33480
    },
    {
      "entropy": 5.468968677520752,
      "epoch": 3.323243350535927,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003921676844658821,
      "loss": 5.0838,
      "mean_token_accuracy": 0.17441871911287307,
      "num_tokens": 73361747.0,
      "step": 33485
    },
    {
      "entropy": 5.4479881763458256,
      "epoch": 3.3237395791980946,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003921374347159026,
      "loss": 5.0193,
      "mean_token_accuracy": 0.17914775460958482,
      "num_tokens": 73372340.0,
      "step": 33490
    },
    {
      "entropy": 5.407903480529785,
      "epoch": 3.324235807860262,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003921071820611671,
      "loss": 5.0154,
      "mean_token_accuracy": 0.17660306692123412,
      "num_tokens": 73383972.0,
      "step": 33495
    },
    {
      "entropy": 5.337342548370361,
      "epoch": 3.3247320365224295,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003920769265024259,
      "loss": 4.9795,
      "mean_token_accuracy": 0.17859248518943788,
      "num_tokens": 73394346.0,
      "step": 33500
    },
    {
      "entropy": 5.411111450195312,
      "epoch": 3.325228265184597,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039204666804042913,
      "loss": 5.0307,
      "mean_token_accuracy": 0.1737190842628479,
      "num_tokens": 73404964.0,
      "step": 33505
    },
    {
      "entropy": 5.479685497283936,
      "epoch": 3.3257244938467645,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003920164066759273,
      "loss": 5.0691,
      "mean_token_accuracy": 0.17339686304330826,
      "num_tokens": 73415487.0,
      "step": 33510
    },
    {
      "entropy": 5.454518127441406,
      "epoch": 3.326220722508932,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003919861424096707,
      "loss": 5.0622,
      "mean_token_accuracy": 0.17667144238948823,
      "num_tokens": 73424909.0,
      "step": 33515
    },
    {
      "entropy": 5.489235258102417,
      "epoch": 3.3267169511710994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003919558752424098,
      "loss": 5.048,
      "mean_token_accuracy": 0.17826171964406967,
      "num_tokens": 73435106.0,
      "step": 33520
    },
    {
      "entropy": 5.477728748321534,
      "epoch": 3.327213179833267,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003919256051748952,
      "loss": 5.0681,
      "mean_token_accuracy": 0.17559683471918106,
      "num_tokens": 73445606.0,
      "step": 33525
    },
    {
      "entropy": 5.374240493774414,
      "epoch": 3.327709408495435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039189533220787765,
      "loss": 5.0076,
      "mean_token_accuracy": 0.17927660197019576,
      "num_tokens": 73456309.0,
      "step": 33530
    },
    {
      "entropy": 5.401327896118164,
      "epoch": 3.3282056371576023,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003918650563421076,
      "loss": 4.9827,
      "mean_token_accuracy": 0.1829048737883568,
      "num_tokens": 73467817.0,
      "step": 33535
    },
    {
      "entropy": 5.385431289672852,
      "epoch": 3.32870186581977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000391834777578336,
      "loss": 4.9379,
      "mean_token_accuracy": 0.18724875003099442,
      "num_tokens": 73477312.0,
      "step": 33540
    },
    {
      "entropy": 5.4236647605896,
      "epoch": 3.3291980944819373,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003918044959173138,
      "loss": 5.1093,
      "mean_token_accuracy": 0.1809835210442543,
      "num_tokens": 73488772.0,
      "step": 33545
    },
    {
      "entropy": 5.394353580474854,
      "epoch": 3.3296943231441047,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003917742113597918,
      "loss": 4.9936,
      "mean_token_accuracy": 0.17890530228614807,
      "num_tokens": 73500570.0,
      "step": 33550
    },
    {
      "entropy": 5.364930057525635,
      "epoch": 3.330190551806272,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003917439239065208,
      "loss": 4.9477,
      "mean_token_accuracy": 0.18487173914909363,
      "num_tokens": 73511679.0,
      "step": 33555
    },
    {
      "entropy": 5.385964488983154,
      "epoch": 3.3306867804684397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039171363355825216,
      "loss": 4.9797,
      "mean_token_accuracy": 0.18832760900259018,
      "num_tokens": 73523312.0,
      "step": 33560
    },
    {
      "entropy": 5.4613609790802,
      "epoch": 3.3311830091306076,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039168334031573695,
      "loss": 5.0557,
      "mean_token_accuracy": 0.17828723043203354,
      "num_tokens": 73534789.0,
      "step": 33565
    },
    {
      "entropy": 5.46058988571167,
      "epoch": 3.331679237792775,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003916530441797262,
      "loss": 5.0166,
      "mean_token_accuracy": 0.18022844344377517,
      "num_tokens": 73545331.0,
      "step": 33570
    },
    {
      "entropy": 5.391769123077393,
      "epoch": 3.3321754664549426,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039162274515097147,
      "loss": 4.9542,
      "mean_token_accuracy": 0.18488970398902893,
      "num_tokens": 73555347.0,
      "step": 33575
    },
    {
      "entropy": 5.384586238861084,
      "epoch": 3.33267169511711,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003915924432302239,
      "loss": 5.023,
      "mean_token_accuracy": 0.18137118518352507,
      "num_tokens": 73567180.0,
      "step": 33580
    },
    {
      "entropy": 5.4436279296875,
      "epoch": 3.3331679237792775,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003915621384182349,
      "loss": 5.0763,
      "mean_token_accuracy": 0.1781131938099861,
      "num_tokens": 73578589.0,
      "step": 33585
    },
    {
      "entropy": 5.450430870056152,
      "epoch": 3.333664152441445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003915318307157561,
      "loss": 4.9677,
      "mean_token_accuracy": 0.18136014491319657,
      "num_tokens": 73588939.0,
      "step": 33590
    },
    {
      "entropy": 5.436947536468506,
      "epoch": 3.3341603811036125,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039150152012353896,
      "loss": 5.0526,
      "mean_token_accuracy": 0.17788740247488022,
      "num_tokens": 73600476.0,
      "step": 33595
    },
    {
      "entropy": 5.4802733898162845,
      "epoch": 3.33465660976578,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039147120664233525,
      "loss": 5.0588,
      "mean_token_accuracy": 0.18714150339365004,
      "num_tokens": 73610826.0,
      "step": 33600
    },
    {
      "entropy": 5.436606550216675,
      "epoch": 3.3351528384279474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039144089027289653,
      "loss": 4.9919,
      "mean_token_accuracy": 0.17988209724426268,
      "num_tokens": 73623029.0,
      "step": 33605
    },
    {
      "entropy": 5.3802149295806885,
      "epoch": 3.335649067090115,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039141057101597465,
      "loss": 4.9871,
      "mean_token_accuracy": 0.18306195437908174,
      "num_tokens": 73634092.0,
      "step": 33610
    },
    {
      "entropy": 5.416773271560669,
      "epoch": 3.336145295752283,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003913802488723215,
      "loss": 4.9729,
      "mean_token_accuracy": 0.19186298102140426,
      "num_tokens": 73643297.0,
      "step": 33615
    },
    {
      "entropy": 5.366734361648559,
      "epoch": 3.3366415244144503,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003913499238426889,
      "loss": 4.9975,
      "mean_token_accuracy": 0.1851107284426689,
      "num_tokens": 73654001.0,
      "step": 33620
    },
    {
      "entropy": 5.393735647201538,
      "epoch": 3.3371377530766178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039131959592782884,
      "loss": 4.9941,
      "mean_token_accuracy": 0.18012942969799042,
      "num_tokens": 73664509.0,
      "step": 33625
    },
    {
      "entropy": 5.348919820785523,
      "epoch": 3.3376339817387852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039128926512849356,
      "loss": 4.8879,
      "mean_token_accuracy": 0.18820395469665527,
      "num_tokens": 73674585.0,
      "step": 33630
    },
    {
      "entropy": 5.3481005191802975,
      "epoch": 3.3381302104009527,
      "grad_norm": 0.96875,
      "learning_rate": 0.000391258931445435,
      "loss": 4.86,
      "mean_token_accuracy": 0.18259898126125335,
      "num_tokens": 73684550.0,
      "step": 33635
    },
    {
      "entropy": 5.42025055885315,
      "epoch": 3.33862643906312,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003912285948794055,
      "loss": 5.037,
      "mean_token_accuracy": 0.18019335865974426,
      "num_tokens": 73694767.0,
      "step": 33640
    },
    {
      "entropy": 5.5762487888336185,
      "epoch": 3.3391226677252877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003911982554311573,
      "loss": 5.0932,
      "mean_token_accuracy": 0.180497470498085,
      "num_tokens": 73707900.0,
      "step": 33645
    },
    {
      "entropy": 5.485906410217285,
      "epoch": 3.3396188963874556,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039116791310144275,
      "loss": 5.0535,
      "mean_token_accuracy": 0.18134179562330247,
      "num_tokens": 73719014.0,
      "step": 33650
    },
    {
      "entropy": 5.466574478149414,
      "epoch": 3.340115125049623,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003911375678910142,
      "loss": 5.1318,
      "mean_token_accuracy": 0.17382141053676606,
      "num_tokens": 73730107.0,
      "step": 33655
    },
    {
      "entropy": 5.489460134506226,
      "epoch": 3.3406113537117905,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039110721980062434,
      "loss": 5.1338,
      "mean_token_accuracy": 0.17698842585086821,
      "num_tokens": 73742355.0,
      "step": 33660
    },
    {
      "entropy": 5.490702152252197,
      "epoch": 3.341107582373958,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039107686883102554,
      "loss": 5.0877,
      "mean_token_accuracy": 0.17717001140117644,
      "num_tokens": 73753372.0,
      "step": 33665
    },
    {
      "entropy": 5.411437892913819,
      "epoch": 3.3416038110361255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039104651498297053,
      "loss": 4.9993,
      "mean_token_accuracy": 0.1823589488863945,
      "num_tokens": 73764547.0,
      "step": 33670
    },
    {
      "entropy": 5.449496412277222,
      "epoch": 3.342100039698293,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000391016158257212,
      "loss": 5.0674,
      "mean_token_accuracy": 0.1810700014233589,
      "num_tokens": 73775505.0,
      "step": 33675
    },
    {
      "entropy": 5.4996339797973635,
      "epoch": 3.3425962683604604,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003909857986545028,
      "loss": 5.0966,
      "mean_token_accuracy": 0.17510397732257843,
      "num_tokens": 73785963.0,
      "step": 33680
    },
    {
      "entropy": 5.484544324874878,
      "epoch": 3.343092497022628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039095543617559566,
      "loss": 4.9571,
      "mean_token_accuracy": 0.18340194523334502,
      "num_tokens": 73796457.0,
      "step": 33685
    },
    {
      "entropy": 5.51483645439148,
      "epoch": 3.3435887256847954,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003909250708212436,
      "loss": 5.0491,
      "mean_token_accuracy": 0.17820413112640382,
      "num_tokens": 73807662.0,
      "step": 33690
    },
    {
      "entropy": 5.414791440963745,
      "epoch": 3.344084954346963,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039089470259219956,
      "loss": 5.0374,
      "mean_token_accuracy": 0.17880839705467225,
      "num_tokens": 73818688.0,
      "step": 33695
    },
    {
      "entropy": 5.44263858795166,
      "epoch": 3.3445811830091308,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039086433148921665,
      "loss": 5.0549,
      "mean_token_accuracy": 0.17674943804740906,
      "num_tokens": 73831234.0,
      "step": 33700
    },
    {
      "entropy": 5.416278934478759,
      "epoch": 3.3450774116712982,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003908339575130479,
      "loss": 5.0269,
      "mean_token_accuracy": 0.18851690143346786,
      "num_tokens": 73841505.0,
      "step": 33705
    },
    {
      "entropy": 5.474075555801392,
      "epoch": 3.3455736403334657,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003908035806644467,
      "loss": 5.052,
      "mean_token_accuracy": 0.1757915735244751,
      "num_tokens": 73851900.0,
      "step": 33710
    },
    {
      "entropy": 5.467041778564453,
      "epoch": 3.346069868995633,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039077320094416617,
      "loss": 4.9857,
      "mean_token_accuracy": 0.18963960260152818,
      "num_tokens": 73862317.0,
      "step": 33715
    },
    {
      "entropy": 5.425394105911255,
      "epoch": 3.3465660976578007,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003907428183529598,
      "loss": 5.0281,
      "mean_token_accuracy": 0.1727563813328743,
      "num_tokens": 73872166.0,
      "step": 33720
    },
    {
      "entropy": 5.421058511734008,
      "epoch": 3.347062326319968,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039071243289158085,
      "loss": 5.0185,
      "mean_token_accuracy": 0.1796514704823494,
      "num_tokens": 73883064.0,
      "step": 33725
    },
    {
      "entropy": 5.428688335418701,
      "epoch": 3.3475585549821356,
      "grad_norm": 1.0,
      "learning_rate": 0.00039068204456078296,
      "loss": 5.0034,
      "mean_token_accuracy": 0.18138481825590133,
      "num_tokens": 73895032.0,
      "step": 33730
    },
    {
      "entropy": 5.4146191596984865,
      "epoch": 3.3480547836443035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039065165336131963,
      "loss": 4.9835,
      "mean_token_accuracy": 0.1874190017580986,
      "num_tokens": 73906309.0,
      "step": 33735
    },
    {
      "entropy": 5.407496452331543,
      "epoch": 3.348551012306471,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003906212592939445,
      "loss": 4.9918,
      "mean_token_accuracy": 0.17784871459007262,
      "num_tokens": 73917240.0,
      "step": 33740
    },
    {
      "entropy": 5.373044061660766,
      "epoch": 3.3490472409686385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039059086235941127,
      "loss": 4.9663,
      "mean_token_accuracy": 0.17885575741529464,
      "num_tokens": 73928216.0,
      "step": 33745
    },
    {
      "entropy": 5.398536539077758,
      "epoch": 3.349543469630806,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003905604625584737,
      "loss": 5.0034,
      "mean_token_accuracy": 0.18305015563964844,
      "num_tokens": 73940115.0,
      "step": 33750
    },
    {
      "entropy": 5.473001623153687,
      "epoch": 3.3500396982929734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003905300598918857,
      "loss": 5.0892,
      "mean_token_accuracy": 0.1777743563055992,
      "num_tokens": 73950756.0,
      "step": 33755
    },
    {
      "entropy": 5.435230159759522,
      "epoch": 3.350535926955141,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039049965436040125,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18574047833681107,
      "num_tokens": 73962269.0,
      "step": 33760
    },
    {
      "entropy": 5.374963998794556,
      "epoch": 3.3510321556173084,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003904692459647741,
      "loss": 4.9586,
      "mean_token_accuracy": 0.18257542401552201,
      "num_tokens": 73971883.0,
      "step": 33765
    },
    {
      "entropy": 5.460019826889038,
      "epoch": 3.351528384279476,
      "grad_norm": 0.90625,
      "learning_rate": 0.00039043883470575856,
      "loss": 5.0916,
      "mean_token_accuracy": 0.18071721941232682,
      "num_tokens": 73983801.0,
      "step": 33770
    },
    {
      "entropy": 5.421697950363159,
      "epoch": 3.3520246129416433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039040842058410863,
      "loss": 4.9982,
      "mean_token_accuracy": 0.18341130018234253,
      "num_tokens": 73994198.0,
      "step": 33775
    },
    {
      "entropy": 5.470569467544555,
      "epoch": 3.352520841603811,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039037800360057864,
      "loss": 5.0557,
      "mean_token_accuracy": 0.17220084071159364,
      "num_tokens": 74005521.0,
      "step": 33780
    },
    {
      "entropy": 5.4682220935821535,
      "epoch": 3.3530170702659787,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003903475837559226,
      "loss": 5.0629,
      "mean_token_accuracy": 0.18724535554647445,
      "num_tokens": 74016606.0,
      "step": 33785
    },
    {
      "entropy": 5.392180633544922,
      "epoch": 3.353513298928146,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039031716105089516,
      "loss": 5.0138,
      "mean_token_accuracy": 0.18750035166740417,
      "num_tokens": 74027111.0,
      "step": 33790
    },
    {
      "entropy": 5.364354372024536,
      "epoch": 3.3540095275903137,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003902867354862506,
      "loss": 4.9303,
      "mean_token_accuracy": 0.18422077894210814,
      "num_tokens": 74037425.0,
      "step": 33795
    },
    {
      "entropy": 5.401456642150879,
      "epoch": 3.354505756252481,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003902563070627434,
      "loss": 5.0245,
      "mean_token_accuracy": 0.1820485234260559,
      "num_tokens": 74048622.0,
      "step": 33800
    },
    {
      "entropy": 5.402909564971924,
      "epoch": 3.3550019849146486,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039022587578112813,
      "loss": 4.9874,
      "mean_token_accuracy": 0.1835455983877182,
      "num_tokens": 74058430.0,
      "step": 33805
    },
    {
      "entropy": 5.446691942214966,
      "epoch": 3.355498213576816,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039019544164215956,
      "loss": 5.0242,
      "mean_token_accuracy": 0.1790039584040642,
      "num_tokens": 74069488.0,
      "step": 33810
    },
    {
      "entropy": 5.426865720748902,
      "epoch": 3.3559944422389836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039016500464659213,
      "loss": 4.9971,
      "mean_token_accuracy": 0.1832887813448906,
      "num_tokens": 74079142.0,
      "step": 33815
    },
    {
      "entropy": 5.457046842575073,
      "epoch": 3.3564906709011515,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003901345647951807,
      "loss": 5.0362,
      "mean_token_accuracy": 0.17562223672866822,
      "num_tokens": 74089984.0,
      "step": 33820
    },
    {
      "entropy": 5.410950422286987,
      "epoch": 3.356986899563319,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039010412208868026,
      "loss": 5.0395,
      "mean_token_accuracy": 0.1769495725631714,
      "num_tokens": 74101225.0,
      "step": 33825
    },
    {
      "entropy": 5.355516958236694,
      "epoch": 3.3574831282254864,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003900736765278455,
      "loss": 4.9197,
      "mean_token_accuracy": 0.19284462332725524,
      "num_tokens": 74111198.0,
      "step": 33830
    },
    {
      "entropy": 5.48547329902649,
      "epoch": 3.357979356887654,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039004322811343165,
      "loss": 5.099,
      "mean_token_accuracy": 0.16950763314962386,
      "num_tokens": 74121458.0,
      "step": 33835
    },
    {
      "entropy": 5.427397060394287,
      "epoch": 3.3584755855498214,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039001277684619357,
      "loss": 5.0422,
      "mean_token_accuracy": 0.18374361246824264,
      "num_tokens": 74133338.0,
      "step": 33840
    },
    {
      "entropy": 5.370839786529541,
      "epoch": 3.358971814211989,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003899823227268865,
      "loss": 5.0564,
      "mean_token_accuracy": 0.177116522192955,
      "num_tokens": 74144062.0,
      "step": 33845
    },
    {
      "entropy": 5.367576313018799,
      "epoch": 3.3594680428741563,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038995186575626554,
      "loss": 4.9938,
      "mean_token_accuracy": 0.18195905089378356,
      "num_tokens": 74154001.0,
      "step": 33850
    },
    {
      "entropy": 5.387108850479126,
      "epoch": 3.359964271536324,
      "grad_norm": 0.953125,
      "learning_rate": 0.000389921405935086,
      "loss": 4.9601,
      "mean_token_accuracy": 0.18431116491556168,
      "num_tokens": 74164254.0,
      "step": 33855
    },
    {
      "entropy": 5.376916933059692,
      "epoch": 3.3604605001984913,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038989094326410327,
      "loss": 4.9376,
      "mean_token_accuracy": 0.19288201928138732,
      "num_tokens": 74173915.0,
      "step": 33860
    },
    {
      "entropy": 5.541948890686035,
      "epoch": 3.3609567288606588,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003898604777440727,
      "loss": 5.139,
      "mean_token_accuracy": 0.16601809859275818,
      "num_tokens": 74185903.0,
      "step": 33865
    },
    {
      "entropy": 5.487831878662109,
      "epoch": 3.3614529575228267,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003898300093757499,
      "loss": 5.0628,
      "mean_token_accuracy": 0.17546252459287642,
      "num_tokens": 74198361.0,
      "step": 33870
    },
    {
      "entropy": 5.35136981010437,
      "epoch": 3.361949186184994,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003897995381598901,
      "loss": 4.9966,
      "mean_token_accuracy": 0.1750118166208267,
      "num_tokens": 74210467.0,
      "step": 33875
    },
    {
      "entropy": 5.354898071289062,
      "epoch": 3.3624454148471616,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003897690640972492,
      "loss": 4.9415,
      "mean_token_accuracy": 0.1885494127869606,
      "num_tokens": 74221717.0,
      "step": 33880
    },
    {
      "entropy": 5.392679452896118,
      "epoch": 3.362941643509329,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038973858718858295,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18122388869524003,
      "num_tokens": 74233164.0,
      "step": 33885
    },
    {
      "entropy": 5.513504981994629,
      "epoch": 3.3634378721714966,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003897081074346468,
      "loss": 5.0876,
      "mean_token_accuracy": 0.1736449733376503,
      "num_tokens": 74245693.0,
      "step": 33890
    },
    {
      "entropy": 5.382313442230225,
      "epoch": 3.363934100833664,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038967762483619684,
      "loss": 4.9415,
      "mean_token_accuracy": 0.19341123402118682,
      "num_tokens": 74256499.0,
      "step": 33895
    },
    {
      "entropy": 5.408911895751953,
      "epoch": 3.3644303294958315,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003896471393939889,
      "loss": 5.0546,
      "mean_token_accuracy": 0.18430915623903274,
      "num_tokens": 74267713.0,
      "step": 33900
    },
    {
      "entropy": 5.424875783920288,
      "epoch": 3.364926558157999,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003896166511087789,
      "loss": 5.0801,
      "mean_token_accuracy": 0.17977195829153061,
      "num_tokens": 74279149.0,
      "step": 33905
    },
    {
      "entropy": 5.418867588043213,
      "epoch": 3.365422786820167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000389586159981323,
      "loss": 4.9938,
      "mean_token_accuracy": 0.1893578588962555,
      "num_tokens": 74289520.0,
      "step": 33910
    },
    {
      "entropy": 5.47594838142395,
      "epoch": 3.3659190154823344,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003895556660123771,
      "loss": 5.0105,
      "mean_token_accuracy": 0.18501646369695662,
      "num_tokens": 74300546.0,
      "step": 33915
    },
    {
      "entropy": 5.354170942306519,
      "epoch": 3.366415244144502,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038952516920269764,
      "loss": 4.9432,
      "mean_token_accuracy": 0.17933772057294844,
      "num_tokens": 74310687.0,
      "step": 33920
    },
    {
      "entropy": 5.413970613479615,
      "epoch": 3.3669114728066694,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003894946695530408,
      "loss": 4.9951,
      "mean_token_accuracy": 0.17689030915498732,
      "num_tokens": 74322251.0,
      "step": 33925
    },
    {
      "entropy": 5.486088275909424,
      "epoch": 3.367407701468837,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038946416706416275,
      "loss": 5.033,
      "mean_token_accuracy": 0.17160936444997787,
      "num_tokens": 74333845.0,
      "step": 33930
    },
    {
      "entropy": 5.469594860076905,
      "epoch": 3.3679039301310043,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003894336617368201,
      "loss": 5.0543,
      "mean_token_accuracy": 0.17727560847997664,
      "num_tokens": 74344188.0,
      "step": 33935
    },
    {
      "entropy": 5.415221118927002,
      "epoch": 3.368400158793172,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00038940315357176926,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1721617490053177,
      "num_tokens": 74355174.0,
      "step": 33940
    },
    {
      "entropy": 5.417795324325562,
      "epoch": 3.3688963874553393,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038937264256976657,
      "loss": 4.9483,
      "mean_token_accuracy": 0.18834576904773712,
      "num_tokens": 74365396.0,
      "step": 33945
    },
    {
      "entropy": 5.4023291110992435,
      "epoch": 3.3693926161175067,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003893421287315689,
      "loss": 4.9926,
      "mean_token_accuracy": 0.17827546298503877,
      "num_tokens": 74377303.0,
      "step": 33950
    },
    {
      "entropy": 5.356221532821655,
      "epoch": 3.3698888447796747,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003893116120579328,
      "loss": 5.0202,
      "mean_token_accuracy": 0.17785131931304932,
      "num_tokens": 74389967.0,
      "step": 33955
    },
    {
      "entropy": 5.440314245223999,
      "epoch": 3.370385073441842,
      "grad_norm": 1.0,
      "learning_rate": 0.0003892810925496151,
      "loss": 4.9928,
      "mean_token_accuracy": 0.17897299975156783,
      "num_tokens": 74400252.0,
      "step": 33960
    },
    {
      "entropy": 5.386027669906616,
      "epoch": 3.3708813021040096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003892505702073725,
      "loss": 4.9848,
      "mean_token_accuracy": 0.18382516801357268,
      "num_tokens": 74412165.0,
      "step": 33965
    },
    {
      "entropy": 5.3498780727386475,
      "epoch": 3.371377530766177,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038922004503196195,
      "loss": 4.9407,
      "mean_token_accuracy": 0.18842052072286605,
      "num_tokens": 74423078.0,
      "step": 33970
    },
    {
      "entropy": 5.3441976547241214,
      "epoch": 3.3718737594283446,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003891895170241405,
      "loss": 4.9258,
      "mean_token_accuracy": 0.18544422090053558,
      "num_tokens": 74434552.0,
      "step": 33975
    },
    {
      "entropy": 5.420249986648559,
      "epoch": 3.372369988090512,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038915898618466487,
      "loss": 5.0396,
      "mean_token_accuracy": 0.17968420833349227,
      "num_tokens": 74446188.0,
      "step": 33980
    },
    {
      "entropy": 5.328199338912964,
      "epoch": 3.3728662167526795,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038912845251429256,
      "loss": 4.9209,
      "mean_token_accuracy": 0.18754867911338807,
      "num_tokens": 74456358.0,
      "step": 33985
    },
    {
      "entropy": 5.303410625457763,
      "epoch": 3.373362445414847,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038909791601378054,
      "loss": 4.8824,
      "mean_token_accuracy": 0.1848541721701622,
      "num_tokens": 74466155.0,
      "step": 33990
    },
    {
      "entropy": 5.467749452590942,
      "epoch": 3.373858674077015,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038906737668388595,
      "loss": 5.1016,
      "mean_token_accuracy": 0.1711037576198578,
      "num_tokens": 74477282.0,
      "step": 33995
    },
    {
      "entropy": 5.3463598728179935,
      "epoch": 3.3743549027391824,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038903683452536634,
      "loss": 4.9644,
      "mean_token_accuracy": 0.18563371598720552,
      "num_tokens": 74489736.0,
      "step": 34000
    },
    {
      "entropy": 5.308917903900147,
      "epoch": 3.37485113140135,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003890062895389789,
      "loss": 4.9734,
      "mean_token_accuracy": 0.17814530432224274,
      "num_tokens": 74500797.0,
      "step": 34005
    },
    {
      "entropy": 5.366389989852905,
      "epoch": 3.3753473600635173,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003889757417254811,
      "loss": 5.0021,
      "mean_token_accuracy": 0.18296977132558823,
      "num_tokens": 74510811.0,
      "step": 34010
    },
    {
      "entropy": 5.4467125415802,
      "epoch": 3.375843588725685,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038894519108563047,
      "loss": 4.9737,
      "mean_token_accuracy": 0.18679150938987732,
      "num_tokens": 74521726.0,
      "step": 34015
    },
    {
      "entropy": 5.394652843475342,
      "epoch": 3.3763398173878523,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038891463762018476,
      "loss": 5.0021,
      "mean_token_accuracy": 0.18762665390968322,
      "num_tokens": 74532260.0,
      "step": 34020
    },
    {
      "entropy": 5.450306940078735,
      "epoch": 3.3768360460500197,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038888408132990146,
      "loss": 5.0587,
      "mean_token_accuracy": 0.1717986658215523,
      "num_tokens": 74544691.0,
      "step": 34025
    },
    {
      "entropy": 5.452050971984863,
      "epoch": 3.377332274712187,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038885352221553827,
      "loss": 5.0336,
      "mean_token_accuracy": 0.17717496752738954,
      "num_tokens": 74555250.0,
      "step": 34030
    },
    {
      "entropy": 5.38316068649292,
      "epoch": 3.3778285033743547,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003888229602778532,
      "loss": 4.9527,
      "mean_token_accuracy": 0.18228165656328202,
      "num_tokens": 74566079.0,
      "step": 34035
    },
    {
      "entropy": 5.411316871643066,
      "epoch": 3.3783247320365226,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00038879239551760386,
      "loss": 5.0403,
      "mean_token_accuracy": 0.17923801988363267,
      "num_tokens": 74577741.0,
      "step": 34040
    },
    {
      "entropy": 5.386161661148071,
      "epoch": 3.37882096069869,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003887618279355483,
      "loss": 4.9517,
      "mean_token_accuracy": 0.18583111017942427,
      "num_tokens": 74588146.0,
      "step": 34045
    },
    {
      "entropy": 5.455913543701172,
      "epoch": 3.3793171893608576,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038873125753244466,
      "loss": 5.1053,
      "mean_token_accuracy": 0.1783323124051094,
      "num_tokens": 74599745.0,
      "step": 34050
    },
    {
      "entropy": 5.410540771484375,
      "epoch": 3.379813418023025,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003887006843090508,
      "loss": 4.9516,
      "mean_token_accuracy": 0.17792655378580094,
      "num_tokens": 74611215.0,
      "step": 34055
    },
    {
      "entropy": 5.485703468322754,
      "epoch": 3.3803096466851925,
      "grad_norm": 1.03125,
      "learning_rate": 0.000388670108266125,
      "loss": 5.056,
      "mean_token_accuracy": 0.18737722784280778,
      "num_tokens": 74620676.0,
      "step": 34060
    },
    {
      "entropy": 5.472678709030151,
      "epoch": 3.38080587534736,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003886395294044255,
      "loss": 5.1056,
      "mean_token_accuracy": 0.1753143846988678,
      "num_tokens": 74633129.0,
      "step": 34065
    },
    {
      "entropy": 5.43122501373291,
      "epoch": 3.3813021040095275,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003886089477247105,
      "loss": 4.9575,
      "mean_token_accuracy": 0.17968078702688217,
      "num_tokens": 74643125.0,
      "step": 34070
    },
    {
      "entropy": 5.5071516036987305,
      "epoch": 3.381798332671695,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003885783632277384,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17513405829668044,
      "num_tokens": 74653573.0,
      "step": 34075
    },
    {
      "entropy": 5.4172135353088375,
      "epoch": 3.382294561333863,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003885477759142676,
      "loss": 4.9289,
      "mean_token_accuracy": 0.19036852866411208,
      "num_tokens": 74663451.0,
      "step": 34080
    },
    {
      "entropy": 5.359428644180298,
      "epoch": 3.3827907899960303,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003885171857850567,
      "loss": 4.9489,
      "mean_token_accuracy": 0.1859660640358925,
      "num_tokens": 74673698.0,
      "step": 34085
    },
    {
      "entropy": 5.367064571380615,
      "epoch": 3.383287018658198,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003884865928408642,
      "loss": 5.0564,
      "mean_token_accuracy": 0.1770259201526642,
      "num_tokens": 74685952.0,
      "step": 34090
    },
    {
      "entropy": 5.399816465377808,
      "epoch": 3.3837832473203653,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003884559970824486,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18341962546110152,
      "num_tokens": 74698159.0,
      "step": 34095
    },
    {
      "entropy": 5.468533945083618,
      "epoch": 3.3842794759825328,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003884253985105689,
      "loss": 5.0615,
      "mean_token_accuracy": 0.17777419686317444,
      "num_tokens": 74708718.0,
      "step": 34100
    },
    {
      "entropy": 5.416344547271729,
      "epoch": 3.3847757046447002,
      "grad_norm": 1.0,
      "learning_rate": 0.0003883947971259837,
      "loss": 4.9429,
      "mean_token_accuracy": 0.18126244843006134,
      "num_tokens": 74718249.0,
      "step": 34105
    },
    {
      "entropy": 5.410548877716065,
      "epoch": 3.3852719333068677,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003883641929294519,
      "loss": 4.9204,
      "mean_token_accuracy": 0.19860609471797944,
      "num_tokens": 74728856.0,
      "step": 34110
    },
    {
      "entropy": 5.379506063461304,
      "epoch": 3.385768161969035,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003883335859217323,
      "loss": 4.9785,
      "mean_token_accuracy": 0.18622382432222367,
      "num_tokens": 74740126.0,
      "step": 34115
    },
    {
      "entropy": 5.351134872436523,
      "epoch": 3.3862643906312027,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000388302976103584,
      "loss": 4.9949,
      "mean_token_accuracy": 0.17692252546548842,
      "num_tokens": 74751277.0,
      "step": 34120
    },
    {
      "entropy": 5.397002458572388,
      "epoch": 3.3867606192933706,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003882723634757661,
      "loss": 5.0369,
      "mean_token_accuracy": 0.17875767052173613,
      "num_tokens": 74761498.0,
      "step": 34125
    },
    {
      "entropy": 5.389218854904175,
      "epoch": 3.387256847955538,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003882417480390376,
      "loss": 4.9658,
      "mean_token_accuracy": 0.1776727706193924,
      "num_tokens": 74771482.0,
      "step": 34130
    },
    {
      "entropy": 5.450253009796143,
      "epoch": 3.3877530766177055,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038821112979415786,
      "loss": 5.0503,
      "mean_token_accuracy": 0.17805011719465255,
      "num_tokens": 74783649.0,
      "step": 34135
    },
    {
      "entropy": 5.4138387680053714,
      "epoch": 3.388249305279873,
      "grad_norm": 1.125,
      "learning_rate": 0.000388180508741886,
      "loss": 4.9764,
      "mean_token_accuracy": 0.18418164700269699,
      "num_tokens": 74794142.0,
      "step": 34140
    },
    {
      "entropy": 5.385027551651001,
      "epoch": 3.3887455339420405,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038814988488298136,
      "loss": 4.9941,
      "mean_token_accuracy": 0.179804790019989,
      "num_tokens": 74805369.0,
      "step": 34145
    },
    {
      "entropy": 5.367738676071167,
      "epoch": 3.389241762604208,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003881192582182034,
      "loss": 4.9888,
      "mean_token_accuracy": 0.18335778564214705,
      "num_tokens": 74816558.0,
      "step": 34150
    },
    {
      "entropy": 5.4258757591247555,
      "epoch": 3.3897379912663754,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038808862874831164,
      "loss": 4.9833,
      "mean_token_accuracy": 0.1835537627339363,
      "num_tokens": 74826838.0,
      "step": 34155
    },
    {
      "entropy": 5.326924419403076,
      "epoch": 3.390234219928543,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003880579964740655,
      "loss": 4.9409,
      "mean_token_accuracy": 0.1789726734161377,
      "num_tokens": 74837182.0,
      "step": 34160
    },
    {
      "entropy": 5.391260814666748,
      "epoch": 3.390730448590711,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038802736139622473,
      "loss": 4.9332,
      "mean_token_accuracy": 0.18947242796421052,
      "num_tokens": 74847973.0,
      "step": 34165
    },
    {
      "entropy": 5.428077983856201,
      "epoch": 3.3912266772528783,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038799672351554895,
      "loss": 5.0068,
      "mean_token_accuracy": 0.18210818022489547,
      "num_tokens": 74858952.0,
      "step": 34170
    },
    {
      "entropy": 5.483993768692017,
      "epoch": 3.3917229059150458,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038796608283279786,
      "loss": 5.0207,
      "mean_token_accuracy": 0.180761881172657,
      "num_tokens": 74870427.0,
      "step": 34175
    },
    {
      "entropy": 5.359162712097168,
      "epoch": 3.3922191345772132,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003879354393487314,
      "loss": 4.9687,
      "mean_token_accuracy": 0.18525552600622178,
      "num_tokens": 74881502.0,
      "step": 34180
    },
    {
      "entropy": 5.41765398979187,
      "epoch": 3.3927153632393807,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038790479306410934,
      "loss": 5.0491,
      "mean_token_accuracy": 0.17483481168746948,
      "num_tokens": 74893065.0,
      "step": 34185
    },
    {
      "entropy": 5.364520645141601,
      "epoch": 3.393211591901548,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003878741439796918,
      "loss": 4.985,
      "mean_token_accuracy": 0.18359198719263076,
      "num_tokens": 74903806.0,
      "step": 34190
    },
    {
      "entropy": 5.522730302810669,
      "epoch": 3.3937078205637157,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003878434920962386,
      "loss": 5.1065,
      "mean_token_accuracy": 0.1777633622288704,
      "num_tokens": 74915025.0,
      "step": 34195
    },
    {
      "entropy": 5.429110956192017,
      "epoch": 3.394204049225883,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00038781283741451,
      "loss": 5.0311,
      "mean_token_accuracy": 0.174177385866642,
      "num_tokens": 74926438.0,
      "step": 34200
    },
    {
      "entropy": 5.447095537185669,
      "epoch": 3.3947002778880506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038778217993526616,
      "loss": 4.99,
      "mean_token_accuracy": 0.1786207988858223,
      "num_tokens": 74937707.0,
      "step": 34205
    },
    {
      "entropy": 5.4224225044250485,
      "epoch": 3.395196506550218,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003877515196592672,
      "loss": 4.9777,
      "mean_token_accuracy": 0.17927611023187637,
      "num_tokens": 74947436.0,
      "step": 34210
    },
    {
      "entropy": 5.434295606613159,
      "epoch": 3.395692735212386,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038772085658727353,
      "loss": 5.0531,
      "mean_token_accuracy": 0.1765843003988266,
      "num_tokens": 74958707.0,
      "step": 34215
    },
    {
      "entropy": 5.4135181427001955,
      "epoch": 3.3961889638745535,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038769019072004555,
      "loss": 5.0178,
      "mean_token_accuracy": 0.1836770296096802,
      "num_tokens": 74970145.0,
      "step": 34220
    },
    {
      "entropy": 5.362081480026245,
      "epoch": 3.396685192536721,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003876595220583436,
      "loss": 5.0064,
      "mean_token_accuracy": 0.17862914651632308,
      "num_tokens": 74982095.0,
      "step": 34225
    },
    {
      "entropy": 5.361257171630859,
      "epoch": 3.3971814211988884,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003876288506029283,
      "loss": 4.9048,
      "mean_token_accuracy": 0.1874033972620964,
      "num_tokens": 74993648.0,
      "step": 34230
    },
    {
      "entropy": 5.413701868057251,
      "epoch": 3.397677649861056,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003875981763545602,
      "loss": 4.9653,
      "mean_token_accuracy": 0.1831018179655075,
      "num_tokens": 75003707.0,
      "step": 34235
    },
    {
      "entropy": 5.343882942199707,
      "epoch": 3.3981738785232234,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038756749931399993,
      "loss": 4.9587,
      "mean_token_accuracy": 0.1798220917582512,
      "num_tokens": 75013029.0,
      "step": 34240
    },
    {
      "entropy": 5.377629327774048,
      "epoch": 3.398670107185391,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003875368194820083,
      "loss": 5.0026,
      "mean_token_accuracy": 0.1811673402786255,
      "num_tokens": 75024555.0,
      "step": 34245
    },
    {
      "entropy": 5.476950693130493,
      "epoch": 3.399166335847559,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000387506136859346,
      "loss": 5.0399,
      "mean_token_accuracy": 0.1779547870159149,
      "num_tokens": 75034529.0,
      "step": 34250
    },
    {
      "entropy": 5.446598672866822,
      "epoch": 3.3996625645097263,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003874754514467739,
      "loss": 5.083,
      "mean_token_accuracy": 0.1769633710384369,
      "num_tokens": 75047733.0,
      "step": 34255
    },
    {
      "entropy": 5.452311706542969,
      "epoch": 3.4001587931718937,
      "grad_norm": 0.984375,
      "learning_rate": 0.000387444763245053,
      "loss": 5.0177,
      "mean_token_accuracy": 0.18342276215553283,
      "num_tokens": 75058576.0,
      "step": 34260
    },
    {
      "entropy": 5.404520845413208,
      "epoch": 3.400655021834061,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003874140722549443,
      "loss": 5.0331,
      "mean_token_accuracy": 0.1845068246126175,
      "num_tokens": 75068470.0,
      "step": 34265
    },
    {
      "entropy": 5.333158206939697,
      "epoch": 3.4011512504962287,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003873833784772088,
      "loss": 4.9152,
      "mean_token_accuracy": 0.18677926510572435,
      "num_tokens": 75079383.0,
      "step": 34270
    },
    {
      "entropy": 5.418491268157959,
      "epoch": 3.401647479158396,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003873526819126077,
      "loss": 5.0557,
      "mean_token_accuracy": 0.17214712202548982,
      "num_tokens": 75089474.0,
      "step": 34275
    },
    {
      "entropy": 5.409856081008911,
      "epoch": 3.4021437078205636,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003873219825619022,
      "loss": 5.0361,
      "mean_token_accuracy": 0.18441493213176727,
      "num_tokens": 75100286.0,
      "step": 34280
    },
    {
      "entropy": 5.478721570968628,
      "epoch": 3.402639936482731,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003872912804258536,
      "loss": 5.0686,
      "mean_token_accuracy": 0.1788042426109314,
      "num_tokens": 75111916.0,
      "step": 34285
    },
    {
      "entropy": 5.436806011199951,
      "epoch": 3.4031361651448986,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038726057550522307,
      "loss": 4.9465,
      "mean_token_accuracy": 0.18906715959310533,
      "num_tokens": 75122231.0,
      "step": 34290
    },
    {
      "entropy": 5.48765606880188,
      "epoch": 3.403632393807066,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038722986780077227,
      "loss": 5.088,
      "mean_token_accuracy": 0.17017096877098084,
      "num_tokens": 75133289.0,
      "step": 34295
    },
    {
      "entropy": 5.371739721298217,
      "epoch": 3.404128622469234,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038719915731326264,
      "loss": 4.9751,
      "mean_token_accuracy": 0.1861788734793663,
      "num_tokens": 75143409.0,
      "step": 34300
    },
    {
      "entropy": 5.407282829284668,
      "epoch": 3.4046248511314015,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038716844404345556,
      "loss": 5.0662,
      "mean_token_accuracy": 0.18068346679210662,
      "num_tokens": 75154848.0,
      "step": 34305
    },
    {
      "entropy": 5.388323402404785,
      "epoch": 3.405121079793569,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038713772799211284,
      "loss": 5.0645,
      "mean_token_accuracy": 0.17791986614465713,
      "num_tokens": 75166861.0,
      "step": 34310
    },
    {
      "entropy": 5.427469873428345,
      "epoch": 3.4056173084557364,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003871070091599961,
      "loss": 4.9751,
      "mean_token_accuracy": 0.18715752959251403,
      "num_tokens": 75176934.0,
      "step": 34315
    },
    {
      "entropy": 5.3345129013061525,
      "epoch": 3.406113537117904,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038707628754786706,
      "loss": 4.9274,
      "mean_token_accuracy": 0.1824156805872917,
      "num_tokens": 75187604.0,
      "step": 34320
    },
    {
      "entropy": 5.349548768997193,
      "epoch": 3.4066097657800714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038704556315648757,
      "loss": 4.9102,
      "mean_token_accuracy": 0.18599169850349426,
      "num_tokens": 75198524.0,
      "step": 34325
    },
    {
      "entropy": 5.418764591217041,
      "epoch": 3.407105994442239,
      "grad_norm": 1.0,
      "learning_rate": 0.0003870148359866196,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18204151093959808,
      "num_tokens": 75209857.0,
      "step": 34330
    },
    {
      "entropy": 5.375080108642578,
      "epoch": 3.4076022231044067,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000386984106039025,
      "loss": 5.0821,
      "mean_token_accuracy": 0.18221911489963533,
      "num_tokens": 75220912.0,
      "step": 34335
    },
    {
      "entropy": 5.328965902328491,
      "epoch": 3.408098451766574,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000386953373314466,
      "loss": 4.9501,
      "mean_token_accuracy": 0.18869590610265732,
      "num_tokens": 75232257.0,
      "step": 34340
    },
    {
      "entropy": 5.412779855728149,
      "epoch": 3.4085946804287417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038692263781370446,
      "loss": 5.0401,
      "mean_token_accuracy": 0.17743041664361953,
      "num_tokens": 75244147.0,
      "step": 34345
    },
    {
      "entropy": 5.4882677555084225,
      "epoch": 3.409090909090909,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038689189953750275,
      "loss": 5.0305,
      "mean_token_accuracy": 0.18191228657960892,
      "num_tokens": 75255910.0,
      "step": 34350
    },
    {
      "entropy": 5.398751878738404,
      "epoch": 3.4095871377530766,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038686115848662303,
      "loss": 5.0365,
      "mean_token_accuracy": 0.176483915746212,
      "num_tokens": 75267611.0,
      "step": 34355
    },
    {
      "entropy": 5.349809026718139,
      "epoch": 3.410083366415244,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003868304146618276,
      "loss": 5.0013,
      "mean_token_accuracy": 0.17517462521791458,
      "num_tokens": 75278835.0,
      "step": 34360
    },
    {
      "entropy": 5.463790321350098,
      "epoch": 3.4105795950774116,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038679966806387886,
      "loss": 5.0407,
      "mean_token_accuracy": 0.1774822175502777,
      "num_tokens": 75289317.0,
      "step": 34365
    },
    {
      "entropy": 5.41486701965332,
      "epoch": 3.411075823739579,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003867689186935393,
      "loss": 5.0813,
      "mean_token_accuracy": 0.18284723907709122,
      "num_tokens": 75302304.0,
      "step": 34370
    },
    {
      "entropy": 5.471863508224487,
      "epoch": 3.4115720524017465,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003867381665515712,
      "loss": 5.015,
      "mean_token_accuracy": 0.18828266412019729,
      "num_tokens": 75312081.0,
      "step": 34375
    },
    {
      "entropy": 5.3731639862060545,
      "epoch": 3.412068281063914,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003867074116387376,
      "loss": 4.931,
      "mean_token_accuracy": 0.18170141130685807,
      "num_tokens": 75322948.0,
      "step": 34380
    },
    {
      "entropy": 5.359010839462281,
      "epoch": 3.412564509726082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038667665395580083,
      "loss": 4.9908,
      "mean_token_accuracy": 0.1802266463637352,
      "num_tokens": 75333817.0,
      "step": 34385
    },
    {
      "entropy": 5.372941160202027,
      "epoch": 3.4130607383882494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003866458935035236,
      "loss": 4.9932,
      "mean_token_accuracy": 0.17814208418130875,
      "num_tokens": 75344014.0,
      "step": 34390
    },
    {
      "entropy": 5.443901920318604,
      "epoch": 3.413556967050417,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038661513028266886,
      "loss": 5.0144,
      "mean_token_accuracy": 0.18127166479825974,
      "num_tokens": 75355350.0,
      "step": 34395
    },
    {
      "entropy": 5.406754636764527,
      "epoch": 3.4140531957125844,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003865843642939994,
      "loss": 5.0095,
      "mean_token_accuracy": 0.1791015535593033,
      "num_tokens": 75364633.0,
      "step": 34400
    },
    {
      "entropy": 5.397401332855225,
      "epoch": 3.414549424374752,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038655359553827807,
      "loss": 4.9393,
      "mean_token_accuracy": 0.19137738645076752,
      "num_tokens": 75375103.0,
      "step": 34405
    },
    {
      "entropy": 5.42176628112793,
      "epoch": 3.4150456530369193,
      "grad_norm": 1.0625,
      "learning_rate": 0.000386522824016268,
      "loss": 5.0932,
      "mean_token_accuracy": 0.17962660342454911,
      "num_tokens": 75386112.0,
      "step": 34410
    },
    {
      "entropy": 5.344005489349366,
      "epoch": 3.415541881699087,
      "grad_norm": 1.0,
      "learning_rate": 0.00038649204972873225,
      "loss": 4.877,
      "mean_token_accuracy": 0.18554601520299913,
      "num_tokens": 75396744.0,
      "step": 34415
    },
    {
      "entropy": 5.4310445308685305,
      "epoch": 3.4160381103612547,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003864612726764339,
      "loss": 5.042,
      "mean_token_accuracy": 0.1767111137509346,
      "num_tokens": 75408706.0,
      "step": 34420
    },
    {
      "entropy": 5.437350702285767,
      "epoch": 3.416534339023422,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003864304928601361,
      "loss": 4.9745,
      "mean_token_accuracy": 0.17862947136163712,
      "num_tokens": 75419036.0,
      "step": 34425
    },
    {
      "entropy": 5.329837608337402,
      "epoch": 3.4170305676855897,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003863997102806023,
      "loss": 4.9773,
      "mean_token_accuracy": 0.1884962424635887,
      "num_tokens": 75429415.0,
      "step": 34430
    },
    {
      "entropy": 5.339147901535034,
      "epoch": 3.417526796347757,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003863689249385956,
      "loss": 4.9218,
      "mean_token_accuracy": 0.1930084526538849,
      "num_tokens": 75441077.0,
      "step": 34435
    },
    {
      "entropy": 5.46747236251831,
      "epoch": 3.4180230250099246,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038633813683487954,
      "loss": 5.1119,
      "mean_token_accuracy": 0.17840615957975386,
      "num_tokens": 75452516.0,
      "step": 34440
    },
    {
      "entropy": 5.405384206771851,
      "epoch": 3.418519253672092,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003863073459702177,
      "loss": 5.0175,
      "mean_token_accuracy": 0.18051045387983322,
      "num_tokens": 75463205.0,
      "step": 34445
    },
    {
      "entropy": 5.42024827003479,
      "epoch": 3.4190154823342596,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038627655234537345,
      "loss": 5.0325,
      "mean_token_accuracy": 0.1787765622138977,
      "num_tokens": 75474201.0,
      "step": 34450
    },
    {
      "entropy": 5.388675212860107,
      "epoch": 3.419511710996427,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038624575596111056,
      "loss": 4.9712,
      "mean_token_accuracy": 0.19362995475530625,
      "num_tokens": 75484310.0,
      "step": 34455
    },
    {
      "entropy": 5.335207271575928,
      "epoch": 3.4200079396585945,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038621495681819256,
      "loss": 4.9197,
      "mean_token_accuracy": 0.1823198288679123,
      "num_tokens": 75494583.0,
      "step": 34460
    },
    {
      "entropy": 5.410931491851807,
      "epoch": 3.420504168320762,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003861841549173833,
      "loss": 5.0733,
      "mean_token_accuracy": 0.17791882753372193,
      "num_tokens": 75505176.0,
      "step": 34465
    },
    {
      "entropy": 5.520517444610595,
      "epoch": 3.42100039698293,
      "grad_norm": 1.0,
      "learning_rate": 0.0003861533502594466,
      "loss": 5.0198,
      "mean_token_accuracy": 0.18252875357866288,
      "num_tokens": 75514795.0,
      "step": 34470
    },
    {
      "entropy": 5.456529235839843,
      "epoch": 3.4214966256450974,
      "grad_norm": 1.0,
      "learning_rate": 0.0003861225428451463,
      "loss": 5.0086,
      "mean_token_accuracy": 0.1821325585246086,
      "num_tokens": 75525975.0,
      "step": 34475
    },
    {
      "entropy": 5.393724155426026,
      "epoch": 3.421992854307265,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003860917326752464,
      "loss": 4.9768,
      "mean_token_accuracy": 0.18291890025138854,
      "num_tokens": 75536648.0,
      "step": 34480
    },
    {
      "entropy": 5.46187195777893,
      "epoch": 3.4224890829694323,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038606091975051097,
      "loss": 5.1747,
      "mean_token_accuracy": 0.17189075350761412,
      "num_tokens": 75547724.0,
      "step": 34485
    },
    {
      "entropy": 5.407605791091919,
      "epoch": 3.4229853116316,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038603010407170393,
      "loss": 5.0575,
      "mean_token_accuracy": 0.18027685284614564,
      "num_tokens": 75559061.0,
      "step": 34490
    },
    {
      "entropy": 5.408925724029541,
      "epoch": 3.4234815402937673,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003859992856395896,
      "loss": 5.0171,
      "mean_token_accuracy": 0.17755882441997528,
      "num_tokens": 75569744.0,
      "step": 34495
    },
    {
      "entropy": 5.520258808135987,
      "epoch": 3.4239777689559348,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003859684644549322,
      "loss": 5.1139,
      "mean_token_accuracy": 0.18215175867080688,
      "num_tokens": 75579780.0,
      "step": 34500
    },
    {
      "entropy": 5.407665777206421,
      "epoch": 3.4244739976181022,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003859376405184959,
      "loss": 5.0051,
      "mean_token_accuracy": 0.17896666675806044,
      "num_tokens": 75591526.0,
      "step": 34505
    },
    {
      "entropy": 5.462966775894165,
      "epoch": 3.42497022628027,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003859068138310453,
      "loss": 5.043,
      "mean_token_accuracy": 0.17721709460020066,
      "num_tokens": 75602573.0,
      "step": 34510
    },
    {
      "entropy": 5.374417400360107,
      "epoch": 3.4254664549424376,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038587598439334466,
      "loss": 5.0407,
      "mean_token_accuracy": 0.17427548617124558,
      "num_tokens": 75613774.0,
      "step": 34515
    },
    {
      "entropy": 5.3507321834564205,
      "epoch": 3.425962683604605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038584515220615843,
      "loss": 4.962,
      "mean_token_accuracy": 0.18853309899568557,
      "num_tokens": 75624435.0,
      "step": 34520
    },
    {
      "entropy": 5.373302030563354,
      "epoch": 3.4264589122667726,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038581431727025136,
      "loss": 4.9503,
      "mean_token_accuracy": 0.178922401368618,
      "num_tokens": 75634429.0,
      "step": 34525
    },
    {
      "entropy": 5.349950122833252,
      "epoch": 3.42695514092894,
      "grad_norm": 0.96875,
      "learning_rate": 0.000385783479586388,
      "loss": 4.9406,
      "mean_token_accuracy": 0.1816437542438507,
      "num_tokens": 75646066.0,
      "step": 34530
    },
    {
      "entropy": 5.427351188659668,
      "epoch": 3.4274513695911075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003857526391553331,
      "loss": 5.0047,
      "mean_token_accuracy": 0.18783362060785294,
      "num_tokens": 75655718.0,
      "step": 34535
    },
    {
      "entropy": 5.390956735610962,
      "epoch": 3.427947598253275,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038572179597785134,
      "loss": 4.9411,
      "mean_token_accuracy": 0.19022223204374314,
      "num_tokens": 75666581.0,
      "step": 34540
    },
    {
      "entropy": 5.292997217178344,
      "epoch": 3.4284438269154425,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038569095005470765,
      "loss": 4.9238,
      "mean_token_accuracy": 0.19145753383636474,
      "num_tokens": 75676405.0,
      "step": 34545
    },
    {
      "entropy": 5.323121070861816,
      "epoch": 3.42894005557761,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038566010138666693,
      "loss": 4.8997,
      "mean_token_accuracy": 0.19546955376863479,
      "num_tokens": 75687723.0,
      "step": 34550
    },
    {
      "entropy": 5.413716840744018,
      "epoch": 3.429436284239778,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038562924997449414,
      "loss": 5.0166,
      "mean_token_accuracy": 0.18129447847604752,
      "num_tokens": 75699461.0,
      "step": 34555
    },
    {
      "entropy": 5.428606700897217,
      "epoch": 3.4299325129019453,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003855983958189543,
      "loss": 5.0684,
      "mean_token_accuracy": 0.17306059151887893,
      "num_tokens": 75711679.0,
      "step": 34560
    },
    {
      "entropy": 5.444021797180175,
      "epoch": 3.430428741564113,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003855675389208126,
      "loss": 5.1137,
      "mean_token_accuracy": 0.1730145886540413,
      "num_tokens": 75721027.0,
      "step": 34565
    },
    {
      "entropy": 5.417173528671265,
      "epoch": 3.4309249702262803,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003855366792808341,
      "loss": 4.9811,
      "mean_token_accuracy": 0.18130682408809662,
      "num_tokens": 75731194.0,
      "step": 34570
    },
    {
      "entropy": 5.47132511138916,
      "epoch": 3.4314211988884478,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038550581689978427,
      "loss": 5.0373,
      "mean_token_accuracy": 0.17319101989269256,
      "num_tokens": 75741730.0,
      "step": 34575
    },
    {
      "entropy": 5.421815347671509,
      "epoch": 3.4319174275506152,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038547495177842817,
      "loss": 5.0398,
      "mean_token_accuracy": 0.18507571816444396,
      "num_tokens": 75752980.0,
      "step": 34580
    },
    {
      "entropy": 5.424503993988037,
      "epoch": 3.4324136562127827,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003854440839175314,
      "loss": 5.0957,
      "mean_token_accuracy": 0.17362979650497437,
      "num_tokens": 75764015.0,
      "step": 34585
    },
    {
      "entropy": 5.4450061321258545,
      "epoch": 3.43290988487495,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038541321331785934,
      "loss": 5.0691,
      "mean_token_accuracy": 0.1740783378481865,
      "num_tokens": 75775267.0,
      "step": 34590
    },
    {
      "entropy": 5.508864307403565,
      "epoch": 3.433406113537118,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038538233998017745,
      "loss": 5.0287,
      "mean_token_accuracy": 0.1816992163658142,
      "num_tokens": 75785456.0,
      "step": 34595
    },
    {
      "entropy": 5.3711395263671875,
      "epoch": 3.4339023421992856,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038535146390525143,
      "loss": 4.9505,
      "mean_token_accuracy": 0.18910387754440308,
      "num_tokens": 75795563.0,
      "step": 34600
    },
    {
      "entropy": 5.414090776443482,
      "epoch": 3.434398570861453,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038532058509384694,
      "loss": 5.0091,
      "mean_token_accuracy": 0.17945578396320344,
      "num_tokens": 75807068.0,
      "step": 34605
    },
    {
      "entropy": 5.396866273880005,
      "epoch": 3.4348947995236205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003852897035467296,
      "loss": 5.0026,
      "mean_token_accuracy": 0.17694084197282792,
      "num_tokens": 75818354.0,
      "step": 34610
    },
    {
      "entropy": 5.4343644142150875,
      "epoch": 3.435391028185788,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038525881926466525,
      "loss": 5.0078,
      "mean_token_accuracy": 0.18800454437732697,
      "num_tokens": 75829382.0,
      "step": 34615
    },
    {
      "entropy": 5.403625726699829,
      "epoch": 3.4358872568479555,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038522793224841977,
      "loss": 4.9406,
      "mean_token_accuracy": 0.1845098003745079,
      "num_tokens": 75839278.0,
      "step": 34620
    },
    {
      "entropy": 5.369133377075196,
      "epoch": 3.436383485510123,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038519704249875917,
      "loss": 4.9736,
      "mean_token_accuracy": 0.18054971545934678,
      "num_tokens": 75851074.0,
      "step": 34625
    },
    {
      "entropy": 5.399808788299561,
      "epoch": 3.4368797141722904,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003851661500164493,
      "loss": 5.0304,
      "mean_token_accuracy": 0.17645980268716813,
      "num_tokens": 75862747.0,
      "step": 34630
    },
    {
      "entropy": 5.335109758377075,
      "epoch": 3.437375942834458,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003851352548022563,
      "loss": 4.97,
      "mean_token_accuracy": 0.185017392039299,
      "num_tokens": 75872807.0,
      "step": 34635
    },
    {
      "entropy": 5.405866003036499,
      "epoch": 3.437872171496626,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003851043568569463,
      "loss": 5.0162,
      "mean_token_accuracy": 0.18085665255784988,
      "num_tokens": 75883630.0,
      "step": 34640
    },
    {
      "entropy": 5.461266994476318,
      "epoch": 3.4383684001587933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003850734561812855,
      "loss": 5.0399,
      "mean_token_accuracy": 0.18142334669828414,
      "num_tokens": 75895147.0,
      "step": 34645
    },
    {
      "entropy": 5.415081357955932,
      "epoch": 3.438864628820961,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038504255277604016,
      "loss": 5.0045,
      "mean_token_accuracy": 0.18231966942548752,
      "num_tokens": 75906685.0,
      "step": 34650
    },
    {
      "entropy": 5.332524919509888,
      "epoch": 3.4393608574831283,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003850116466419766,
      "loss": 4.9174,
      "mean_token_accuracy": 0.18524434119462968,
      "num_tokens": 75916859.0,
      "step": 34655
    },
    {
      "entropy": 5.403206920623779,
      "epoch": 3.4398570861452957,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003849807377798614,
      "loss": 5.1031,
      "mean_token_accuracy": 0.17471079528331757,
      "num_tokens": 75927138.0,
      "step": 34660
    },
    {
      "entropy": 5.403244066238403,
      "epoch": 3.440353314807463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038494982619046076,
      "loss": 4.9864,
      "mean_token_accuracy": 0.1805132195353508,
      "num_tokens": 75939542.0,
      "step": 34665
    },
    {
      "entropy": 5.436344146728516,
      "epoch": 3.4408495434696307,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003849189118745414,
      "loss": 5.0633,
      "mean_token_accuracy": 0.18698889315128325,
      "num_tokens": 75951252.0,
      "step": 34670
    },
    {
      "entropy": 5.398120403289795,
      "epoch": 3.441345772131798,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003848879948328699,
      "loss": 4.9904,
      "mean_token_accuracy": 0.17888158112764357,
      "num_tokens": 75962440.0,
      "step": 34675
    },
    {
      "entropy": 5.380470609664917,
      "epoch": 3.441842000793966,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003848570750662129,
      "loss": 4.9374,
      "mean_token_accuracy": 0.18874251693487168,
      "num_tokens": 75971177.0,
      "step": 34680
    },
    {
      "entropy": 5.374517822265625,
      "epoch": 3.4423382294561335,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003848261525753371,
      "loss": 5.0646,
      "mean_token_accuracy": 0.17738610357046128,
      "num_tokens": 75983112.0,
      "step": 34685
    },
    {
      "entropy": 5.429327154159546,
      "epoch": 3.442834458118301,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003847952273610094,
      "loss": 5.0301,
      "mean_token_accuracy": 0.17672648131847382,
      "num_tokens": 75994347.0,
      "step": 34690
    },
    {
      "entropy": 5.4362894058227536,
      "epoch": 3.4433306867804685,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003847642994239968,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1888373613357544,
      "num_tokens": 76004712.0,
      "step": 34695
    },
    {
      "entropy": 5.425391292572021,
      "epoch": 3.443826915442636,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038473336876506585,
      "loss": 5.0496,
      "mean_token_accuracy": 0.17900082916021348,
      "num_tokens": 76014586.0,
      "step": 34700
    },
    {
      "entropy": 5.430922031402588,
      "epoch": 3.4443231441048034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038470243538498393,
      "loss": 4.9695,
      "mean_token_accuracy": 0.18746112883090973,
      "num_tokens": 76024132.0,
      "step": 34705
    },
    {
      "entropy": 5.456184244155883,
      "epoch": 3.444819372766971,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038467149928451803,
      "loss": 5.039,
      "mean_token_accuracy": 0.17347747832536697,
      "num_tokens": 76035122.0,
      "step": 34710
    },
    {
      "entropy": 5.405946922302246,
      "epoch": 3.4453156014291384,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038464056046443523,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1769217222929001,
      "num_tokens": 76047027.0,
      "step": 34715
    },
    {
      "entropy": 5.350467395782471,
      "epoch": 3.445811830091306,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003846096189255027,
      "loss": 4.9763,
      "mean_token_accuracy": 0.17732470333576203,
      "num_tokens": 76057743.0,
      "step": 34720
    },
    {
      "entropy": 5.329965686798095,
      "epoch": 3.446308058753474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038457867466848796,
      "loss": 4.8862,
      "mean_token_accuracy": 0.1896825522184372,
      "num_tokens": 76068433.0,
      "step": 34725
    },
    {
      "entropy": 5.455346298217774,
      "epoch": 3.4468042874156413,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003845477276941582,
      "loss": 5.0335,
      "mean_token_accuracy": 0.18680101186037062,
      "num_tokens": 76080681.0,
      "step": 34730
    },
    {
      "entropy": 5.4059107303619385,
      "epoch": 3.4473005160778087,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003845167780032807,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1779945284128189,
      "num_tokens": 76092584.0,
      "step": 34735
    },
    {
      "entropy": 5.337603425979614,
      "epoch": 3.447796744739976,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038448582559662324,
      "loss": 4.931,
      "mean_token_accuracy": 0.18763526976108552,
      "num_tokens": 76103629.0,
      "step": 34740
    },
    {
      "entropy": 5.401053810119629,
      "epoch": 3.4482929734021437,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038445487047495313,
      "loss": 4.9813,
      "mean_token_accuracy": 0.18550268709659576,
      "num_tokens": 76114069.0,
      "step": 34745
    },
    {
      "entropy": 5.418249273300171,
      "epoch": 3.448789202064311,
      "grad_norm": 1.0,
      "learning_rate": 0.0003844239126390382,
      "loss": 5.0502,
      "mean_token_accuracy": 0.18411883264780043,
      "num_tokens": 76125896.0,
      "step": 34750
    },
    {
      "entropy": 5.41906771659851,
      "epoch": 3.4492854307264786,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038439295208964595,
      "loss": 5.0067,
      "mean_token_accuracy": 0.18579675108194352,
      "num_tokens": 76137066.0,
      "step": 34755
    },
    {
      "entropy": 5.369965982437134,
      "epoch": 3.449781659388646,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003843619888275442,
      "loss": 4.942,
      "mean_token_accuracy": 0.19237753450870515,
      "num_tokens": 76148778.0,
      "step": 34760
    },
    {
      "entropy": 5.3793432235717775,
      "epoch": 3.450277888050814,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038433102285350074,
      "loss": 5.0318,
      "mean_token_accuracy": 0.17664385735988616,
      "num_tokens": 76158760.0,
      "step": 34765
    },
    {
      "entropy": 5.408417892456055,
      "epoch": 3.4507741167129815,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038430005416828354,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18464981019496918,
      "num_tokens": 76169459.0,
      "step": 34770
    },
    {
      "entropy": 5.407119655609131,
      "epoch": 3.451270345375149,
      "grad_norm": 1.0,
      "learning_rate": 0.00038426908277266054,
      "loss": 5.0097,
      "mean_token_accuracy": 0.18407472372055053,
      "num_tokens": 76179074.0,
      "step": 34775
    },
    {
      "entropy": 5.446936464309692,
      "epoch": 3.4517665740373165,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038423810866739975,
      "loss": 5.0175,
      "mean_token_accuracy": 0.18432100862264633,
      "num_tokens": 76189470.0,
      "step": 34780
    },
    {
      "entropy": 5.373384237289429,
      "epoch": 3.452262802699484,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038420713185326926,
      "loss": 4.9627,
      "mean_token_accuracy": 0.18100893795490264,
      "num_tokens": 76200427.0,
      "step": 34785
    },
    {
      "entropy": 5.35047755241394,
      "epoch": 3.4527590313616514,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003841761523310372,
      "loss": 5.0211,
      "mean_token_accuracy": 0.17954607903957367,
      "num_tokens": 76211132.0,
      "step": 34790
    },
    {
      "entropy": 5.416629838943481,
      "epoch": 3.453255260023819,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003841451701014718,
      "loss": 5.03,
      "mean_token_accuracy": 0.181217697262764,
      "num_tokens": 76222661.0,
      "step": 34795
    },
    {
      "entropy": 5.443408679962158,
      "epoch": 3.4537514886859864,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038411418516534135,
      "loss": 5.0103,
      "mean_token_accuracy": 0.18386436104774476,
      "num_tokens": 76232727.0,
      "step": 34800
    },
    {
      "entropy": 5.4601093292236325,
      "epoch": 3.454247717348154,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038408319752341425,
      "loss": 5.0271,
      "mean_token_accuracy": 0.18569287657737732,
      "num_tokens": 76242783.0,
      "step": 34805
    },
    {
      "entropy": 5.378458786010742,
      "epoch": 3.4547439460103213,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000384052207176459,
      "loss": 4.9094,
      "mean_token_accuracy": 0.17868120819330216,
      "num_tokens": 76254223.0,
      "step": 34810
    },
    {
      "entropy": 5.437292814254761,
      "epoch": 3.4552401746724892,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003840212141252438,
      "loss": 5.0406,
      "mean_token_accuracy": 0.17754711359739303,
      "num_tokens": 76265094.0,
      "step": 34815
    },
    {
      "entropy": 5.389361381530762,
      "epoch": 3.4557364033346567,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038399021837053754,
      "loss": 4.9981,
      "mean_token_accuracy": 0.17574485540390014,
      "num_tokens": 76276722.0,
      "step": 34820
    },
    {
      "entropy": 5.388948011398315,
      "epoch": 3.456232631996824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003839592199131088,
      "loss": 4.9753,
      "mean_token_accuracy": 0.18008440285921096,
      "num_tokens": 76288027.0,
      "step": 34825
    },
    {
      "entropy": 5.421294212341309,
      "epoch": 3.4567288606589917,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038392821875372603,
      "loss": 4.9804,
      "mean_token_accuracy": 0.1893290787935257,
      "num_tokens": 76299147.0,
      "step": 34830
    },
    {
      "entropy": 5.401208782196045,
      "epoch": 3.457225089321159,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003838972148931582,
      "loss": 5.1249,
      "mean_token_accuracy": 0.17610291987657548,
      "num_tokens": 76309375.0,
      "step": 34835
    },
    {
      "entropy": 5.402150297164917,
      "epoch": 3.4577213179833266,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003838662083321741,
      "loss": 4.9213,
      "mean_token_accuracy": 0.1861485496163368,
      "num_tokens": 76319953.0,
      "step": 34840
    },
    {
      "entropy": 5.4518993377685545,
      "epoch": 3.458217546645494,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038383519907154263,
      "loss": 5.0188,
      "mean_token_accuracy": 0.1842680111527443,
      "num_tokens": 76331525.0,
      "step": 34845
    },
    {
      "entropy": 5.34639310836792,
      "epoch": 3.458713775307662,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038380418711203275,
      "loss": 4.9241,
      "mean_token_accuracy": 0.19518654942512512,
      "num_tokens": 76343290.0,
      "step": 34850
    },
    {
      "entropy": 5.43895993232727,
      "epoch": 3.4592100039698295,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038377317245441345,
      "loss": 5.1292,
      "mean_token_accuracy": 0.1709782212972641,
      "num_tokens": 76354993.0,
      "step": 34855
    },
    {
      "entropy": 5.404227542877197,
      "epoch": 3.459706232631997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038374215509945386,
      "loss": 5.0604,
      "mean_token_accuracy": 0.1758579894900322,
      "num_tokens": 76365206.0,
      "step": 34860
    },
    {
      "entropy": 5.448332166671753,
      "epoch": 3.4602024612941644,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003837111350479232,
      "loss": 5.0439,
      "mean_token_accuracy": 0.17619819790124894,
      "num_tokens": 76376549.0,
      "step": 34865
    },
    {
      "entropy": 5.438089990615845,
      "epoch": 3.460698689956332,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038368011230059055,
      "loss": 4.9789,
      "mean_token_accuracy": 0.17954225838184357,
      "num_tokens": 76387654.0,
      "step": 34870
    },
    {
      "entropy": 5.485519361495972,
      "epoch": 3.4611949186184994,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038364908685822543,
      "loss": 5.05,
      "mean_token_accuracy": 0.17326681613922118,
      "num_tokens": 76399462.0,
      "step": 34875
    },
    {
      "entropy": 5.438061237335205,
      "epoch": 3.461691147280667,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038361805872159693,
      "loss": 5.0126,
      "mean_token_accuracy": 0.1840846285223961,
      "num_tokens": 76411394.0,
      "step": 34880
    },
    {
      "entropy": 5.41704478263855,
      "epoch": 3.4621873759428343,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003835870278914747,
      "loss": 5.0508,
      "mean_token_accuracy": 0.17361980080604553,
      "num_tokens": 76423251.0,
      "step": 34885
    },
    {
      "entropy": 5.4222430229187015,
      "epoch": 3.462683604605002,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003835559943686282,
      "loss": 4.956,
      "mean_token_accuracy": 0.18402546942234038,
      "num_tokens": 76433828.0,
      "step": 34890
    },
    {
      "entropy": 5.362557411193848,
      "epoch": 3.4631798332671693,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003835249581538269,
      "loss": 4.9271,
      "mean_token_accuracy": 0.18935859054327012,
      "num_tokens": 76445006.0,
      "step": 34895
    },
    {
      "entropy": 5.3230955600738525,
      "epoch": 3.463676061929337,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038349391924784055,
      "loss": 4.9973,
      "mean_token_accuracy": 0.1815429538488388,
      "num_tokens": 76456296.0,
      "step": 34900
    },
    {
      "entropy": 5.3133360862731935,
      "epoch": 3.4641722905915047,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038346287765143866,
      "loss": 4.9285,
      "mean_token_accuracy": 0.18435630798339844,
      "num_tokens": 76467330.0,
      "step": 34905
    },
    {
      "entropy": 5.477258014678955,
      "epoch": 3.464668519253672,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038343183336539123,
      "loss": 5.0648,
      "mean_token_accuracy": 0.17393265664577484,
      "num_tokens": 76478393.0,
      "step": 34910
    },
    {
      "entropy": 5.408223295211792,
      "epoch": 3.4651647479158396,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000383400786390468,
      "loss": 4.9805,
      "mean_token_accuracy": 0.18472786694765092,
      "num_tokens": 76489695.0,
      "step": 34915
    },
    {
      "entropy": 5.354113435745239,
      "epoch": 3.465660976578007,
      "grad_norm": 1.0,
      "learning_rate": 0.00038336973672743876,
      "loss": 4.9847,
      "mean_token_accuracy": 0.17807235866785048,
      "num_tokens": 76500831.0,
      "step": 34920
    },
    {
      "entropy": 5.379671335220337,
      "epoch": 3.4661572052401746,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038333868437707367,
      "loss": 4.9568,
      "mean_token_accuracy": 0.18617449402809144,
      "num_tokens": 76511054.0,
      "step": 34925
    },
    {
      "entropy": 5.3884148597717285,
      "epoch": 3.466653433902342,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003833076293401425,
      "loss": 5.0085,
      "mean_token_accuracy": 0.18022496551275252,
      "num_tokens": 76522948.0,
      "step": 34930
    },
    {
      "entropy": 5.340939235687256,
      "epoch": 3.46714966256451,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003832765716174156,
      "loss": 4.9685,
      "mean_token_accuracy": 0.18865820169448852,
      "num_tokens": 76533827.0,
      "step": 34935
    },
    {
      "entropy": 5.486875724792481,
      "epoch": 3.4676458912266774,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003832455112096631,
      "loss": 5.1142,
      "mean_token_accuracy": 0.18417088985443114,
      "num_tokens": 76545252.0,
      "step": 34940
    },
    {
      "entropy": 5.407190895080566,
      "epoch": 3.468142119888845,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003832144481176551,
      "loss": 4.9557,
      "mean_token_accuracy": 0.18604169636964799,
      "num_tokens": 76555774.0,
      "step": 34945
    },
    {
      "entropy": 5.39130539894104,
      "epoch": 3.4686383485510124,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038318338234216194,
      "loss": 4.9843,
      "mean_token_accuracy": 0.1796030282974243,
      "num_tokens": 76566403.0,
      "step": 34950
    },
    {
      "entropy": 5.3861900806427006,
      "epoch": 3.46913457721318,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038315231388395403,
      "loss": 5.023,
      "mean_token_accuracy": 0.18400018215179442,
      "num_tokens": 76578938.0,
      "step": 34955
    },
    {
      "entropy": 5.4309916496276855,
      "epoch": 3.4696308058753473,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003831212427438018,
      "loss": 5.0524,
      "mean_token_accuracy": 0.18311774730682373,
      "num_tokens": 76589355.0,
      "step": 34960
    },
    {
      "entropy": 5.391110801696778,
      "epoch": 3.470127034537515,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003830901689224757,
      "loss": 4.9459,
      "mean_token_accuracy": 0.18807214945554734,
      "num_tokens": 76601674.0,
      "step": 34965
    },
    {
      "entropy": 5.469771957397461,
      "epoch": 3.4706232631996823,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038305909242074645,
      "loss": 4.9847,
      "mean_token_accuracy": 0.19052737951278687,
      "num_tokens": 76612814.0,
      "step": 34970
    },
    {
      "entropy": 5.351305341720581,
      "epoch": 3.4711194918618498,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003830280132393844,
      "loss": 4.9953,
      "mean_token_accuracy": 0.17712455838918686,
      "num_tokens": 76623177.0,
      "step": 34975
    },
    {
      "entropy": 5.377119398117065,
      "epoch": 3.4716157205240172,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038299693137916046,
      "loss": 5.0075,
      "mean_token_accuracy": 0.18182770311832427,
      "num_tokens": 76633286.0,
      "step": 34980
    },
    {
      "entropy": 5.486572313308716,
      "epoch": 3.472111949186185,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038296584684084536,
      "loss": 5.0482,
      "mean_token_accuracy": 0.18350258022546767,
      "num_tokens": 76644851.0,
      "step": 34985
    },
    {
      "entropy": 5.387198495864868,
      "epoch": 3.4726081778483526,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003829347596252099,
      "loss": 4.9665,
      "mean_token_accuracy": 0.18077638745307922,
      "num_tokens": 76655165.0,
      "step": 34990
    },
    {
      "entropy": 5.348215341567993,
      "epoch": 3.47310440651052,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038290366973302493,
      "loss": 4.9692,
      "mean_token_accuracy": 0.18342647403478624,
      "num_tokens": 76666282.0,
      "step": 34995
    },
    {
      "entropy": 5.419886875152588,
      "epoch": 3.4736006351726876,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038287257716506154,
      "loss": 4.9956,
      "mean_token_accuracy": 0.17941206693649292,
      "num_tokens": 76675600.0,
      "step": 35000
    },
    {
      "entropy": 5.378925561904907,
      "epoch": 3.474096863834855,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038284148192209064,
      "loss": 4.9788,
      "mean_token_accuracy": 0.18490706831216813,
      "num_tokens": 76687412.0,
      "step": 35005
    },
    {
      "entropy": 5.4103155612945555,
      "epoch": 3.4745930924970225,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038281038400488326,
      "loss": 5.0568,
      "mean_token_accuracy": 0.17510709017515183,
      "num_tokens": 76698799.0,
      "step": 35010
    },
    {
      "entropy": 5.44095139503479,
      "epoch": 3.47508932115919,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038277928341421086,
      "loss": 5.0144,
      "mean_token_accuracy": 0.17918485552072524,
      "num_tokens": 76709672.0,
      "step": 35015
    },
    {
      "entropy": 5.3785299301147464,
      "epoch": 3.475585549821358,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003827481801508444,
      "loss": 4.967,
      "mean_token_accuracy": 0.18404096812009813,
      "num_tokens": 76720641.0,
      "step": 35020
    },
    {
      "entropy": 5.328287649154663,
      "epoch": 3.4760817784835254,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038271707421555517,
      "loss": 4.934,
      "mean_token_accuracy": 0.18562791049480437,
      "num_tokens": 76731173.0,
      "step": 35025
    },
    {
      "entropy": 5.461834812164307,
      "epoch": 3.476578007145693,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003826859656091147,
      "loss": 5.058,
      "mean_token_accuracy": 0.177250275015831,
      "num_tokens": 76742801.0,
      "step": 35030
    },
    {
      "entropy": 5.422087717056274,
      "epoch": 3.4770742358078603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003826548543322943,
      "loss": 4.9874,
      "mean_token_accuracy": 0.18487960696220399,
      "num_tokens": 76753088.0,
      "step": 35035
    },
    {
      "entropy": 5.395896005630493,
      "epoch": 3.477570464470028,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003826237403858654,
      "loss": 5.0165,
      "mean_token_accuracy": 0.18404140323400497,
      "num_tokens": 76765715.0,
      "step": 35040
    },
    {
      "entropy": 5.443727779388428,
      "epoch": 3.4780666931321953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003825926237705998,
      "loss": 5.0251,
      "mean_token_accuracy": 0.18218721449375153,
      "num_tokens": 76775626.0,
      "step": 35045
    },
    {
      "entropy": 5.349295711517334,
      "epoch": 3.4785629217943628,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000382561504487269,
      "loss": 4.9017,
      "mean_token_accuracy": 0.19220537394285203,
      "num_tokens": 76785357.0,
      "step": 35050
    },
    {
      "entropy": 5.37229552268982,
      "epoch": 3.4790591504565302,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038253038253664446,
      "loss": 4.9934,
      "mean_token_accuracy": 0.18520217686891555,
      "num_tokens": 76797553.0,
      "step": 35055
    },
    {
      "entropy": 5.341104125976562,
      "epoch": 3.4795553791186977,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038249925791949827,
      "loss": 4.961,
      "mean_token_accuracy": 0.18921779096126556,
      "num_tokens": 76807564.0,
      "step": 35060
    },
    {
      "entropy": 5.407150983810425,
      "epoch": 3.480051607780865,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003824681306366021,
      "loss": 5.0219,
      "mean_token_accuracy": 0.1778075084090233,
      "num_tokens": 76818012.0,
      "step": 35065
    },
    {
      "entropy": 5.440543365478516,
      "epoch": 3.480547836443033,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038243700068872784,
      "loss": 5.0325,
      "mean_token_accuracy": 0.18203782588243483,
      "num_tokens": 76828552.0,
      "step": 35070
    },
    {
      "entropy": 5.495623302459717,
      "epoch": 3.4810440651052006,
      "grad_norm": 1.0,
      "learning_rate": 0.00038240586807664745,
      "loss": 5.0302,
      "mean_token_accuracy": 0.18282975554466246,
      "num_tokens": 76839652.0,
      "step": 35075
    },
    {
      "entropy": 5.400187826156616,
      "epoch": 3.481540293767368,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000382374732801133,
      "loss": 4.9883,
      "mean_token_accuracy": 0.1862170860171318,
      "num_tokens": 76851563.0,
      "step": 35080
    },
    {
      "entropy": 5.36224331855774,
      "epoch": 3.4820365224295355,
      "grad_norm": 1.0,
      "learning_rate": 0.00038234359486295663,
      "loss": 4.941,
      "mean_token_accuracy": 0.18105631470680236,
      "num_tokens": 76862769.0,
      "step": 35085
    },
    {
      "entropy": 5.425376224517822,
      "epoch": 3.482532751091703,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003823124542628904,
      "loss": 5.0061,
      "mean_token_accuracy": 0.17947500199079514,
      "num_tokens": 76873447.0,
      "step": 35090
    },
    {
      "entropy": 5.357279825210571,
      "epoch": 3.4830289797538705,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038228131100170637,
      "loss": 4.9742,
      "mean_token_accuracy": 0.18539926260709763,
      "num_tokens": 76884187.0,
      "step": 35095
    },
    {
      "entropy": 5.435159730911255,
      "epoch": 3.483525208416038,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038225016508017717,
      "loss": 4.9907,
      "mean_token_accuracy": 0.1839264377951622,
      "num_tokens": 76895592.0,
      "step": 35100
    },
    {
      "entropy": 5.466398763656616,
      "epoch": 3.484021437078206,
      "grad_norm": 0.953125,
      "learning_rate": 0.000382219016499075,
      "loss": 5.0579,
      "mean_token_accuracy": 0.1787136137485504,
      "num_tokens": 76907257.0,
      "step": 35105
    },
    {
      "entropy": 5.434612512588501,
      "epoch": 3.4845176657403734,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038218786525917207,
      "loss": 5.0544,
      "mean_token_accuracy": 0.17716766744852067,
      "num_tokens": 76917926.0,
      "step": 35110
    },
    {
      "entropy": 5.412615823745727,
      "epoch": 3.485013894402541,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003821567113612412,
      "loss": 4.9746,
      "mean_token_accuracy": 0.18821098506450654,
      "num_tokens": 76929060.0,
      "step": 35115
    },
    {
      "entropy": 5.331369209289551,
      "epoch": 3.4855101230647083,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038212555480605475,
      "loss": 4.9346,
      "mean_token_accuracy": 0.1881317526102066,
      "num_tokens": 76939694.0,
      "step": 35120
    },
    {
      "entropy": 5.370525598526001,
      "epoch": 3.486006351726876,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038209439559438535,
      "loss": 4.9696,
      "mean_token_accuracy": 0.18972153067588807,
      "num_tokens": 76951015.0,
      "step": 35125
    },
    {
      "entropy": 5.372125053405762,
      "epoch": 3.4865025803890433,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003820632337270057,
      "loss": 5.0198,
      "mean_token_accuracy": 0.17959978729486464,
      "num_tokens": 76962606.0,
      "step": 35130
    },
    {
      "entropy": 5.438270568847656,
      "epoch": 3.4869988090512107,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038203206920468867,
      "loss": 4.9849,
      "mean_token_accuracy": 0.1811990424990654,
      "num_tokens": 76972957.0,
      "step": 35135
    },
    {
      "entropy": 5.352398920059204,
      "epoch": 3.487495037713378,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038200090202820686,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18769451677799226,
      "num_tokens": 76983396.0,
      "step": 35140
    },
    {
      "entropy": 5.363202095031738,
      "epoch": 3.4879912663755457,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038196973219833323,
      "loss": 4.9274,
      "mean_token_accuracy": 0.18814205676317214,
      "num_tokens": 76994340.0,
      "step": 35145
    },
    {
      "entropy": 5.44134840965271,
      "epoch": 3.488487495037713,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038193855971584086,
      "loss": 4.9836,
      "mean_token_accuracy": 0.1819687843322754,
      "num_tokens": 77004723.0,
      "step": 35150
    },
    {
      "entropy": 5.352181720733642,
      "epoch": 3.488983723699881,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038190738458150256,
      "loss": 5.0766,
      "mean_token_accuracy": 0.17352047562599182,
      "num_tokens": 77017031.0,
      "step": 35155
    },
    {
      "entropy": 5.379154396057129,
      "epoch": 3.4894799523620486,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00038187620679609144,
      "loss": 4.9437,
      "mean_token_accuracy": 0.1874968469142914,
      "num_tokens": 77028629.0,
      "step": 35160
    },
    {
      "entropy": 5.439510011672974,
      "epoch": 3.489976181024216,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038184502636038074,
      "loss": 5.0333,
      "mean_token_accuracy": 0.18345480114221574,
      "num_tokens": 77037811.0,
      "step": 35165
    },
    {
      "entropy": 5.510025453567505,
      "epoch": 3.4904724096863835,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003818138432751436,
      "loss": 5.1313,
      "mean_token_accuracy": 0.1804363250732422,
      "num_tokens": 77049213.0,
      "step": 35170
    },
    {
      "entropy": 5.427469062805176,
      "epoch": 3.490968638348551,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003817826575411533,
      "loss": 4.998,
      "mean_token_accuracy": 0.18027654588222503,
      "num_tokens": 77059090.0,
      "step": 35175
    },
    {
      "entropy": 5.3929798126220705,
      "epoch": 3.4914648670107185,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038175146915918323,
      "loss": 4.9776,
      "mean_token_accuracy": 0.188216869533062,
      "num_tokens": 77070092.0,
      "step": 35180
    },
    {
      "entropy": 5.422583293914795,
      "epoch": 3.491961095672886,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038172027813000676,
      "loss": 5.0051,
      "mean_token_accuracy": 0.18388280868530274,
      "num_tokens": 77079808.0,
      "step": 35185
    },
    {
      "entropy": 5.363872432708741,
      "epoch": 3.4924573243350534,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038168908445439725,
      "loss": 4.9733,
      "mean_token_accuracy": 0.18293115496635437,
      "num_tokens": 77090772.0,
      "step": 35190
    },
    {
      "entropy": 5.382416677474976,
      "epoch": 3.4929535529972213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038165788813312844,
      "loss": 4.995,
      "mean_token_accuracy": 0.18196018040180206,
      "num_tokens": 77101358.0,
      "step": 35195
    },
    {
      "entropy": 5.384260892868042,
      "epoch": 3.493449781659389,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038162668916697385,
      "loss": 4.9097,
      "mean_token_accuracy": 0.19096811562776567,
      "num_tokens": 77111198.0,
      "step": 35200
    },
    {
      "entropy": 5.4307451248168945,
      "epoch": 3.4939460103215563,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038159548755670707,
      "loss": 5.0787,
      "mean_token_accuracy": 0.18057731091976165,
      "num_tokens": 77121990.0,
      "step": 35205
    },
    {
      "entropy": 5.421067762374878,
      "epoch": 3.4944422389837237,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038156428330310193,
      "loss": 5.034,
      "mean_token_accuracy": 0.17953920662403106,
      "num_tokens": 77132952.0,
      "step": 35210
    },
    {
      "entropy": 5.400482273101806,
      "epoch": 3.4949384676458912,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003815330764069322,
      "loss": 4.9822,
      "mean_token_accuracy": 0.18442022949457168,
      "num_tokens": 77142870.0,
      "step": 35215
    },
    {
      "entropy": 5.4701642990112305,
      "epoch": 3.4954346963080587,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003815018668689717,
      "loss": 5.0852,
      "mean_token_accuracy": 0.1793295308947563,
      "num_tokens": 77154180.0,
      "step": 35220
    },
    {
      "entropy": 5.461076736450195,
      "epoch": 3.495930924970226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038147065468999435,
      "loss": 5.0742,
      "mean_token_accuracy": 0.1828949585556984,
      "num_tokens": 77165083.0,
      "step": 35225
    },
    {
      "entropy": 5.4393394947052,
      "epoch": 3.4964271536323936,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038143943987077426,
      "loss": 5.0008,
      "mean_token_accuracy": 0.1797925978899002,
      "num_tokens": 77176809.0,
      "step": 35230
    },
    {
      "entropy": 5.3574456691741945,
      "epoch": 3.496923382294561,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003814082224120854,
      "loss": 4.9587,
      "mean_token_accuracy": 0.17892360985279082,
      "num_tokens": 77187325.0,
      "step": 35235
    },
    {
      "entropy": 5.490878915786743,
      "epoch": 3.497419610956729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003813770023147019,
      "loss": 5.0383,
      "mean_token_accuracy": 0.17965473979711533,
      "num_tokens": 77197403.0,
      "step": 35240
    },
    {
      "entropy": 5.376110792160034,
      "epoch": 3.4979158396188965,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038134577957939795,
      "loss": 4.9608,
      "mean_token_accuracy": 0.18787499368190766,
      "num_tokens": 77208262.0,
      "step": 35245
    },
    {
      "entropy": 5.436704921722412,
      "epoch": 3.498412068281064,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038131455420694786,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1873202458024025,
      "num_tokens": 77219327.0,
      "step": 35250
    },
    {
      "entropy": 5.425765943527222,
      "epoch": 3.4989082969432315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003812833261981259,
      "loss": 5.1118,
      "mean_token_accuracy": 0.1814822658896446,
      "num_tokens": 77230380.0,
      "step": 35255
    },
    {
      "entropy": 5.4117053031921385,
      "epoch": 3.499404525605399,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038125209555370644,
      "loss": 4.9466,
      "mean_token_accuracy": 0.18757576644420623,
      "num_tokens": 77241173.0,
      "step": 35260
    },
    {
      "entropy": 5.375750732421875,
      "epoch": 3.4999007542675664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000381220862274464,
      "loss": 4.987,
      "mean_token_accuracy": 0.1754270687699318,
      "num_tokens": 77252674.0,
      "step": 35265
    },
    {
      "entropy": 5.410860347747803,
      "epoch": 3.500396982929734,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038118962636117304,
      "loss": 5.0171,
      "mean_token_accuracy": 0.18462812453508376,
      "num_tokens": 77264826.0,
      "step": 35270
    },
    {
      "entropy": 5.489904069900513,
      "epoch": 3.500893211591902,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038115838781460816,
      "loss": 5.0191,
      "mean_token_accuracy": 0.18076796680688859,
      "num_tokens": 77276068.0,
      "step": 35275
    },
    {
      "entropy": 5.422795629501342,
      "epoch": 3.5013894402540693,
      "grad_norm": 1.078125,
      "learning_rate": 0.000381127146635544,
      "loss": 5.032,
      "mean_token_accuracy": 0.18150772899389267,
      "num_tokens": 77285394.0,
      "step": 35280
    },
    {
      "entropy": 5.385997295379639,
      "epoch": 3.5018856689162368,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003810959028247554,
      "loss": 5.0116,
      "mean_token_accuracy": 0.17525086402893067,
      "num_tokens": 77295843.0,
      "step": 35285
    },
    {
      "entropy": 5.442922592163086,
      "epoch": 3.5023818975784042,
      "grad_norm": 1.0,
      "learning_rate": 0.0003810646563830169,
      "loss": 5.0584,
      "mean_token_accuracy": 0.17888209521770476,
      "num_tokens": 77306064.0,
      "step": 35290
    },
    {
      "entropy": 5.413690376281738,
      "epoch": 3.5028781262405717,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038103340731110354,
      "loss": 4.9561,
      "mean_token_accuracy": 0.18432590812444688,
      "num_tokens": 77316425.0,
      "step": 35295
    },
    {
      "entropy": 5.531447124481201,
      "epoch": 3.503374354902739,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038100215560979016,
      "loss": 5.1476,
      "mean_token_accuracy": 0.1685477688908577,
      "num_tokens": 77327185.0,
      "step": 35300
    },
    {
      "entropy": 5.439564228057861,
      "epoch": 3.5038705835649067,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038097090127985165,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1718806728720665,
      "num_tokens": 77338791.0,
      "step": 35305
    },
    {
      "entropy": 5.424860572814941,
      "epoch": 3.504366812227074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038093964432206316,
      "loss": 4.9751,
      "mean_token_accuracy": 0.18133553266525268,
      "num_tokens": 77350035.0,
      "step": 35310
    },
    {
      "entropy": 5.4239301681518555,
      "epoch": 3.5048630408892416,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003809083847371999,
      "loss": 5.0149,
      "mean_token_accuracy": 0.17882667034864425,
      "num_tokens": 77362422.0,
      "step": 35315
    },
    {
      "entropy": 5.367153692245483,
      "epoch": 3.505359269551409,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003808771225260368,
      "loss": 4.9211,
      "mean_token_accuracy": 0.1844573825597763,
      "num_tokens": 77373383.0,
      "step": 35320
    },
    {
      "entropy": 5.35972409248352,
      "epoch": 3.5058554982135766,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038084585768934917,
      "loss": 5.0099,
      "mean_token_accuracy": 0.18005014210939407,
      "num_tokens": 77383782.0,
      "step": 35325
    },
    {
      "entropy": 5.376561975479126,
      "epoch": 3.5063517268757445,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038081459022791243,
      "loss": 4.9886,
      "mean_token_accuracy": 0.18461474180221557,
      "num_tokens": 77394266.0,
      "step": 35330
    },
    {
      "entropy": 5.377936124801636,
      "epoch": 3.506847955537912,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038078332014250185,
      "loss": 4.927,
      "mean_token_accuracy": 0.1882869556546211,
      "num_tokens": 77405042.0,
      "step": 35335
    },
    {
      "entropy": 5.385925340652466,
      "epoch": 3.5073441842000794,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003807520474338928,
      "loss": 4.9794,
      "mean_token_accuracy": 0.18777644783258438,
      "num_tokens": 77415411.0,
      "step": 35340
    },
    {
      "entropy": 5.37573275566101,
      "epoch": 3.507840412862247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003807207721028609,
      "loss": 4.9339,
      "mean_token_accuracy": 0.1900595873594284,
      "num_tokens": 77423991.0,
      "step": 35345
    },
    {
      "entropy": 5.374426698684692,
      "epoch": 3.5083366415244144,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003806894941501817,
      "loss": 5.0116,
      "mean_token_accuracy": 0.17792379707098008,
      "num_tokens": 77435701.0,
      "step": 35350
    },
    {
      "entropy": 5.358694267272949,
      "epoch": 3.508832870186582,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003806582135766307,
      "loss": 4.9071,
      "mean_token_accuracy": 0.18779427707195281,
      "num_tokens": 77445618.0,
      "step": 35355
    },
    {
      "entropy": 5.366492795944214,
      "epoch": 3.5093290988487498,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038062693038298367,
      "loss": 4.928,
      "mean_token_accuracy": 0.18786842674016951,
      "num_tokens": 77456674.0,
      "step": 35360
    },
    {
      "entropy": 5.33056492805481,
      "epoch": 3.5098253275109172,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003805956445700165,
      "loss": 4.9755,
      "mean_token_accuracy": 0.18193348944187165,
      "num_tokens": 77467841.0,
      "step": 35365
    },
    {
      "entropy": 5.371614408493042,
      "epoch": 3.5103215561730847,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003805643561385048,
      "loss": 4.872,
      "mean_token_accuracy": 0.1883884683251381,
      "num_tokens": 77477883.0,
      "step": 35370
    },
    {
      "entropy": 5.318430852890015,
      "epoch": 3.510817784835252,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038053306508922455,
      "loss": 4.91,
      "mean_token_accuracy": 0.18598982989788054,
      "num_tokens": 77488714.0,
      "step": 35375
    },
    {
      "entropy": 5.313405704498291,
      "epoch": 3.5113140134974197,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038050177142295166,
      "loss": 4.957,
      "mean_token_accuracy": 0.18376001268625258,
      "num_tokens": 77499325.0,
      "step": 35380
    },
    {
      "entropy": 5.356580972671509,
      "epoch": 3.511810242159587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038047047514046223,
      "loss": 4.9602,
      "mean_token_accuracy": 0.17882158756256103,
      "num_tokens": 77510682.0,
      "step": 35385
    },
    {
      "entropy": 5.391067171096802,
      "epoch": 3.5123064708217546,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003804391762425322,
      "loss": 4.9897,
      "mean_token_accuracy": 0.18299271166324615,
      "num_tokens": 77520604.0,
      "step": 35390
    },
    {
      "entropy": 5.388978338241577,
      "epoch": 3.512802699483922,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038040787472993793,
      "loss": 5.034,
      "mean_token_accuracy": 0.18418212831020356,
      "num_tokens": 77532481.0,
      "step": 35395
    },
    {
      "entropy": 5.440984773635864,
      "epoch": 3.5132989281460896,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003803765706034554,
      "loss": 5.0652,
      "mean_token_accuracy": 0.17001494765281677,
      "num_tokens": 77543940.0,
      "step": 35400
    },
    {
      "entropy": 5.406626462936401,
      "epoch": 3.513795156808257,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038034526386386097,
      "loss": 4.8852,
      "mean_token_accuracy": 0.18454800844192504,
      "num_tokens": 77553765.0,
      "step": 35405
    },
    {
      "entropy": 5.413919305801391,
      "epoch": 3.5142913854704245,
      "grad_norm": 1.0,
      "learning_rate": 0.00038031395451193105,
      "loss": 4.9996,
      "mean_token_accuracy": 0.18388922214508058,
      "num_tokens": 77566132.0,
      "step": 35410
    },
    {
      "entropy": 5.380902719497681,
      "epoch": 3.5147876141325924,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038028264254844204,
      "loss": 4.9462,
      "mean_token_accuracy": 0.18644925057888032,
      "num_tokens": 77577706.0,
      "step": 35415
    },
    {
      "entropy": 5.402522945404053,
      "epoch": 3.51528384279476,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038025132797417024,
      "loss": 4.9448,
      "mean_token_accuracy": 0.188187275826931,
      "num_tokens": 77589113.0,
      "step": 35420
    },
    {
      "entropy": 5.316028165817261,
      "epoch": 3.5157800714569274,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003802200107898924,
      "loss": 4.9246,
      "mean_token_accuracy": 0.19087208658456803,
      "num_tokens": 77599012.0,
      "step": 35425
    },
    {
      "entropy": 5.345059633255005,
      "epoch": 3.516276300119095,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038018869099638506,
      "loss": 5.0446,
      "mean_token_accuracy": 0.1783007189631462,
      "num_tokens": 77610389.0,
      "step": 35430
    },
    {
      "entropy": 5.395476531982422,
      "epoch": 3.5167725287812623,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003801573685944247,
      "loss": 4.9085,
      "mean_token_accuracy": 0.18891381323337555,
      "num_tokens": 77621245.0,
      "step": 35435
    },
    {
      "entropy": 5.390676021575928,
      "epoch": 3.51726875744343,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038012604358478823,
      "loss": 4.9345,
      "mean_token_accuracy": 0.18559474349021912,
      "num_tokens": 77631003.0,
      "step": 35440
    },
    {
      "entropy": 5.337047290802002,
      "epoch": 3.5177649861055977,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038009471596825244,
      "loss": 4.8862,
      "mean_token_accuracy": 0.18809996396303177,
      "num_tokens": 77641222.0,
      "step": 35445
    },
    {
      "entropy": 5.434998226165772,
      "epoch": 3.518261214767765,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003800633857455942,
      "loss": 5.0534,
      "mean_token_accuracy": 0.18056721985340118,
      "num_tokens": 77652705.0,
      "step": 35450
    },
    {
      "entropy": 5.4161018371582035,
      "epoch": 3.5187574434299327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003800320529175903,
      "loss": 5.0513,
      "mean_token_accuracy": 0.17902516275644303,
      "num_tokens": 77662883.0,
      "step": 35455
    },
    {
      "entropy": 5.349888896942138,
      "epoch": 3.5192536720921,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038000071748501787,
      "loss": 4.9605,
      "mean_token_accuracy": 0.18033334463834763,
      "num_tokens": 77673979.0,
      "step": 35460
    },
    {
      "entropy": 5.430091714859008,
      "epoch": 3.5197499007542676,
      "grad_norm": 1.125,
      "learning_rate": 0.00037996937944865383,
      "loss": 4.9422,
      "mean_token_accuracy": 0.1870006114244461,
      "num_tokens": 77684387.0,
      "step": 35465
    },
    {
      "entropy": 5.442556142807007,
      "epoch": 3.520246129416435,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003799380388092754,
      "loss": 4.9964,
      "mean_token_accuracy": 0.18336276113986968,
      "num_tokens": 77695218.0,
      "step": 35470
    },
    {
      "entropy": 5.383476686477661,
      "epoch": 3.5207423580786026,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003799066955676598,
      "loss": 4.9973,
      "mean_token_accuracy": 0.18377732038497924,
      "num_tokens": 77705118.0,
      "step": 35475
    },
    {
      "entropy": 5.2790830612182615,
      "epoch": 3.52123858674077,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003798753497245841,
      "loss": 4.9198,
      "mean_token_accuracy": 0.18836069405078887,
      "num_tokens": 77715440.0,
      "step": 35480
    },
    {
      "entropy": 5.412485599517822,
      "epoch": 3.5217348154029375,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037984400128082576,
      "loss": 5.0938,
      "mean_token_accuracy": 0.17732673287391662,
      "num_tokens": 77725146.0,
      "step": 35485
    },
    {
      "entropy": 5.528080463409424,
      "epoch": 3.522231044065105,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037981265023716217,
      "loss": 5.0183,
      "mean_token_accuracy": 0.1842538520693779,
      "num_tokens": 77735987.0,
      "step": 35490
    },
    {
      "entropy": 5.443262147903442,
      "epoch": 3.5227272727272725,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037978129659437063,
      "loss": 4.9428,
      "mean_token_accuracy": 0.18937551975250244,
      "num_tokens": 77747003.0,
      "step": 35495
    },
    {
      "entropy": 5.287317419052124,
      "epoch": 3.5232235013894404,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003797499403532287,
      "loss": 4.9469,
      "mean_token_accuracy": 0.18604174703359605,
      "num_tokens": 77757866.0,
      "step": 35500
    },
    {
      "entropy": 5.371984577178955,
      "epoch": 3.523719730051608,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037971858151451403,
      "loss": 4.9785,
      "mean_token_accuracy": 0.18943011611700059,
      "num_tokens": 77769448.0,
      "step": 35505
    },
    {
      "entropy": 5.437814617156983,
      "epoch": 3.5242159587137754,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037968722007900415,
      "loss": 4.9184,
      "mean_token_accuracy": 0.1917823627591133,
      "num_tokens": 77781207.0,
      "step": 35510
    },
    {
      "entropy": 5.388255834579468,
      "epoch": 3.524712187375943,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037965585604747684,
      "loss": 5.0218,
      "mean_token_accuracy": 0.17293210476636886,
      "num_tokens": 77791050.0,
      "step": 35515
    },
    {
      "entropy": 5.3876560688018795,
      "epoch": 3.5252084160381103,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003796244894207097,
      "loss": 4.9566,
      "mean_token_accuracy": 0.18879414051771165,
      "num_tokens": 77802486.0,
      "step": 35520
    },
    {
      "entropy": 5.36789231300354,
      "epoch": 3.525704644700278,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037959312019948075,
      "loss": 4.9589,
      "mean_token_accuracy": 0.18408476412296296,
      "num_tokens": 77815581.0,
      "step": 35525
    },
    {
      "entropy": 5.393506002426148,
      "epoch": 3.5262008733624453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003795617483845678,
      "loss": 4.9772,
      "mean_token_accuracy": 0.19150342494249345,
      "num_tokens": 77826139.0,
      "step": 35530
    },
    {
      "entropy": 5.469369888305664,
      "epoch": 3.526697102024613,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037953037397674876,
      "loss": 5.0862,
      "mean_token_accuracy": 0.1777235046029091,
      "num_tokens": 77837929.0,
      "step": 35535
    },
    {
      "entropy": 5.378912305831909,
      "epoch": 3.5271933306867806,
      "grad_norm": 1.0,
      "learning_rate": 0.0003794989969768017,
      "loss": 4.9445,
      "mean_token_accuracy": 0.1854723647236824,
      "num_tokens": 77848118.0,
      "step": 35540
    },
    {
      "entropy": 5.390123605728149,
      "epoch": 3.527689559348948,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003794676173855048,
      "loss": 4.9645,
      "mean_token_accuracy": 0.1830987185239792,
      "num_tokens": 77859378.0,
      "step": 35545
    },
    {
      "entropy": 5.355823802947998,
      "epoch": 3.5281857880111156,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00037943623520363595,
      "loss": 4.9656,
      "mean_token_accuracy": 0.18017413467168808,
      "num_tokens": 77872524.0,
      "step": 35550
    },
    {
      "entropy": 5.359723663330078,
      "epoch": 3.528682016673283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003794048504319736,
      "loss": 4.9582,
      "mean_token_accuracy": 0.18424235433340072,
      "num_tokens": 77883053.0,
      "step": 35555
    },
    {
      "entropy": 5.438723802566528,
      "epoch": 3.5291782453354505,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037937346307129584,
      "loss": 5.0848,
      "mean_token_accuracy": 0.18079889416694642,
      "num_tokens": 77894709.0,
      "step": 35560
    },
    {
      "entropy": 5.450226974487305,
      "epoch": 3.529674473997618,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037934207312238116,
      "loss": 5.039,
      "mean_token_accuracy": 0.17841133028268813,
      "num_tokens": 77906213.0,
      "step": 35565
    },
    {
      "entropy": 5.409271144866944,
      "epoch": 3.5301707026597855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037931068058600785,
      "loss": 4.9648,
      "mean_token_accuracy": 0.17717690020799637,
      "num_tokens": 77917015.0,
      "step": 35570
    },
    {
      "entropy": 5.451410722732544,
      "epoch": 3.530666931321953,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037927928546295454,
      "loss": 5.0737,
      "mean_token_accuracy": 0.1789613664150238,
      "num_tokens": 77927504.0,
      "step": 35575
    },
    {
      "entropy": 5.4320765972137455,
      "epoch": 3.5311631599841204,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003792478877539995,
      "loss": 4.9829,
      "mean_token_accuracy": 0.1876608908176422,
      "num_tokens": 77937656.0,
      "step": 35580
    },
    {
      "entropy": 5.450017690658569,
      "epoch": 3.5316593886462884,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003792164874599216,
      "loss": 5.1014,
      "mean_token_accuracy": 0.17469649761915207,
      "num_tokens": 77948626.0,
      "step": 35585
    },
    {
      "entropy": 5.470351505279541,
      "epoch": 3.532155617308456,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003791850845814994,
      "loss": 5.0088,
      "mean_token_accuracy": 0.17825121879577638,
      "num_tokens": 77960055.0,
      "step": 35590
    },
    {
      "entropy": 5.404789066314697,
      "epoch": 3.5326518459706233,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037915367911951146,
      "loss": 4.9837,
      "mean_token_accuracy": 0.18101276010274886,
      "num_tokens": 77971379.0,
      "step": 35595
    },
    {
      "entropy": 5.380224132537842,
      "epoch": 3.533148074632791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037912227107473684,
      "loss": 4.9502,
      "mean_token_accuracy": 0.1851690948009491,
      "num_tokens": 77982746.0,
      "step": 35600
    },
    {
      "entropy": 5.354017066955566,
      "epoch": 3.5336443032949583,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037909086044795427,
      "loss": 5.0053,
      "mean_token_accuracy": 0.1847306400537491,
      "num_tokens": 77994561.0,
      "step": 35605
    },
    {
      "entropy": 5.47634859085083,
      "epoch": 3.5341405319571257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037905944723994257,
      "loss": 5.1131,
      "mean_token_accuracy": 0.1802698105573654,
      "num_tokens": 78004835.0,
      "step": 35610
    },
    {
      "entropy": 5.377177715301514,
      "epoch": 3.534636760619293,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037902803145148085,
      "loss": 4.8897,
      "mean_token_accuracy": 0.18688472509384155,
      "num_tokens": 78015669.0,
      "step": 35615
    },
    {
      "entropy": 5.419980049133301,
      "epoch": 3.535132989281461,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003789966130833481,
      "loss": 5.0333,
      "mean_token_accuracy": 0.17977844476699828,
      "num_tokens": 78025340.0,
      "step": 35620
    },
    {
      "entropy": 5.403515291213989,
      "epoch": 3.5356292179436286,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003789651921363234,
      "loss": 4.9945,
      "mean_token_accuracy": 0.18364507108926773,
      "num_tokens": 78034790.0,
      "step": 35625
    },
    {
      "entropy": 5.432247638702393,
      "epoch": 3.536125446605796,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037893376861118606,
      "loss": 5.0848,
      "mean_token_accuracy": 0.18156198859214784,
      "num_tokens": 78045803.0,
      "step": 35630
    },
    {
      "entropy": 5.389259958267212,
      "epoch": 3.5366216752679636,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037890234250871515,
      "loss": 4.9708,
      "mean_token_accuracy": 0.18246087431907654,
      "num_tokens": 78056690.0,
      "step": 35635
    },
    {
      "entropy": 5.412456750869751,
      "epoch": 3.537117903930131,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037887091382969,
      "loss": 4.9645,
      "mean_token_accuracy": 0.19486597180366516,
      "num_tokens": 78067888.0,
      "step": 35640
    },
    {
      "entropy": 5.39998779296875,
      "epoch": 3.5376141325922985,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003788394825748901,
      "loss": 4.9717,
      "mean_token_accuracy": 0.1846923530101776,
      "num_tokens": 78077877.0,
      "step": 35645
    },
    {
      "entropy": 5.342907571792603,
      "epoch": 3.538110361254466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037880804874509467,
      "loss": 4.9812,
      "mean_token_accuracy": 0.18113266825675964,
      "num_tokens": 78090114.0,
      "step": 35650
    },
    {
      "entropy": 5.4766881465911865,
      "epoch": 3.5386065899166335,
      "grad_norm": 1.125,
      "learning_rate": 0.0003787766123410834,
      "loss": 5.1441,
      "mean_token_accuracy": 0.17572843730449678,
      "num_tokens": 78099737.0,
      "step": 35655
    },
    {
      "entropy": 5.3728187561035154,
      "epoch": 3.539102818578801,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003787451733636357,
      "loss": 4.917,
      "mean_token_accuracy": 0.18693676590919495,
      "num_tokens": 78110979.0,
      "step": 35660
    },
    {
      "entropy": 5.408726692199707,
      "epoch": 3.5395990472409684,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00037871373181353135,
      "loss": 4.9453,
      "mean_token_accuracy": 0.18322530537843704,
      "num_tokens": 78123107.0,
      "step": 35665
    },
    {
      "entropy": 5.4043889999389645,
      "epoch": 3.5400952759031363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037868228769154987,
      "loss": 5.0316,
      "mean_token_accuracy": 0.1826969414949417,
      "num_tokens": 78134128.0,
      "step": 35670
    },
    {
      "entropy": 5.437457132339477,
      "epoch": 3.540591504565304,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003786508409984711,
      "loss": 5.0432,
      "mean_token_accuracy": 0.1726970136165619,
      "num_tokens": 78144561.0,
      "step": 35675
    },
    {
      "entropy": 5.354578161239624,
      "epoch": 3.5410877332274713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037861939173507484,
      "loss": 4.9128,
      "mean_token_accuracy": 0.1909685105085373,
      "num_tokens": 78155700.0,
      "step": 35680
    },
    {
      "entropy": 5.462960433959961,
      "epoch": 3.5415839618896388,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037858793990214087,
      "loss": 5.0616,
      "mean_token_accuracy": 0.17929994016885759,
      "num_tokens": 78166904.0,
      "step": 35685
    },
    {
      "entropy": 5.330289840698242,
      "epoch": 3.5420801905518062,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037855648550044925,
      "loss": 4.9407,
      "mean_token_accuracy": 0.18712171018123627,
      "num_tokens": 78178409.0,
      "step": 35690
    },
    {
      "entropy": 5.460771274566651,
      "epoch": 3.5425764192139737,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003785250285307799,
      "loss": 5.0154,
      "mean_token_accuracy": 0.17856587320566178,
      "num_tokens": 78190639.0,
      "step": 35695
    },
    {
      "entropy": 5.415554666519165,
      "epoch": 3.543072647876141,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037849356899391295,
      "loss": 4.9938,
      "mean_token_accuracy": 0.182511043548584,
      "num_tokens": 78202345.0,
      "step": 35700
    },
    {
      "entropy": 5.344880628585815,
      "epoch": 3.543568876538309,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003784621068906285,
      "loss": 4.8889,
      "mean_token_accuracy": 0.18997801542282106,
      "num_tokens": 78212767.0,
      "step": 35705
    },
    {
      "entropy": 5.300110149383545,
      "epoch": 3.5440651052004766,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037843064222170677,
      "loss": 4.9341,
      "mean_token_accuracy": 0.1847782999277115,
      "num_tokens": 78223338.0,
      "step": 35710
    },
    {
      "entropy": 5.333527231216431,
      "epoch": 3.544561333862644,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037839917498792787,
      "loss": 4.957,
      "mean_token_accuracy": 0.18786016702651978,
      "num_tokens": 78234903.0,
      "step": 35715
    },
    {
      "entropy": 5.361989259719849,
      "epoch": 3.5450575625248115,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037836770519007236,
      "loss": 4.9764,
      "mean_token_accuracy": 0.1883337378501892,
      "num_tokens": 78245936.0,
      "step": 35720
    },
    {
      "entropy": 5.3568510055542,
      "epoch": 3.545553791186979,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003783362328289205,
      "loss": 4.9878,
      "mean_token_accuracy": 0.18780911415815355,
      "num_tokens": 78256294.0,
      "step": 35725
    },
    {
      "entropy": 5.434877395629883,
      "epoch": 3.5460500198491465,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003783047579052526,
      "loss": 5.0407,
      "mean_token_accuracy": 0.18264245092868805,
      "num_tokens": 78266385.0,
      "step": 35730
    },
    {
      "entropy": 5.416906833648682,
      "epoch": 3.546546248511314,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037827328041984944,
      "loss": 4.9553,
      "mean_token_accuracy": 0.1909652054309845,
      "num_tokens": 78276044.0,
      "step": 35735
    },
    {
      "entropy": 5.396893167495728,
      "epoch": 3.5470424771734814,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003782418003734914,
      "loss": 4.9923,
      "mean_token_accuracy": 0.1774732366204262,
      "num_tokens": 78286943.0,
      "step": 35740
    },
    {
      "entropy": 5.408217906951904,
      "epoch": 3.547538705835649,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003782103177669592,
      "loss": 4.9424,
      "mean_token_accuracy": 0.19021696150302886,
      "num_tokens": 78298313.0,
      "step": 35745
    },
    {
      "entropy": 5.407057857513427,
      "epoch": 3.5480349344978164,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037817883260103354,
      "loss": 5.0164,
      "mean_token_accuracy": 0.17613512873649598,
      "num_tokens": 78309750.0,
      "step": 35750
    },
    {
      "entropy": 5.314993667602539,
      "epoch": 3.548531163159984,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037814734487649516,
      "loss": 4.9202,
      "mean_token_accuracy": 0.18784330785274506,
      "num_tokens": 78321998.0,
      "step": 35755
    },
    {
      "entropy": 5.448619031906128,
      "epoch": 3.5490273918221518,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003781158545941248,
      "loss": 5.0979,
      "mean_token_accuracy": 0.1738488644361496,
      "num_tokens": 78333973.0,
      "step": 35760
    },
    {
      "entropy": 5.5138466358184814,
      "epoch": 3.5495236204843192,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037808436175470344,
      "loss": 5.0936,
      "mean_token_accuracy": 0.17825772762298583,
      "num_tokens": 78345281.0,
      "step": 35765
    },
    {
      "entropy": 5.4530738353729244,
      "epoch": 3.5500198491464867,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003780528663590122,
      "loss": 5.0912,
      "mean_token_accuracy": 0.17892976254224777,
      "num_tokens": 78356995.0,
      "step": 35770
    },
    {
      "entropy": 5.429361152648926,
      "epoch": 3.550516077808654,
      "grad_norm": 1.0,
      "learning_rate": 0.0003780213684078318,
      "loss": 4.981,
      "mean_token_accuracy": 0.18156906664371492,
      "num_tokens": 78367910.0,
      "step": 35775
    },
    {
      "entropy": 5.370654010772705,
      "epoch": 3.5510123064708217,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037798986790194347,
      "loss": 4.9222,
      "mean_token_accuracy": 0.19085272997617722,
      "num_tokens": 78378995.0,
      "step": 35780
    },
    {
      "entropy": 5.447018671035766,
      "epoch": 3.551508535132989,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037795836484212837,
      "loss": 5.0149,
      "mean_token_accuracy": 0.1837821200489998,
      "num_tokens": 78389881.0,
      "step": 35785
    },
    {
      "entropy": 5.321249389648438,
      "epoch": 3.552004763795157,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037792685922916763,
      "loss": 4.875,
      "mean_token_accuracy": 0.18481937795877457,
      "num_tokens": 78400242.0,
      "step": 35790
    },
    {
      "entropy": 5.331661176681519,
      "epoch": 3.5525009924573245,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037789535106384266,
      "loss": 4.8891,
      "mean_token_accuracy": 0.18241699337959288,
      "num_tokens": 78410057.0,
      "step": 35795
    },
    {
      "entropy": 5.343530130386353,
      "epoch": 3.552997221119492,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003778638403469346,
      "loss": 5.0019,
      "mean_token_accuracy": 0.18498039692640306,
      "num_tokens": 78419684.0,
      "step": 35800
    },
    {
      "entropy": 5.353309917449951,
      "epoch": 3.5534934497816595,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037783232707922507,
      "loss": 4.9259,
      "mean_token_accuracy": 0.18322086632251738,
      "num_tokens": 78430569.0,
      "step": 35805
    },
    {
      "entropy": 5.47004337310791,
      "epoch": 3.553989678443827,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037780081126149523,
      "loss": 4.9913,
      "mean_token_accuracy": 0.18547094613313675,
      "num_tokens": 78441290.0,
      "step": 35810
    },
    {
      "entropy": 5.4068831443786625,
      "epoch": 3.5544859071059944,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037776929289452696,
      "loss": 4.9824,
      "mean_token_accuracy": 0.18604299426078796,
      "num_tokens": 78451638.0,
      "step": 35815
    },
    {
      "entropy": 5.42078742980957,
      "epoch": 3.554982135768162,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003777377719791016,
      "loss": 4.9858,
      "mean_token_accuracy": 0.18741633594036103,
      "num_tokens": 78462759.0,
      "step": 35820
    },
    {
      "entropy": 5.330410289764404,
      "epoch": 3.5554783644303294,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037770624851600086,
      "loss": 4.9736,
      "mean_token_accuracy": 0.1883223757147789,
      "num_tokens": 78475465.0,
      "step": 35825
    },
    {
      "entropy": 5.312044095993042,
      "epoch": 3.555974593092497,
      "grad_norm": 1.0,
      "learning_rate": 0.0003776747225060065,
      "loss": 4.8613,
      "mean_token_accuracy": 0.1850554257631302,
      "num_tokens": 78487900.0,
      "step": 35830
    },
    {
      "entropy": 5.352413845062256,
      "epoch": 3.5564708217546643,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003776431939499003,
      "loss": 4.9051,
      "mean_token_accuracy": 0.19020790606737137,
      "num_tokens": 78497538.0,
      "step": 35835
    },
    {
      "entropy": 5.388669347763061,
      "epoch": 3.556967050416832,
      "grad_norm": 0.9375,
      "learning_rate": 0.000377611662848464,
      "loss": 4.9957,
      "mean_token_accuracy": 0.18331486731767654,
      "num_tokens": 78508299.0,
      "step": 35840
    },
    {
      "entropy": 5.360615062713623,
      "epoch": 3.5574632790789997,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003775801292024796,
      "loss": 4.8989,
      "mean_token_accuracy": 0.19157810807228087,
      "num_tokens": 78518188.0,
      "step": 35845
    },
    {
      "entropy": 5.35721869468689,
      "epoch": 3.557959507741167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037754859301272905,
      "loss": 4.9861,
      "mean_token_accuracy": 0.18445123881101608,
      "num_tokens": 78528390.0,
      "step": 35850
    },
    {
      "entropy": 5.314705944061279,
      "epoch": 3.5584557364033347,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037751705427999436,
      "loss": 5.0438,
      "mean_token_accuracy": 0.18262302577495576,
      "num_tokens": 78540266.0,
      "step": 35855
    },
    {
      "entropy": 5.405879211425781,
      "epoch": 3.558951965065502,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003774855130050576,
      "loss": 5.0227,
      "mean_token_accuracy": 0.1781691014766693,
      "num_tokens": 78550937.0,
      "step": 35860
    },
    {
      "entropy": 5.406703901290894,
      "epoch": 3.5594481937276696,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037745396918870095,
      "loss": 4.9457,
      "mean_token_accuracy": 0.18231339901685714,
      "num_tokens": 78561619.0,
      "step": 35865
    },
    {
      "entropy": 5.328273487091065,
      "epoch": 3.559944422389837,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003774224228317067,
      "loss": 4.9387,
      "mean_token_accuracy": 0.18810761123895645,
      "num_tokens": 78573840.0,
      "step": 35870
    },
    {
      "entropy": 5.315991687774658,
      "epoch": 3.560440651052005,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000377390873934857,
      "loss": 4.9766,
      "mean_token_accuracy": 0.18078119605779647,
      "num_tokens": 78585477.0,
      "step": 35875
    },
    {
      "entropy": 5.453083753585815,
      "epoch": 3.5609368797141725,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003773593224989343,
      "loss": 5.0896,
      "mean_token_accuracy": 0.178088241815567,
      "num_tokens": 78597281.0,
      "step": 35880
    },
    {
      "entropy": 5.387767744064331,
      "epoch": 3.56143310837634,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003773277685247209,
      "loss": 4.9682,
      "mean_token_accuracy": 0.1855141893029213,
      "num_tokens": 78609300.0,
      "step": 35885
    },
    {
      "entropy": 5.440272283554077,
      "epoch": 3.5619293370385074,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037729621201299933,
      "loss": 4.993,
      "mean_token_accuracy": 0.18676108121871948,
      "num_tokens": 78620443.0,
      "step": 35890
    },
    {
      "entropy": 5.394781970977784,
      "epoch": 3.562425565700675,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003772646529645522,
      "loss": 4.9754,
      "mean_token_accuracy": 0.183668614923954,
      "num_tokens": 78630113.0,
      "step": 35895
    },
    {
      "entropy": 5.386165523529053,
      "epoch": 3.5629217943628424,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003772330913801621,
      "loss": 4.9183,
      "mean_token_accuracy": 0.18434310853481292,
      "num_tokens": 78640092.0,
      "step": 35900
    },
    {
      "entropy": 5.384771203994751,
      "epoch": 3.56341802302501,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003772015272606115,
      "loss": 4.9428,
      "mean_token_accuracy": 0.1896195501089096,
      "num_tokens": 78651092.0,
      "step": 35905
    },
    {
      "entropy": 5.323941373825074,
      "epoch": 3.5639142516871773,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003771699606066833,
      "loss": 4.9769,
      "mean_token_accuracy": 0.187209752202034,
      "num_tokens": 78662435.0,
      "step": 35910
    },
    {
      "entropy": 5.363180875778198,
      "epoch": 3.564410480349345,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003771383914191602,
      "loss": 4.9579,
      "mean_token_accuracy": 0.18060072958469392,
      "num_tokens": 78673299.0,
      "step": 35915
    },
    {
      "entropy": 5.408843517303467,
      "epoch": 3.5649067090115123,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003771068196988251,
      "loss": 4.9744,
      "mean_token_accuracy": 0.1868484303355217,
      "num_tokens": 78683286.0,
      "step": 35920
    },
    {
      "entropy": 5.368365001678467,
      "epoch": 3.5654029376736798,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037707524544646094,
      "loss": 4.9157,
      "mean_token_accuracy": 0.18708733916282655,
      "num_tokens": 78695282.0,
      "step": 35925
    },
    {
      "entropy": 5.310205554962158,
      "epoch": 3.5658991663358477,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003770436686628506,
      "loss": 4.9492,
      "mean_token_accuracy": 0.18208978176116944,
      "num_tokens": 78706880.0,
      "step": 35930
    },
    {
      "entropy": 5.442976379394532,
      "epoch": 3.566395394998015,
      "grad_norm": 1.0,
      "learning_rate": 0.00037701208934877716,
      "loss": 4.991,
      "mean_token_accuracy": 0.17986783683300017,
      "num_tokens": 78717515.0,
      "step": 35935
    },
    {
      "entropy": 5.417661952972412,
      "epoch": 3.5668916236601826,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037698050750502375,
      "loss": 4.963,
      "mean_token_accuracy": 0.18276464492082595,
      "num_tokens": 78728959.0,
      "step": 35940
    },
    {
      "entropy": 5.32805118560791,
      "epoch": 3.56738785232235,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003769489231323735,
      "loss": 4.9124,
      "mean_token_accuracy": 0.18387677520513535,
      "num_tokens": 78740580.0,
      "step": 35945
    },
    {
      "entropy": 5.391315221786499,
      "epoch": 3.5678840809845176,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003769173362316096,
      "loss": 4.9523,
      "mean_token_accuracy": 0.18548175394535066,
      "num_tokens": 78751388.0,
      "step": 35950
    },
    {
      "entropy": 5.385159492492676,
      "epoch": 3.568380309646685,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037688574680351544,
      "loss": 5.0226,
      "mean_token_accuracy": 0.17867135107517243,
      "num_tokens": 78763474.0,
      "step": 35955
    },
    {
      "entropy": 5.334959363937378,
      "epoch": 3.568876538308853,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037685415484887426,
      "loss": 4.9622,
      "mean_token_accuracy": 0.1834461659193039,
      "num_tokens": 78775570.0,
      "step": 35960
    },
    {
      "entropy": 5.388579893112182,
      "epoch": 3.5693727669710205,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037682256036846947,
      "loss": 4.9438,
      "mean_token_accuracy": 0.18081192970275878,
      "num_tokens": 78786558.0,
      "step": 35965
    },
    {
      "entropy": 5.405415391921997,
      "epoch": 3.569868995633188,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037679096336308473,
      "loss": 4.9675,
      "mean_token_accuracy": 0.18648997098207473,
      "num_tokens": 78796872.0,
      "step": 35970
    },
    {
      "entropy": 5.4038349151611325,
      "epoch": 3.5703652242953554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037675936383350326,
      "loss": 4.966,
      "mean_token_accuracy": 0.17778484523296356,
      "num_tokens": 78807971.0,
      "step": 35975
    },
    {
      "entropy": 5.397624111175537,
      "epoch": 3.570861452957523,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003767277617805089,
      "loss": 5.0166,
      "mean_token_accuracy": 0.18026231676340104,
      "num_tokens": 78818593.0,
      "step": 35980
    },
    {
      "entropy": 5.397526264190674,
      "epoch": 3.5713576816196904,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003766961572048853,
      "loss": 4.9841,
      "mean_token_accuracy": 0.18494823724031448,
      "num_tokens": 78830573.0,
      "step": 35985
    },
    {
      "entropy": 5.397661638259888,
      "epoch": 3.571853910281858,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003766645501074161,
      "loss": 5.0373,
      "mean_token_accuracy": 0.18071119338274003,
      "num_tokens": 78842041.0,
      "step": 35990
    },
    {
      "entropy": 5.409921360015869,
      "epoch": 3.5723501389440253,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037663294048888513,
      "loss": 4.9946,
      "mean_token_accuracy": 0.18527311682701111,
      "num_tokens": 78852549.0,
      "step": 35995
    },
    {
      "entropy": 5.433162450790405,
      "epoch": 3.572846367606193,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037660132835007625,
      "loss": 5.064,
      "mean_token_accuracy": 0.18733496963977814,
      "num_tokens": 78862706.0,
      "step": 36000
    },
    {
      "epoch": 3.572846367606193,
      "eval_entropy": 5.214634483742527,
      "eval_loss": 5.140158176422119,
      "eval_mean_token_accuracy": 0.18137347217083657,
      "eval_num_tokens": 78862706.0,
      "eval_runtime": 26.6001,
      "eval_samples_per_second": 1265.746,
      "eval_steps_per_second": 158.232,
      "step": 36000
    },
    {
      "entropy": 5.418843126296997,
      "epoch": 3.5733425962683603,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003765697136917734,
      "loss": 5.033,
      "mean_token_accuracy": 0.17715100347995758,
      "num_tokens": 78874214.0,
      "step": 36005
    },
    {
      "entropy": 5.418095588684082,
      "epoch": 3.5738388249305277,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003765380965147604,
      "loss": 5.0623,
      "mean_token_accuracy": 0.18120596408843995,
      "num_tokens": 78886178.0,
      "step": 36010
    },
    {
      "entropy": 5.371176815032959,
      "epoch": 3.5743350535926957,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037650647681982145,
      "loss": 4.9475,
      "mean_token_accuracy": 0.18180815428495406,
      "num_tokens": 78896114.0,
      "step": 36015
    },
    {
      "entropy": 5.453980207443237,
      "epoch": 3.574831282254863,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037647485460774065,
      "loss": 4.9619,
      "mean_token_accuracy": 0.1900964081287384,
      "num_tokens": 78906447.0,
      "step": 36020
    },
    {
      "entropy": 5.336492729187012,
      "epoch": 3.5753275109170306,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003764432298793021,
      "loss": 4.8692,
      "mean_token_accuracy": 0.18858197182416916,
      "num_tokens": 78917737.0,
      "step": 36025
    },
    {
      "entropy": 5.3382774829864506,
      "epoch": 3.575823739579198,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037641160263529003,
      "loss": 4.9912,
      "mean_token_accuracy": 0.18276125490665435,
      "num_tokens": 78929682.0,
      "step": 36030
    },
    {
      "entropy": 5.365615463256836,
      "epoch": 3.5763199682413656,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037637997287648877,
      "loss": 4.9375,
      "mean_token_accuracy": 0.19095201939344406,
      "num_tokens": 78941875.0,
      "step": 36035
    },
    {
      "entropy": 5.3991344451904295,
      "epoch": 3.576816196903533,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037634834060368256,
      "loss": 4.9959,
      "mean_token_accuracy": 0.18141243159770964,
      "num_tokens": 78951891.0,
      "step": 36040
    },
    {
      "entropy": 5.45051589012146,
      "epoch": 3.577312425565701,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003763167058176559,
      "loss": 5.0797,
      "mean_token_accuracy": 0.178875832259655,
      "num_tokens": 78964141.0,
      "step": 36045
    },
    {
      "entropy": 5.400818920135498,
      "epoch": 3.5778086542278684,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003762850685191933,
      "loss": 4.969,
      "mean_token_accuracy": 0.1919109046459198,
      "num_tokens": 78973644.0,
      "step": 36050
    },
    {
      "entropy": 5.327863693237305,
      "epoch": 3.578304882890036,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003762534287090792,
      "loss": 4.9698,
      "mean_token_accuracy": 0.18246931880712508,
      "num_tokens": 78984666.0,
      "step": 36055
    },
    {
      "entropy": 5.3494460105896,
      "epoch": 3.5788011115522034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003762217863880983,
      "loss": 4.9278,
      "mean_token_accuracy": 0.18810644149780273,
      "num_tokens": 78996064.0,
      "step": 36060
    },
    {
      "entropy": 5.452240514755249,
      "epoch": 3.579297340214371,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003761901415570351,
      "loss": 5.0429,
      "mean_token_accuracy": 0.17291307896375657,
      "num_tokens": 79006733.0,
      "step": 36065
    },
    {
      "entropy": 5.423266267776489,
      "epoch": 3.5797935688765383,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037615849421667466,
      "loss": 5.0045,
      "mean_token_accuracy": 0.18291288912296294,
      "num_tokens": 79018027.0,
      "step": 36070
    },
    {
      "entropy": 5.410660028457642,
      "epoch": 3.580289797538706,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003761268443678014,
      "loss": 4.9951,
      "mean_token_accuracy": 0.1811971679329872,
      "num_tokens": 79028281.0,
      "step": 36075
    },
    {
      "entropy": 5.426606845855713,
      "epoch": 3.5807860262008733,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003760951920112003,
      "loss": 4.9587,
      "mean_token_accuracy": 0.182107050716877,
      "num_tokens": 79039110.0,
      "step": 36080
    },
    {
      "entropy": 5.445136213302613,
      "epoch": 3.5812822548630407,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037606353714765626,
      "loss": 5.0505,
      "mean_token_accuracy": 0.18885496258735657,
      "num_tokens": 79051294.0,
      "step": 36085
    },
    {
      "entropy": 5.379706621170044,
      "epoch": 3.5817784835252082,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003760318797779543,
      "loss": 5.0657,
      "mean_token_accuracy": 0.1762421950697899,
      "num_tokens": 79062790.0,
      "step": 36090
    },
    {
      "entropy": 5.389630460739136,
      "epoch": 3.5822747121873757,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003760002199028794,
      "loss": 5.0164,
      "mean_token_accuracy": 0.184000563621521,
      "num_tokens": 79073699.0,
      "step": 36095
    },
    {
      "entropy": 5.392884969711304,
      "epoch": 3.5827709408495436,
      "grad_norm": 1.0,
      "learning_rate": 0.0003759685575232168,
      "loss": 4.8978,
      "mean_token_accuracy": 0.18859148472547532,
      "num_tokens": 79084588.0,
      "step": 36100
    },
    {
      "entropy": 5.334650707244873,
      "epoch": 3.583267169511711,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003759368926397514,
      "loss": 4.9864,
      "mean_token_accuracy": 0.1866183191537857,
      "num_tokens": 79094767.0,
      "step": 36105
    },
    {
      "entropy": 5.343237257003784,
      "epoch": 3.5837633981738786,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003759052252532686,
      "loss": 4.9861,
      "mean_token_accuracy": 0.18212573528289794,
      "num_tokens": 79105186.0,
      "step": 36110
    },
    {
      "entropy": 5.360953283309937,
      "epoch": 3.584259626836046,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003758735553645537,
      "loss": 4.9547,
      "mean_token_accuracy": 0.1865053340792656,
      "num_tokens": 79116184.0,
      "step": 36115
    },
    {
      "entropy": 5.3608787059783936,
      "epoch": 3.5847558554982135,
      "grad_norm": 1.09375,
      "learning_rate": 0.000375841882974392,
      "loss": 4.933,
      "mean_token_accuracy": 0.18503332287073135,
      "num_tokens": 79126676.0,
      "step": 36120
    },
    {
      "entropy": 5.4335822582244875,
      "epoch": 3.585252084160381,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003758102080835688,
      "loss": 5.0882,
      "mean_token_accuracy": 0.17358813881874086,
      "num_tokens": 79138090.0,
      "step": 36125
    },
    {
      "entropy": 5.387348222732544,
      "epoch": 3.5857483128225485,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003757785306928697,
      "loss": 4.9457,
      "mean_token_accuracy": 0.19484758824110032,
      "num_tokens": 79150087.0,
      "step": 36130
    },
    {
      "entropy": 5.414798021316528,
      "epoch": 3.5862445414847164,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0003757468508030803,
      "loss": 5.03,
      "mean_token_accuracy": 0.1867468073964119,
      "num_tokens": 79162296.0,
      "step": 36135
    },
    {
      "entropy": 5.401835918426514,
      "epoch": 3.586740770146884,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003757151684149859,
      "loss": 5.0139,
      "mean_token_accuracy": 0.1841002196073532,
      "num_tokens": 79173763.0,
      "step": 36140
    },
    {
      "entropy": 5.445752716064453,
      "epoch": 3.5872369988090513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003756834835293725,
      "loss": 5.0603,
      "mean_token_accuracy": 0.18351387679576875,
      "num_tokens": 79184323.0,
      "step": 36145
    },
    {
      "entropy": 5.392802715301514,
      "epoch": 3.587733227471219,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003756517961470256,
      "loss": 5.0238,
      "mean_token_accuracy": 0.17688233703374862,
      "num_tokens": 79195811.0,
      "step": 36150
    },
    {
      "entropy": 5.477212715148926,
      "epoch": 3.5882294561333863,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003756201062687311,
      "loss": 5.0852,
      "mean_token_accuracy": 0.18201405107975005,
      "num_tokens": 79207243.0,
      "step": 36155
    },
    {
      "entropy": 5.421522235870361,
      "epoch": 3.5887256847955538,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037558841389527466,
      "loss": 5.0184,
      "mean_token_accuracy": 0.18423397094011307,
      "num_tokens": 79218909.0,
      "step": 36160
    },
    {
      "entropy": 5.433234119415284,
      "epoch": 3.5892219134577212,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037555671902744244,
      "loss": 4.9779,
      "mean_token_accuracy": 0.18365270346403123,
      "num_tokens": 79229170.0,
      "step": 36165
    },
    {
      "entropy": 5.418811798095703,
      "epoch": 3.5897181421198887,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037552502166602014,
      "loss": 5.0108,
      "mean_token_accuracy": 0.1849789097905159,
      "num_tokens": 79240001.0,
      "step": 36170
    },
    {
      "entropy": 5.406105709075928,
      "epoch": 3.590214370782056,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037549332181179403,
      "loss": 5.0304,
      "mean_token_accuracy": 0.1839681535959244,
      "num_tokens": 79251525.0,
      "step": 36175
    },
    {
      "entropy": 5.4210621356964115,
      "epoch": 3.5907105994442237,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037546161946555004,
      "loss": 5.046,
      "mean_token_accuracy": 0.1771787226200104,
      "num_tokens": 79264125.0,
      "step": 36180
    },
    {
      "entropy": 5.383986711502075,
      "epoch": 3.5912068281063916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037542991462807433,
      "loss": 4.947,
      "mean_token_accuracy": 0.18529943376779556,
      "num_tokens": 79273880.0,
      "step": 36185
    },
    {
      "entropy": 5.384987258911133,
      "epoch": 3.591703056768559,
      "grad_norm": 1.0,
      "learning_rate": 0.00037539820730015316,
      "loss": 4.9979,
      "mean_token_accuracy": 0.1807626962661743,
      "num_tokens": 79284542.0,
      "step": 36190
    },
    {
      "entropy": 5.325160980224609,
      "epoch": 3.5921992854307265,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037536649748257285,
      "loss": 4.8853,
      "mean_token_accuracy": 0.19323171824216842,
      "num_tokens": 79295182.0,
      "step": 36195
    },
    {
      "entropy": 5.292289543151855,
      "epoch": 3.592695514092894,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037533478517611955,
      "loss": 4.9326,
      "mean_token_accuracy": 0.19627914130687713,
      "num_tokens": 79305613.0,
      "step": 36200
    },
    {
      "entropy": 5.39312424659729,
      "epoch": 3.5931917427550615,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003753030703815799,
      "loss": 5.0003,
      "mean_token_accuracy": 0.18827341943979264,
      "num_tokens": 79316566.0,
      "step": 36205
    },
    {
      "entropy": 5.368745279312134,
      "epoch": 3.593687971417229,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003752713530997402,
      "loss": 4.9544,
      "mean_token_accuracy": 0.18960920721292496,
      "num_tokens": 79327224.0,
      "step": 36210
    },
    {
      "entropy": 5.448434734344483,
      "epoch": 3.5941842000793964,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00037523963333138693,
      "loss": 5.0596,
      "mean_token_accuracy": 0.17596156746149064,
      "num_tokens": 79338069.0,
      "step": 36215
    },
    {
      "entropy": 5.4690680503845215,
      "epoch": 3.5946804287415643,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003752079110773068,
      "loss": 5.0889,
      "mean_token_accuracy": 0.17901404500007628,
      "num_tokens": 79348677.0,
      "step": 36220
    },
    {
      "entropy": 5.31224479675293,
      "epoch": 3.595176657403732,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003751761863382865,
      "loss": 4.9628,
      "mean_token_accuracy": 0.1866798773407936,
      "num_tokens": 79359253.0,
      "step": 36225
    },
    {
      "entropy": 5.350127172470093,
      "epoch": 3.5956728860658993,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037514445911511257,
      "loss": 4.9187,
      "mean_token_accuracy": 0.1904713675379753,
      "num_tokens": 79370409.0,
      "step": 36230
    },
    {
      "entropy": 5.368513917922973,
      "epoch": 3.5961691147280668,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003751127294085719,
      "loss": 4.9949,
      "mean_token_accuracy": 0.1790927916765213,
      "num_tokens": 79381259.0,
      "step": 36235
    },
    {
      "entropy": 5.342346906661987,
      "epoch": 3.5966653433902342,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037508099721945113,
      "loss": 4.9226,
      "mean_token_accuracy": 0.18286459296941757,
      "num_tokens": 79391723.0,
      "step": 36240
    },
    {
      "entropy": 5.3965864181518555,
      "epoch": 3.5971615720524017,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003750492625485374,
      "loss": 4.9583,
      "mean_token_accuracy": 0.18446271121501923,
      "num_tokens": 79401015.0,
      "step": 36245
    },
    {
      "entropy": 5.316458415985108,
      "epoch": 3.597657800714569,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003750175253966175,
      "loss": 4.8737,
      "mean_token_accuracy": 0.1972323626279831,
      "num_tokens": 79411923.0,
      "step": 36250
    },
    {
      "entropy": 5.367449855804443,
      "epoch": 3.5981540293767367,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003749857857644785,
      "loss": 4.9671,
      "mean_token_accuracy": 0.1795554429292679,
      "num_tokens": 79423934.0,
      "step": 36255
    },
    {
      "entropy": 5.332741022109985,
      "epoch": 3.598650258038904,
      "grad_norm": 1.0,
      "learning_rate": 0.0003749540436529074,
      "loss": 4.9343,
      "mean_token_accuracy": 0.1936354085803032,
      "num_tokens": 79433718.0,
      "step": 36260
    },
    {
      "entropy": 5.391946744918823,
      "epoch": 3.5991464867010716,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037492229906269147,
      "loss": 4.9479,
      "mean_token_accuracy": 0.17904288619756697,
      "num_tokens": 79443935.0,
      "step": 36265
    },
    {
      "entropy": 5.39227409362793,
      "epoch": 3.5996427153632395,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037489055199461783,
      "loss": 4.9353,
      "mean_token_accuracy": 0.19500216841697693,
      "num_tokens": 79455442.0,
      "step": 36270
    },
    {
      "entropy": 5.366465997695923,
      "epoch": 3.600138944025407,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003748588024494737,
      "loss": 4.9338,
      "mean_token_accuracy": 0.18555177003145218,
      "num_tokens": 79466591.0,
      "step": 36275
    },
    {
      "entropy": 5.396089553833008,
      "epoch": 3.6006351726875745,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037482705042804655,
      "loss": 5.0626,
      "mean_token_accuracy": 0.17544782906770706,
      "num_tokens": 79477064.0,
      "step": 36280
    },
    {
      "entropy": 5.454707050323487,
      "epoch": 3.601131401349742,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037479529593112355,
      "loss": 5.0423,
      "mean_token_accuracy": 0.18436944633722305,
      "num_tokens": 79488799.0,
      "step": 36285
    },
    {
      "entropy": 5.416239547729492,
      "epoch": 3.6016276300119094,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003747635389594923,
      "loss": 4.9867,
      "mean_token_accuracy": 0.18848845213651658,
      "num_tokens": 79499406.0,
      "step": 36290
    },
    {
      "entropy": 5.32406325340271,
      "epoch": 3.602123858674077,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037473177951394027,
      "loss": 4.8948,
      "mean_token_accuracy": 0.1919069543480873,
      "num_tokens": 79509347.0,
      "step": 36295
    },
    {
      "entropy": 5.473528861999512,
      "epoch": 3.6026200873362444,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037470001759525495,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17670040726661682,
      "num_tokens": 79520515.0,
      "step": 36300
    },
    {
      "entropy": 5.335767459869385,
      "epoch": 3.6031163159984123,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037466825320422407,
      "loss": 4.8888,
      "mean_token_accuracy": 0.18801933228969575,
      "num_tokens": 79530563.0,
      "step": 36305
    },
    {
      "entropy": 5.351221084594727,
      "epoch": 3.60361254466058,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037463648634163534,
      "loss": 4.9997,
      "mean_token_accuracy": 0.17989348918199538,
      "num_tokens": 79541578.0,
      "step": 36310
    },
    {
      "entropy": 5.333033227920533,
      "epoch": 3.6041087733227473,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003746047170082764,
      "loss": 4.8965,
      "mean_token_accuracy": 0.19223072081804277,
      "num_tokens": 79552170.0,
      "step": 36315
    },
    {
      "entropy": 5.408327960968018,
      "epoch": 3.6046050019849147,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037457294520493504,
      "loss": 5.0112,
      "mean_token_accuracy": 0.18291790783405304,
      "num_tokens": 79562205.0,
      "step": 36320
    },
    {
      "entropy": 5.404194688796997,
      "epoch": 3.605101230647082,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037454117093239925,
      "loss": 5.0038,
      "mean_token_accuracy": 0.18465862721204757,
      "num_tokens": 79572683.0,
      "step": 36325
    },
    {
      "entropy": 5.439736795425415,
      "epoch": 3.6055974593092497,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003745093941914569,
      "loss": 5.0595,
      "mean_token_accuracy": 0.1681389942765236,
      "num_tokens": 79583453.0,
      "step": 36330
    },
    {
      "entropy": 5.393152332305908,
      "epoch": 3.606093687971417,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037447761498289595,
      "loss": 4.9234,
      "mean_token_accuracy": 0.1820240542292595,
      "num_tokens": 79593902.0,
      "step": 36335
    },
    {
      "entropy": 5.346751976013183,
      "epoch": 3.6065899166335846,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003744458333075046,
      "loss": 4.9378,
      "mean_token_accuracy": 0.18576498478651046,
      "num_tokens": 79606511.0,
      "step": 36340
    },
    {
      "entropy": 5.402992391586304,
      "epoch": 3.607086145295752,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003744140491660708,
      "loss": 4.9582,
      "mean_token_accuracy": 0.18034877628087997,
      "num_tokens": 79617082.0,
      "step": 36345
    },
    {
      "entropy": 5.441939973831177,
      "epoch": 3.6075823739579196,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003743822625593829,
      "loss": 5.0455,
      "mean_token_accuracy": 0.18259080052375792,
      "num_tokens": 79627721.0,
      "step": 36350
    },
    {
      "entropy": 5.391405820846558,
      "epoch": 3.608078602620087,
      "grad_norm": 1.046875,
      "learning_rate": 0.000374350473488229,
      "loss": 4.9737,
      "mean_token_accuracy": 0.1845546305179596,
      "num_tokens": 79637464.0,
      "step": 36355
    },
    {
      "entropy": 5.323117780685425,
      "epoch": 3.608574831282255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037431868195339735,
      "loss": 4.9779,
      "mean_token_accuracy": 0.1822695627808571,
      "num_tokens": 79647128.0,
      "step": 36360
    },
    {
      "entropy": 5.334505319595337,
      "epoch": 3.6090710599444225,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003742868879556764,
      "loss": 4.9246,
      "mean_token_accuracy": 0.1911790519952774,
      "num_tokens": 79658569.0,
      "step": 36365
    },
    {
      "entropy": 5.407737159729004,
      "epoch": 3.60956728860659,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037425509149585466,
      "loss": 4.9941,
      "mean_token_accuracy": 0.1841077536344528,
      "num_tokens": 79669660.0,
      "step": 36370
    },
    {
      "entropy": 5.431107139587402,
      "epoch": 3.6100635172687574,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003742232925747205,
      "loss": 4.9954,
      "mean_token_accuracy": 0.17439635396003722,
      "num_tokens": 79680017.0,
      "step": 36375
    },
    {
      "entropy": 5.319172525405884,
      "epoch": 3.610559745930925,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003741914911930625,
      "loss": 4.9269,
      "mean_token_accuracy": 0.20251391381025313,
      "num_tokens": 79691291.0,
      "step": 36380
    },
    {
      "entropy": 5.370221185684204,
      "epoch": 3.6110559745930924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003741596873516693,
      "loss": 4.9884,
      "mean_token_accuracy": 0.1846747502684593,
      "num_tokens": 79701539.0,
      "step": 36385
    },
    {
      "entropy": 5.354009675979614,
      "epoch": 3.6115522032552603,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003741278810513294,
      "loss": 4.9731,
      "mean_token_accuracy": 0.19188348948955536,
      "num_tokens": 79711869.0,
      "step": 36390
    },
    {
      "entropy": 5.3571418762207035,
      "epoch": 3.6120484319174277,
      "grad_norm": 1.0,
      "learning_rate": 0.00037409607229283175,
      "loss": 5.0323,
      "mean_token_accuracy": 0.1866254076361656,
      "num_tokens": 79722171.0,
      "step": 36395
    },
    {
      "entropy": 5.43939733505249,
      "epoch": 3.6125446605795952,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000374064261076965,
      "loss": 4.9852,
      "mean_token_accuracy": 0.1816947028040886,
      "num_tokens": 79733877.0,
      "step": 36400
    },
    {
      "entropy": 5.3657063961029055,
      "epoch": 3.6130408892417627,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000374032447404518,
      "loss": 4.9512,
      "mean_token_accuracy": 0.18462421745061874,
      "num_tokens": 79744494.0,
      "step": 36405
    },
    {
      "entropy": 5.429012966156006,
      "epoch": 3.61353711790393,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003740006312762799,
      "loss": 5.0544,
      "mean_token_accuracy": 0.17551883608102797,
      "num_tokens": 79755989.0,
      "step": 36410
    },
    {
      "entropy": 5.44075756072998,
      "epoch": 3.6140333465660976,
      "grad_norm": 1.640625,
      "learning_rate": 0.00037396881269303924,
      "loss": 5.0728,
      "mean_token_accuracy": 0.18064682483673095,
      "num_tokens": 79767184.0,
      "step": 36415
    },
    {
      "entropy": 5.431661128997803,
      "epoch": 3.614529575228265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037393699165558534,
      "loss": 5.0123,
      "mean_token_accuracy": 0.18394550532102585,
      "num_tokens": 79778493.0,
      "step": 36420
    },
    {
      "entropy": 5.392191982269287,
      "epoch": 3.6150258038904326,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003739051681647073,
      "loss": 5.0096,
      "mean_token_accuracy": 0.18513911217451096,
      "num_tokens": 79787998.0,
      "step": 36425
    },
    {
      "entropy": 5.421238040924072,
      "epoch": 3.6155220325526,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037387334222119424,
      "loss": 5.0526,
      "mean_token_accuracy": 0.18224956542253495,
      "num_tokens": 79799469.0,
      "step": 36430
    },
    {
      "entropy": 5.406774568557739,
      "epoch": 3.6160182612147675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037384151382583526,
      "loss": 4.9458,
      "mean_token_accuracy": 0.18570978790521622,
      "num_tokens": 79808656.0,
      "step": 36435
    },
    {
      "entropy": 5.39739499092102,
      "epoch": 3.616514489876935,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003738096829794197,
      "loss": 4.9616,
      "mean_token_accuracy": 0.1873839318752289,
      "num_tokens": 79818127.0,
      "step": 36440
    },
    {
      "entropy": 5.431743383407593,
      "epoch": 3.617010718539103,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000373777849682737,
      "loss": 4.9898,
      "mean_token_accuracy": 0.18464479446411133,
      "num_tokens": 79830451.0,
      "step": 36445
    },
    {
      "entropy": 5.41642918586731,
      "epoch": 3.6175069472012704,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003737460139365764,
      "loss": 5.0356,
      "mean_token_accuracy": 0.18072081357240677,
      "num_tokens": 79842623.0,
      "step": 36450
    },
    {
      "entropy": 5.407133722305298,
      "epoch": 3.618003175863438,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003737141757417275,
      "loss": 4.9558,
      "mean_token_accuracy": 0.18460366278886794,
      "num_tokens": 79852208.0,
      "step": 36455
    },
    {
      "entropy": 5.395886945724487,
      "epoch": 3.6184994045256054,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003736823350989797,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18225777447223662,
      "num_tokens": 79862356.0,
      "step": 36460
    },
    {
      "entropy": 5.360284805297852,
      "epoch": 3.618995633187773,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037365049200912274,
      "loss": 4.9934,
      "mean_token_accuracy": 0.18238188475370407,
      "num_tokens": 79873045.0,
      "step": 36465
    },
    {
      "entropy": 5.455604553222656,
      "epoch": 3.6194918618499403,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037361864647294596,
      "loss": 5.0141,
      "mean_token_accuracy": 0.18533723652362824,
      "num_tokens": 79882738.0,
      "step": 36470
    },
    {
      "entropy": 5.4361992359161375,
      "epoch": 3.6199880905121082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003735867984912394,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18025058656930923,
      "num_tokens": 79894047.0,
      "step": 36475
    },
    {
      "entropy": 5.367565107345581,
      "epoch": 3.6204843191742757,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037355494806479264,
      "loss": 4.9284,
      "mean_token_accuracy": 0.18727579414844514,
      "num_tokens": 79905607.0,
      "step": 36480
    },
    {
      "entropy": 5.364643383026123,
      "epoch": 3.620980547836443,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003735230951943955,
      "loss": 5.0454,
      "mean_token_accuracy": 0.18340959250926972,
      "num_tokens": 79917995.0,
      "step": 36485
    },
    {
      "entropy": 5.422769927978516,
      "epoch": 3.6214767764986107,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037349123988083795,
      "loss": 4.9924,
      "mean_token_accuracy": 0.18397824019193648,
      "num_tokens": 79929600.0,
      "step": 36490
    },
    {
      "entropy": 5.420239734649658,
      "epoch": 3.621973005160778,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037345938212490987,
      "loss": 5.0063,
      "mean_token_accuracy": 0.1862459659576416,
      "num_tokens": 79940597.0,
      "step": 36495
    },
    {
      "entropy": 5.46601095199585,
      "epoch": 3.6224692338229456,
      "grad_norm": 0.875,
      "learning_rate": 0.00037342752192740127,
      "loss": 4.9911,
      "mean_token_accuracy": 0.18656711876392365,
      "num_tokens": 79951531.0,
      "step": 36500
    },
    {
      "entropy": 5.400364875793457,
      "epoch": 3.622965462485113,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003733956592891022,
      "loss": 5.0132,
      "mean_token_accuracy": 0.1786504328250885,
      "num_tokens": 79962651.0,
      "step": 36505
    },
    {
      "entropy": 5.3964368343353275,
      "epoch": 3.6234616911472806,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037336379421080274,
      "loss": 5.0143,
      "mean_token_accuracy": 0.18541034311056137,
      "num_tokens": 79973219.0,
      "step": 36510
    },
    {
      "entropy": 5.408788967132568,
      "epoch": 3.623957919809448,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003733319266932932,
      "loss": 4.958,
      "mean_token_accuracy": 0.18277024179697038,
      "num_tokens": 79985037.0,
      "step": 36515
    },
    {
      "entropy": 5.424822616577148,
      "epoch": 3.6244541484716155,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037330005673736385,
      "loss": 5.0473,
      "mean_token_accuracy": 0.17528901994228363,
      "num_tokens": 79994596.0,
      "step": 36520
    },
    {
      "entropy": 5.459197235107422,
      "epoch": 3.624950377133783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003732681843438048,
      "loss": 4.992,
      "mean_token_accuracy": 0.18327537178993225,
      "num_tokens": 80006045.0,
      "step": 36525
    },
    {
      "entropy": 5.398852348327637,
      "epoch": 3.625446605795951,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037323630951340654,
      "loss": 4.9709,
      "mean_token_accuracy": 0.17811596691608428,
      "num_tokens": 80016423.0,
      "step": 36530
    },
    {
      "entropy": 5.402554941177368,
      "epoch": 3.6259428344581184,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037320443224695955,
      "loss": 5.0263,
      "mean_token_accuracy": 0.17886980772018432,
      "num_tokens": 80026579.0,
      "step": 36535
    },
    {
      "entropy": 5.365498685836792,
      "epoch": 3.626439063120286,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003731725525452542,
      "loss": 4.9563,
      "mean_token_accuracy": 0.18633898347616196,
      "num_tokens": 80037227.0,
      "step": 36540
    },
    {
      "entropy": 5.413009405136108,
      "epoch": 3.6269352917824533,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003731406704090811,
      "loss": 4.9984,
      "mean_token_accuracy": 0.1819295823574066,
      "num_tokens": 80049421.0,
      "step": 36545
    },
    {
      "entropy": 5.426993656158447,
      "epoch": 3.627431520444621,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037310878583923085,
      "loss": 4.9873,
      "mean_token_accuracy": 0.18611658066511155,
      "num_tokens": 80059710.0,
      "step": 36550
    },
    {
      "entropy": 5.403588438034058,
      "epoch": 3.6279277491067883,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003730768988364941,
      "loss": 4.9711,
      "mean_token_accuracy": 0.18591624051332473,
      "num_tokens": 80071403.0,
      "step": 36555
    },
    {
      "entropy": 5.3399097442626955,
      "epoch": 3.628423977768956,
      "grad_norm": 1.0,
      "learning_rate": 0.00037304500940166165,
      "loss": 4.8564,
      "mean_token_accuracy": 0.20060357600450515,
      "num_tokens": 80082978.0,
      "step": 36560
    },
    {
      "entropy": 5.400908899307251,
      "epoch": 3.6289202064311237,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037301311753552407,
      "loss": 5.0323,
      "mean_token_accuracy": 0.190423583984375,
      "num_tokens": 80093002.0,
      "step": 36565
    },
    {
      "entropy": 5.408464002609253,
      "epoch": 3.629416435093291,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003729812232388726,
      "loss": 4.9873,
      "mean_token_accuracy": 0.18474471867084502,
      "num_tokens": 80104738.0,
      "step": 36570
    },
    {
      "entropy": 5.328877544403076,
      "epoch": 3.6299126637554586,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003729493265124978,
      "loss": 4.9411,
      "mean_token_accuracy": 0.18286317139863967,
      "num_tokens": 80113746.0,
      "step": 36575
    },
    {
      "entropy": 5.463904094696045,
      "epoch": 3.630408892417626,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037291742735719077,
      "loss": 5.0322,
      "mean_token_accuracy": 0.17980654984712602,
      "num_tokens": 80126558.0,
      "step": 36580
    },
    {
      "entropy": 5.443748998641968,
      "epoch": 3.6309051210797936,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003728855257737425,
      "loss": 5.0431,
      "mean_token_accuracy": 0.17913201600313186,
      "num_tokens": 80138174.0,
      "step": 36585
    },
    {
      "entropy": 5.382057619094849,
      "epoch": 3.631401349741961,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003728536217629442,
      "loss": 4.9651,
      "mean_token_accuracy": 0.1829465374350548,
      "num_tokens": 80148044.0,
      "step": 36590
    },
    {
      "entropy": 5.450418329238891,
      "epoch": 3.6318975784041285,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037282171532558697,
      "loss": 5.048,
      "mean_token_accuracy": 0.17781291604042054,
      "num_tokens": 80158725.0,
      "step": 36595
    },
    {
      "entropy": 5.518087530136109,
      "epoch": 3.632393807066296,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037278980646246186,
      "loss": 5.121,
      "mean_token_accuracy": 0.18071606606245041,
      "num_tokens": 80170360.0,
      "step": 36600
    },
    {
      "entropy": 5.4039207935333256,
      "epoch": 3.6328900357284635,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037275789517436035,
      "loss": 4.9533,
      "mean_token_accuracy": 0.18123001009225845,
      "num_tokens": 80180167.0,
      "step": 36605
    },
    {
      "entropy": 5.314858293533325,
      "epoch": 3.633386264390631,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037272598146207363,
      "loss": 4.8895,
      "mean_token_accuracy": 0.18532888144254683,
      "num_tokens": 80190867.0,
      "step": 36610
    },
    {
      "entropy": 5.400083780288696,
      "epoch": 3.633882493052799,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037269406532639324,
      "loss": 4.95,
      "mean_token_accuracy": 0.18761820048093797,
      "num_tokens": 80201258.0,
      "step": 36615
    },
    {
      "entropy": 5.366726064682007,
      "epoch": 3.6343787217149663,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003726621467681105,
      "loss": 4.9957,
      "mean_token_accuracy": 0.18349240571260453,
      "num_tokens": 80212499.0,
      "step": 36620
    },
    {
      "entropy": 5.350361442565918,
      "epoch": 3.634874950377134,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003726302257880169,
      "loss": 4.9537,
      "mean_token_accuracy": 0.1846958354115486,
      "num_tokens": 80223611.0,
      "step": 36625
    },
    {
      "entropy": 5.359004926681519,
      "epoch": 3.6353711790393013,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003725983023869041,
      "loss": 4.9755,
      "mean_token_accuracy": 0.18218985944986343,
      "num_tokens": 80233678.0,
      "step": 36630
    },
    {
      "entropy": 5.44401478767395,
      "epoch": 3.6358674077014688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003725663765655638,
      "loss": 5.0499,
      "mean_token_accuracy": 0.18066326677799224,
      "num_tokens": 80246307.0,
      "step": 36635
    },
    {
      "entropy": 5.352375793457031,
      "epoch": 3.6363636363636362,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003725344483247875,
      "loss": 4.843,
      "mean_token_accuracy": 0.1946706786751747,
      "num_tokens": 80256945.0,
      "step": 36640
    },
    {
      "entropy": 5.340617609024048,
      "epoch": 3.636859865025804,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037250251766536715,
      "loss": 4.981,
      "mean_token_accuracy": 0.18700138628482818,
      "num_tokens": 80266299.0,
      "step": 36645
    },
    {
      "entropy": 5.351466512680053,
      "epoch": 3.6373560936879716,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003724705845880944,
      "loss": 4.9995,
      "mean_token_accuracy": 0.18007356971502303,
      "num_tokens": 80276004.0,
      "step": 36650
    },
    {
      "entropy": 5.435752248764038,
      "epoch": 3.637852322350139,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003724386490937611,
      "loss": 5.0472,
      "mean_token_accuracy": 0.17607703059911728,
      "num_tokens": 80285962.0,
      "step": 36655
    },
    {
      "entropy": 5.454711532592773,
      "epoch": 3.6383485510123066,
      "grad_norm": 1.0,
      "learning_rate": 0.0003724067111831594,
      "loss": 5.043,
      "mean_token_accuracy": 0.18673239797353744,
      "num_tokens": 80295457.0,
      "step": 36660
    },
    {
      "entropy": 5.318313407897949,
      "epoch": 3.638844779674474,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003723747708570811,
      "loss": 4.9196,
      "mean_token_accuracy": 0.19290270954370498,
      "num_tokens": 80307125.0,
      "step": 36665
    },
    {
      "entropy": 5.3700926303863525,
      "epoch": 3.6393410083366415,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037234282811631826,
      "loss": 4.9519,
      "mean_token_accuracy": 0.17962894439697266,
      "num_tokens": 80318629.0,
      "step": 36670
    },
    {
      "entropy": 5.43610143661499,
      "epoch": 3.639837236998809,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003723108829616631,
      "loss": 5.0482,
      "mean_token_accuracy": 0.18013983070850373,
      "num_tokens": 80330901.0,
      "step": 36675
    },
    {
      "entropy": 5.382685565948487,
      "epoch": 3.6403334656609765,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003722789353939077,
      "loss": 4.928,
      "mean_token_accuracy": 0.1903747573494911,
      "num_tokens": 80341888.0,
      "step": 36680
    },
    {
      "entropy": 5.373898553848266,
      "epoch": 3.640829694323144,
      "grad_norm": 1.0,
      "learning_rate": 0.00037224698541384434,
      "loss": 4.8989,
      "mean_token_accuracy": 0.18764120787382127,
      "num_tokens": 80352903.0,
      "step": 36685
    },
    {
      "entropy": 5.340475845336914,
      "epoch": 3.6413259229853114,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037221503302226533,
      "loss": 4.9242,
      "mean_token_accuracy": 0.18956861793994903,
      "num_tokens": 80364609.0,
      "step": 36690
    },
    {
      "entropy": 5.321328163146973,
      "epoch": 3.641822151647479,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003721830782199629,
      "loss": 4.971,
      "mean_token_accuracy": 0.18437660187482835,
      "num_tokens": 80375431.0,
      "step": 36695
    },
    {
      "entropy": 5.359927368164063,
      "epoch": 3.642318380309647,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003721511210077296,
      "loss": 4.9969,
      "mean_token_accuracy": 0.18463854789733886,
      "num_tokens": 80385503.0,
      "step": 36700
    },
    {
      "entropy": 5.419887447357178,
      "epoch": 3.6428146089718143,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037211916138635775,
      "loss": 5.0404,
      "mean_token_accuracy": 0.178862327337265,
      "num_tokens": 80396752.0,
      "step": 36705
    },
    {
      "entropy": 5.3956279277801515,
      "epoch": 3.643310837633982,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037208719935664003,
      "loss": 4.9474,
      "mean_token_accuracy": 0.18474121689796447,
      "num_tokens": 80407342.0,
      "step": 36710
    },
    {
      "entropy": 5.398024702072144,
      "epoch": 3.6438070662961493,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000372055234919369,
      "loss": 4.9303,
      "mean_token_accuracy": 0.18755723983049394,
      "num_tokens": 80418356.0,
      "step": 36715
    },
    {
      "entropy": 5.358344507217407,
      "epoch": 3.6443032949583167,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037202326807533724,
      "loss": 4.9364,
      "mean_token_accuracy": 0.18923420161008836,
      "num_tokens": 80428579.0,
      "step": 36720
    },
    {
      "entropy": 5.44270281791687,
      "epoch": 3.644799523620484,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003719912988253376,
      "loss": 5.088,
      "mean_token_accuracy": 0.17531269192695617,
      "num_tokens": 80441494.0,
      "step": 36725
    },
    {
      "entropy": 5.436481857299805,
      "epoch": 3.6452957522826517,
      "grad_norm": 1.0,
      "learning_rate": 0.00037195932717016254,
      "loss": 5.0167,
      "mean_token_accuracy": 0.18323474526405334,
      "num_tokens": 80453766.0,
      "step": 36730
    },
    {
      "entropy": 5.426730155944824,
      "epoch": 3.6457919809448196,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003719273531106053,
      "loss": 4.9132,
      "mean_token_accuracy": 0.18974177837371825,
      "num_tokens": 80464382.0,
      "step": 36735
    },
    {
      "entropy": 5.441700458526611,
      "epoch": 3.646288209606987,
      "grad_norm": 1.0,
      "learning_rate": 0.00037189537664745846,
      "loss": 5.0898,
      "mean_token_accuracy": 0.1817108139395714,
      "num_tokens": 80475605.0,
      "step": 36740
    },
    {
      "entropy": 5.382927799224854,
      "epoch": 3.6467844382691545,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003718633977815151,
      "loss": 4.9781,
      "mean_token_accuracy": 0.18295138478279113,
      "num_tokens": 80485612.0,
      "step": 36745
    },
    {
      "entropy": 5.423728799819946,
      "epoch": 3.647280666931322,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037183141651356814,
      "loss": 5.0087,
      "mean_token_accuracy": 0.18407478779554368,
      "num_tokens": 80497651.0,
      "step": 36750
    },
    {
      "entropy": 5.447868394851684,
      "epoch": 3.6477768955934895,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003717994328444108,
      "loss": 5.072,
      "mean_token_accuracy": 0.1887338399887085,
      "num_tokens": 80509897.0,
      "step": 36755
    },
    {
      "entropy": 5.3980677127838135,
      "epoch": 3.648273124255657,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037176744677483605,
      "loss": 5.0167,
      "mean_token_accuracy": 0.1812607765197754,
      "num_tokens": 80521883.0,
      "step": 36760
    },
    {
      "entropy": 5.351589059829712,
      "epoch": 3.6487693529178244,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037173545830563714,
      "loss": 4.891,
      "mean_token_accuracy": 0.18954287767410277,
      "num_tokens": 80532580.0,
      "step": 36765
    },
    {
      "entropy": 5.416403293609619,
      "epoch": 3.649265581579992,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003717034674376073,
      "loss": 5.0027,
      "mean_token_accuracy": 0.18079604357481002,
      "num_tokens": 80543163.0,
      "step": 36770
    },
    {
      "entropy": 5.382970809936523,
      "epoch": 3.6497618102421594,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037167147417153983,
      "loss": 5.0062,
      "mean_token_accuracy": 0.18424480855464936,
      "num_tokens": 80553721.0,
      "step": 36775
    },
    {
      "entropy": 5.444493246078491,
      "epoch": 3.650258038904327,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037163947850822816,
      "loss": 4.9985,
      "mean_token_accuracy": 0.18075242191553115,
      "num_tokens": 80564340.0,
      "step": 36780
    },
    {
      "entropy": 5.489942026138306,
      "epoch": 3.650754267566495,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037160748044846566,
      "loss": 5.0932,
      "mean_token_accuracy": 0.1800874575972557,
      "num_tokens": 80575669.0,
      "step": 36785
    },
    {
      "entropy": 5.403423547744751,
      "epoch": 3.6512504962286623,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003715754799930458,
      "loss": 4.9592,
      "mean_token_accuracy": 0.18571726828813553,
      "num_tokens": 80586946.0,
      "step": 36790
    },
    {
      "entropy": 5.349408292770386,
      "epoch": 3.6517467248908297,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003715434771427621,
      "loss": 4.9441,
      "mean_token_accuracy": 0.18552385568618773,
      "num_tokens": 80596876.0,
      "step": 36795
    },
    {
      "entropy": 5.374856853485108,
      "epoch": 3.652242953552997,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0003715114718984083,
      "loss": 4.9864,
      "mean_token_accuracy": 0.18125345408916474,
      "num_tokens": 80608243.0,
      "step": 36800
    },
    {
      "entropy": 5.372229814529419,
      "epoch": 3.6527391822151647,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037147946426077785,
      "loss": 4.9873,
      "mean_token_accuracy": 0.18219967931509018,
      "num_tokens": 80618514.0,
      "step": 36805
    },
    {
      "entropy": 5.41916708946228,
      "epoch": 3.653235410877332,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003714474542306646,
      "loss": 4.9742,
      "mean_token_accuracy": 0.17971839755773544,
      "num_tokens": 80628686.0,
      "step": 36810
    },
    {
      "entropy": 5.4034195899963375,
      "epoch": 3.6537316395394996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037141544180886233,
      "loss": 4.8959,
      "mean_token_accuracy": 0.1859565332531929,
      "num_tokens": 80639414.0,
      "step": 36815
    },
    {
      "entropy": 5.273688173294067,
      "epoch": 3.6542278682016676,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003713834269961649,
      "loss": 4.9001,
      "mean_token_accuracy": 0.19004531502723693,
      "num_tokens": 80652504.0,
      "step": 36820
    },
    {
      "entropy": 5.373208045959473,
      "epoch": 3.654724096863835,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037135140979336614,
      "loss": 5.0012,
      "mean_token_accuracy": 0.18111428320407869,
      "num_tokens": 80663923.0,
      "step": 36825
    },
    {
      "entropy": 5.4177680015563965,
      "epoch": 3.6552203255260025,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037131939020126,
      "loss": 4.9716,
      "mean_token_accuracy": 0.18380447775125502,
      "num_tokens": 80674150.0,
      "step": 36830
    },
    {
      "entropy": 5.4716269969940186,
      "epoch": 3.65571655418817,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003712873682206406,
      "loss": 5.0935,
      "mean_token_accuracy": 0.18037766963243484,
      "num_tokens": 80685963.0,
      "step": 36835
    },
    {
      "entropy": 5.364214563369751,
      "epoch": 3.6562127828503375,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003712553438523019,
      "loss": 4.9066,
      "mean_token_accuracy": 0.19544523805379868,
      "num_tokens": 80696838.0,
      "step": 36840
    },
    {
      "entropy": 5.392987632751465,
      "epoch": 3.656709011512505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037122331709703803,
      "loss": 5.0451,
      "mean_token_accuracy": 0.184031879901886,
      "num_tokens": 80708333.0,
      "step": 36845
    },
    {
      "entropy": 5.37519268989563,
      "epoch": 3.6572052401746724,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003711912879556433,
      "loss": 4.961,
      "mean_token_accuracy": 0.19123016744852067,
      "num_tokens": 80718875.0,
      "step": 36850
    },
    {
      "entropy": 5.464037656784058,
      "epoch": 3.65770146883684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000371159256428912,
      "loss": 5.0191,
      "mean_token_accuracy": 0.17863593995571136,
      "num_tokens": 80728637.0,
      "step": 36855
    },
    {
      "entropy": 5.4211163997650145,
      "epoch": 3.6581976974990074,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037112722251763817,
      "loss": 4.9756,
      "mean_token_accuracy": 0.17988142669200896,
      "num_tokens": 80740083.0,
      "step": 36860
    },
    {
      "entropy": 5.415176868438721,
      "epoch": 3.658693926161175,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003710951862226165,
      "loss": 5.0244,
      "mean_token_accuracy": 0.18127388507127762,
      "num_tokens": 80752412.0,
      "step": 36865
    },
    {
      "entropy": 5.339845991134643,
      "epoch": 3.6591901548233428,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037106314754464125,
      "loss": 4.9442,
      "mean_token_accuracy": 0.18406764268875123,
      "num_tokens": 80761653.0,
      "step": 36870
    },
    {
      "entropy": 5.310042858123779,
      "epoch": 3.6596863834855102,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003710311064845069,
      "loss": 4.886,
      "mean_token_accuracy": 0.19003617763519287,
      "num_tokens": 80771135.0,
      "step": 36875
    },
    {
      "entropy": 5.4068028926849365,
      "epoch": 3.6601826121476777,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000370999063043008,
      "loss": 4.9939,
      "mean_token_accuracy": 0.1785968393087387,
      "num_tokens": 80782119.0,
      "step": 36880
    },
    {
      "entropy": 5.411668586730957,
      "epoch": 3.660678840809845,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003709670172209393,
      "loss": 4.9784,
      "mean_token_accuracy": 0.18140937238931656,
      "num_tokens": 80793917.0,
      "step": 36885
    },
    {
      "entropy": 5.430563306808471,
      "epoch": 3.6611750694720127,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003709349690190952,
      "loss": 5.1103,
      "mean_token_accuracy": 0.1745972827076912,
      "num_tokens": 80807256.0,
      "step": 36890
    },
    {
      "entropy": 5.382863235473633,
      "epoch": 3.66167129813418,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003709029184382708,
      "loss": 4.9425,
      "mean_token_accuracy": 0.19162751436233522,
      "num_tokens": 80817224.0,
      "step": 36895
    },
    {
      "entropy": 5.375216484069824,
      "epoch": 3.6621675267963476,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003708708654792605,
      "loss": 4.9066,
      "mean_token_accuracy": 0.18750384896993638,
      "num_tokens": 80828692.0,
      "step": 36900
    },
    {
      "entropy": 5.393818426132202,
      "epoch": 3.6626637554585155,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037083881014285944,
      "loss": 4.9906,
      "mean_token_accuracy": 0.18401738852262498,
      "num_tokens": 80839201.0,
      "step": 36905
    },
    {
      "entropy": 5.464559030532837,
      "epoch": 3.663159984120683,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037080675242986234,
      "loss": 5.0634,
      "mean_token_accuracy": 0.1776882901787758,
      "num_tokens": 80853184.0,
      "step": 36910
    },
    {
      "entropy": 5.312836933135986,
      "epoch": 3.6636562127828505,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003707746923410642,
      "loss": 4.8534,
      "mean_token_accuracy": 0.1875860035419464,
      "num_tokens": 80862425.0,
      "step": 36915
    },
    {
      "entropy": 5.375034189224243,
      "epoch": 3.664152441445018,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037074262987726013,
      "loss": 4.9162,
      "mean_token_accuracy": 0.1894046723842621,
      "num_tokens": 80873652.0,
      "step": 36920
    },
    {
      "entropy": 5.380974626541137,
      "epoch": 3.6646486701071854,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003707105650392451,
      "loss": 4.97,
      "mean_token_accuracy": 0.1854327917098999,
      "num_tokens": 80884505.0,
      "step": 36925
    },
    {
      "entropy": 5.352917242050171,
      "epoch": 3.665144898769353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003706784978278144,
      "loss": 4.9613,
      "mean_token_accuracy": 0.18890138417482377,
      "num_tokens": 80895728.0,
      "step": 36930
    },
    {
      "entropy": 5.4320629119873045,
      "epoch": 3.6656411274315204,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003706464282437629,
      "loss": 4.9925,
      "mean_token_accuracy": 0.1798323154449463,
      "num_tokens": 80907054.0,
      "step": 36935
    },
    {
      "entropy": 5.412231349945069,
      "epoch": 3.666137356093688,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003706143562878863,
      "loss": 4.9389,
      "mean_token_accuracy": 0.19168450832366943,
      "num_tokens": 80917331.0,
      "step": 36940
    },
    {
      "entropy": 5.335580110549927,
      "epoch": 3.6666335847558553,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003705822819609796,
      "loss": 4.9336,
      "mean_token_accuracy": 0.18680620938539505,
      "num_tokens": 80929768.0,
      "step": 36945
    },
    {
      "entropy": 5.383995246887207,
      "epoch": 3.667129813418023,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003705502052638383,
      "loss": 4.983,
      "mean_token_accuracy": 0.18187232911586762,
      "num_tokens": 80940226.0,
      "step": 36950
    },
    {
      "entropy": 5.333879852294922,
      "epoch": 3.6676260420801907,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037051812619725785,
      "loss": 4.9608,
      "mean_token_accuracy": 0.1872262805700302,
      "num_tokens": 80950622.0,
      "step": 36955
    },
    {
      "entropy": 5.372058582305908,
      "epoch": 3.668122270742358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003704860447620335,
      "loss": 4.943,
      "mean_token_accuracy": 0.18901578187942505,
      "num_tokens": 80961439.0,
      "step": 36960
    },
    {
      "entropy": 5.4268828392028805,
      "epoch": 3.6686184994045257,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003704539609589612,
      "loss": 5.0543,
      "mean_token_accuracy": 0.17731310427188873,
      "num_tokens": 80972898.0,
      "step": 36965
    },
    {
      "entropy": 5.334868907928467,
      "epoch": 3.669114728066693,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003704218747888363,
      "loss": 4.9095,
      "mean_token_accuracy": 0.19657402187585832,
      "num_tokens": 80984758.0,
      "step": 36970
    },
    {
      "entropy": 5.393038082122803,
      "epoch": 3.6696109567288606,
      "grad_norm": 1.0,
      "learning_rate": 0.00037038978625245447,
      "loss": 4.964,
      "mean_token_accuracy": 0.18779475539922713,
      "num_tokens": 80995987.0,
      "step": 36975
    },
    {
      "entropy": 5.437883472442627,
      "epoch": 3.670107185391028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003703576953506115,
      "loss": 5.0329,
      "mean_token_accuracy": 0.17892197966575624,
      "num_tokens": 81006834.0,
      "step": 36980
    },
    {
      "entropy": 5.394991540908814,
      "epoch": 3.6706034140531956,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037032560208410313,
      "loss": 4.9643,
      "mean_token_accuracy": 0.18752804398536682,
      "num_tokens": 81016273.0,
      "step": 36985
    },
    {
      "entropy": 5.385824298858642,
      "epoch": 3.6710996427153635,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037029350645372526,
      "loss": 4.9656,
      "mean_token_accuracy": 0.1843208834528923,
      "num_tokens": 81026292.0,
      "step": 36990
    },
    {
      "entropy": 5.349711751937866,
      "epoch": 3.671595871377531,
      "grad_norm": 1.0,
      "learning_rate": 0.00037026140846027383,
      "loss": 4.9733,
      "mean_token_accuracy": 0.1832132264971733,
      "num_tokens": 81036824.0,
      "step": 36995
    },
    {
      "entropy": 5.366300344467163,
      "epoch": 3.6720921000396984,
      "grad_norm": 1.0,
      "learning_rate": 0.0003702293081045448,
      "loss": 4.9695,
      "mean_token_accuracy": 0.1891883134841919,
      "num_tokens": 81047379.0,
      "step": 37000
    },
    {
      "entropy": 5.469265556335449,
      "epoch": 3.672588328701866,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037019720538733394,
      "loss": 5.0862,
      "mean_token_accuracy": 0.17705198675394057,
      "num_tokens": 81059097.0,
      "step": 37005
    },
    {
      "entropy": 5.425981330871582,
      "epoch": 3.6730845573640334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003701651003094376,
      "loss": 4.9832,
      "mean_token_accuracy": 0.1837664872407913,
      "num_tokens": 81069747.0,
      "step": 37010
    },
    {
      "entropy": 5.439412021636963,
      "epoch": 3.673580786026201,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003701329928716519,
      "loss": 5.0234,
      "mean_token_accuracy": 0.18342453986406326,
      "num_tokens": 81080947.0,
      "step": 37015
    },
    {
      "entropy": 5.438124752044677,
      "epoch": 3.6740770146883683,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037010088307477286,
      "loss": 4.9756,
      "mean_token_accuracy": 0.19492780268192292,
      "num_tokens": 81091736.0,
      "step": 37020
    },
    {
      "entropy": 5.384000682830811,
      "epoch": 3.674573243350536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003700687709195969,
      "loss": 4.9693,
      "mean_token_accuracy": 0.18928191661834717,
      "num_tokens": 81103289.0,
      "step": 37025
    },
    {
      "entropy": 5.3875738143920895,
      "epoch": 3.6750694720127033,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003700366564069203,
      "loss": 4.9396,
      "mean_token_accuracy": 0.18942797034978867,
      "num_tokens": 81116738.0,
      "step": 37030
    },
    {
      "entropy": 5.344467544555664,
      "epoch": 3.6755657006748708,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037000453953753943,
      "loss": 4.9267,
      "mean_token_accuracy": 0.19139126390218736,
      "num_tokens": 81126642.0,
      "step": 37035
    },
    {
      "entropy": 5.4456977367401125,
      "epoch": 3.6760619293370382,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003699724203122506,
      "loss": 5.0567,
      "mean_token_accuracy": 0.1825786277651787,
      "num_tokens": 81137744.0,
      "step": 37040
    },
    {
      "entropy": 5.318710136413574,
      "epoch": 3.676558157999206,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003699402987318504,
      "loss": 4.8443,
      "mean_token_accuracy": 0.19422439783811568,
      "num_tokens": 81149420.0,
      "step": 37045
    },
    {
      "entropy": 5.424037313461303,
      "epoch": 3.6770543866613736,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036990817479713543,
      "loss": 5.0444,
      "mean_token_accuracy": 0.180436609685421,
      "num_tokens": 81158516.0,
      "step": 37050
    },
    {
      "entropy": 5.3392292022705075,
      "epoch": 3.677550615323541,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036987604850890224,
      "loss": 4.9553,
      "mean_token_accuracy": 0.1828247144818306,
      "num_tokens": 81170230.0,
      "step": 37055
    },
    {
      "entropy": 5.3591419696807865,
      "epoch": 3.6780468439857086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036984391986794745,
      "loss": 4.9055,
      "mean_token_accuracy": 0.1891096606850624,
      "num_tokens": 81180841.0,
      "step": 37060
    },
    {
      "entropy": 5.4082128524780275,
      "epoch": 3.678543072647876,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036981178887506776,
      "loss": 5.0172,
      "mean_token_accuracy": 0.1833957925438881,
      "num_tokens": 81191808.0,
      "step": 37065
    },
    {
      "entropy": 5.400483703613281,
      "epoch": 3.6790393013100435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036977965553106014,
      "loss": 4.9417,
      "mean_token_accuracy": 0.18900518864393234,
      "num_tokens": 81201326.0,
      "step": 37070
    },
    {
      "entropy": 5.387190246582032,
      "epoch": 3.6795355299722114,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003697475198367212,
      "loss": 4.9327,
      "mean_token_accuracy": 0.18842213302850724,
      "num_tokens": 81213458.0,
      "step": 37075
    },
    {
      "entropy": 5.429676151275634,
      "epoch": 3.680031758634379,
      "grad_norm": 0.921875,
      "learning_rate": 0.000369715381792848,
      "loss": 5.0475,
      "mean_token_accuracy": 0.17841103821992874,
      "num_tokens": 81225133.0,
      "step": 37080
    },
    {
      "entropy": 5.3928327560424805,
      "epoch": 3.6805279872965464,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003696832414002374,
      "loss": 5.0462,
      "mean_token_accuracy": 0.17756138145923614,
      "num_tokens": 81237592.0,
      "step": 37085
    },
    {
      "entropy": 5.449064779281616,
      "epoch": 3.681024215958714,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003696510986596864,
      "loss": 5.0088,
      "mean_token_accuracy": 0.1870705932378769,
      "num_tokens": 81247527.0,
      "step": 37090
    },
    {
      "entropy": 5.444961500167847,
      "epoch": 3.6815204446208813,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036961895357199214,
      "loss": 5.0991,
      "mean_token_accuracy": 0.17863359302282333,
      "num_tokens": 81259109.0,
      "step": 37095
    },
    {
      "entropy": 5.3849118709564205,
      "epoch": 3.682016673283049,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003695868061379518,
      "loss": 4.9702,
      "mean_token_accuracy": 0.1872912660241127,
      "num_tokens": 81269747.0,
      "step": 37100
    },
    {
      "entropy": 5.445422506332397,
      "epoch": 3.6825129019452163,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003695546563583623,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18231143206357955,
      "num_tokens": 81279332.0,
      "step": 37105
    },
    {
      "entropy": 5.356952571868897,
      "epoch": 3.6830091306073838,
      "grad_norm": 1.5,
      "learning_rate": 0.0003695225042340212,
      "loss": 4.8959,
      "mean_token_accuracy": 0.19678807109594346,
      "num_tokens": 81290274.0,
      "step": 37110
    },
    {
      "entropy": 5.442086458206177,
      "epoch": 3.6835053592695512,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003694903497657256,
      "loss": 4.9997,
      "mean_token_accuracy": 0.18240921646356584,
      "num_tokens": 81301665.0,
      "step": 37115
    },
    {
      "entropy": 5.393161201477051,
      "epoch": 3.6840015879317187,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000369458192954273,
      "loss": 4.9445,
      "mean_token_accuracy": 0.1825572669506073,
      "num_tokens": 81311581.0,
      "step": 37120
    },
    {
      "entropy": 5.412784242630005,
      "epoch": 3.684497816593886,
      "grad_norm": 1.125,
      "learning_rate": 0.0003694260338004607,
      "loss": 5.0408,
      "mean_token_accuracy": 0.1835364207625389,
      "num_tokens": 81322526.0,
      "step": 37125
    },
    {
      "entropy": 5.354240083694458,
      "epoch": 3.684994045256054,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036939387230508627,
      "loss": 4.9178,
      "mean_token_accuracy": 0.18866579681634904,
      "num_tokens": 81332132.0,
      "step": 37130
    },
    {
      "entropy": 5.384891510009766,
      "epoch": 3.6854902739182216,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036936170846894715,
      "loss": 4.9571,
      "mean_token_accuracy": 0.18545924872159958,
      "num_tokens": 81343425.0,
      "step": 37135
    },
    {
      "entropy": 5.380729150772095,
      "epoch": 3.685986502580389,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036932954229284106,
      "loss": 4.9303,
      "mean_token_accuracy": 0.19109421819448472,
      "num_tokens": 81354262.0,
      "step": 37140
    },
    {
      "entropy": 5.384098911285401,
      "epoch": 3.6864827312425565,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036929737377756563,
      "loss": 4.9378,
      "mean_token_accuracy": 0.18224654346704483,
      "num_tokens": 81365324.0,
      "step": 37145
    },
    {
      "entropy": 5.321290779113769,
      "epoch": 3.686978959904724,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003692652029239184,
      "loss": 4.9262,
      "mean_token_accuracy": 0.19153830260038376,
      "num_tokens": 81376083.0,
      "step": 37150
    },
    {
      "entropy": 5.471001529693604,
      "epoch": 3.6874751885668915,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036923302973269727,
      "loss": 4.9921,
      "mean_token_accuracy": 0.17808472663164138,
      "num_tokens": 81386283.0,
      "step": 37155
    },
    {
      "entropy": 5.318043565750122,
      "epoch": 3.6879714172290594,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003692008542047001,
      "loss": 4.9039,
      "mean_token_accuracy": 0.1891833558678627,
      "num_tokens": 81397878.0,
      "step": 37160
    },
    {
      "entropy": 5.415350341796875,
      "epoch": 3.688467645891227,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003691686763407248,
      "loss": 5.0505,
      "mean_token_accuracy": 0.18425189107656478,
      "num_tokens": 81409464.0,
      "step": 37165
    },
    {
      "entropy": 5.466192817687988,
      "epoch": 3.6889638745533944,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003691364961415691,
      "loss": 5.0983,
      "mean_token_accuracy": 0.18161174952983855,
      "num_tokens": 81421352.0,
      "step": 37170
    },
    {
      "entropy": 5.360564184188843,
      "epoch": 3.689460103215562,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036910431360803124,
      "loss": 4.8414,
      "mean_token_accuracy": 0.19360606372356415,
      "num_tokens": 81431045.0,
      "step": 37175
    },
    {
      "entropy": 5.3450319290161135,
      "epoch": 3.6899563318777293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036907212874090923,
      "loss": 4.8739,
      "mean_token_accuracy": 0.1968681365251541,
      "num_tokens": 81441772.0,
      "step": 37180
    },
    {
      "entropy": 5.268412780761719,
      "epoch": 3.690452560539897,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000369039941541001,
      "loss": 4.8877,
      "mean_token_accuracy": 0.18960559815168382,
      "num_tokens": 81451964.0,
      "step": 37185
    },
    {
      "entropy": 5.328274440765381,
      "epoch": 3.6909487892020643,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036900775200910497,
      "loss": 4.8867,
      "mean_token_accuracy": 0.18984704464673996,
      "num_tokens": 81462598.0,
      "step": 37190
    },
    {
      "entropy": 5.419655084609985,
      "epoch": 3.6914450178642317,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036897556014601926,
      "loss": 4.9946,
      "mean_token_accuracy": 0.18253720700740814,
      "num_tokens": 81474415.0,
      "step": 37195
    },
    {
      "entropy": 5.432569408416748,
      "epoch": 3.691941246526399,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036894336595254203,
      "loss": 4.981,
      "mean_token_accuracy": 0.18517823666334152,
      "num_tokens": 81485510.0,
      "step": 37200
    },
    {
      "entropy": 5.409546899795532,
      "epoch": 3.6924374751885667,
      "grad_norm": 1.0,
      "learning_rate": 0.00036891116942947184,
      "loss": 5.0093,
      "mean_token_accuracy": 0.1793094739317894,
      "num_tokens": 81496103.0,
      "step": 37205
    },
    {
      "entropy": 5.400683498382568,
      "epoch": 3.692933703850734,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000368878970577607,
      "loss": 5.0546,
      "mean_token_accuracy": 0.1838851109147072,
      "num_tokens": 81507876.0,
      "step": 37210
    },
    {
      "entropy": 5.37025785446167,
      "epoch": 3.693429932512902,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036884676939774597,
      "loss": 4.9439,
      "mean_token_accuracy": 0.18683214783668517,
      "num_tokens": 81520615.0,
      "step": 37215
    },
    {
      "entropy": 5.342367219924927,
      "epoch": 3.6939261611750696,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036881456589068725,
      "loss": 4.9298,
      "mean_token_accuracy": 0.18653721064329148,
      "num_tokens": 81531614.0,
      "step": 37220
    },
    {
      "entropy": 5.3818670272827145,
      "epoch": 3.694422389837237,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036878236005722954,
      "loss": 4.9267,
      "mean_token_accuracy": 0.18761082142591476,
      "num_tokens": 81541027.0,
      "step": 37225
    },
    {
      "entropy": 5.356039524078369,
      "epoch": 3.6949186184994045,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003687501518981713,
      "loss": 4.9692,
      "mean_token_accuracy": 0.18124381750822066,
      "num_tokens": 81551569.0,
      "step": 37230
    },
    {
      "entropy": 5.362717962265014,
      "epoch": 3.695414847161572,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003687179414143114,
      "loss": 4.9286,
      "mean_token_accuracy": 0.1889091059565544,
      "num_tokens": 81561490.0,
      "step": 37235
    },
    {
      "entropy": 5.331531715393067,
      "epoch": 3.6959110758237395,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036868572860644845,
      "loss": 4.8611,
      "mean_token_accuracy": 0.18749651163816453,
      "num_tokens": 81572434.0,
      "step": 37240
    },
    {
      "entropy": 5.395431089401245,
      "epoch": 3.6964073044859074,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036865351347538123,
      "loss": 4.9742,
      "mean_token_accuracy": 0.17753081321716307,
      "num_tokens": 81585135.0,
      "step": 37245
    },
    {
      "entropy": 5.31577434539795,
      "epoch": 3.696903533148075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003686212960219088,
      "loss": 4.8722,
      "mean_token_accuracy": 0.19691140502691268,
      "num_tokens": 81596919.0,
      "step": 37250
    },
    {
      "entropy": 5.338291978836059,
      "epoch": 3.6973997618102423,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036858907624682987,
      "loss": 4.9603,
      "mean_token_accuracy": 0.1833590880036354,
      "num_tokens": 81608335.0,
      "step": 37255
    },
    {
      "entropy": 5.392196607589722,
      "epoch": 3.69789599047241,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003685568541509435,
      "loss": 4.9546,
      "mean_token_accuracy": 0.18608352392911912,
      "num_tokens": 81618940.0,
      "step": 37260
    },
    {
      "entropy": 5.383130025863648,
      "epoch": 3.6983922191345773,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036852462973504884,
      "loss": 4.9475,
      "mean_token_accuracy": 0.1839813619852066,
      "num_tokens": 81629215.0,
      "step": 37265
    },
    {
      "entropy": 5.399757814407349,
      "epoch": 3.6988884477967447,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003684924029999449,
      "loss": 5.0529,
      "mean_token_accuracy": 0.18394280076026917,
      "num_tokens": 81642201.0,
      "step": 37270
    },
    {
      "entropy": 5.376278018951416,
      "epoch": 3.6993846764589122,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036846017394643064,
      "loss": 4.8725,
      "mean_token_accuracy": 0.19436758309602736,
      "num_tokens": 81652307.0,
      "step": 37275
    },
    {
      "entropy": 5.395110130310059,
      "epoch": 3.6998809051210797,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003684279425753057,
      "loss": 4.9924,
      "mean_token_accuracy": 0.1850610613822937,
      "num_tokens": 81662904.0,
      "step": 37280
    },
    {
      "entropy": 5.440134000778198,
      "epoch": 3.700377133783247,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036839570888736893,
      "loss": 5.0517,
      "mean_token_accuracy": 0.18450997918844222,
      "num_tokens": 81674456.0,
      "step": 37285
    },
    {
      "entropy": 5.426989221572876,
      "epoch": 3.7008733624454146,
      "grad_norm": 1.0,
      "learning_rate": 0.0003683634728834199,
      "loss": 4.9171,
      "mean_token_accuracy": 0.18469078689813614,
      "num_tokens": 81685128.0,
      "step": 37290
    },
    {
      "entropy": 5.349720144271851,
      "epoch": 3.701369591107582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036833123456425794,
      "loss": 4.9856,
      "mean_token_accuracy": 0.18747872710227967,
      "num_tokens": 81696661.0,
      "step": 37295
    },
    {
      "entropy": 5.4392900466918945,
      "epoch": 3.70186581976975,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003682989939306824,
      "loss": 5.0187,
      "mean_token_accuracy": 0.18475190699100494,
      "num_tokens": 81708095.0,
      "step": 37300
    },
    {
      "entropy": 5.460628414154053,
      "epoch": 3.7023620484319175,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003682667509834929,
      "loss": 5.0767,
      "mean_token_accuracy": 0.17737533450126647,
      "num_tokens": 81719798.0,
      "step": 37305
    },
    {
      "entropy": 5.347486782073974,
      "epoch": 3.702858277094085,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036823450572348896,
      "loss": 4.9013,
      "mean_token_accuracy": 0.19052979350090027,
      "num_tokens": 81731041.0,
      "step": 37310
    },
    {
      "entropy": 5.382733011245728,
      "epoch": 3.7033545057562525,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003682022581514702,
      "loss": 4.9646,
      "mean_token_accuracy": 0.18469977229833603,
      "num_tokens": 81742198.0,
      "step": 37315
    },
    {
      "entropy": 5.380887508392334,
      "epoch": 3.70385073441842,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036817000826823616,
      "loss": 4.908,
      "mean_token_accuracy": 0.18720412254333496,
      "num_tokens": 81752824.0,
      "step": 37320
    },
    {
      "entropy": 5.390272903442383,
      "epoch": 3.7043469630805874,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036813775607458677,
      "loss": 4.9464,
      "mean_token_accuracy": 0.18975413888692855,
      "num_tokens": 81764785.0,
      "step": 37325
    },
    {
      "entropy": 5.320296144485473,
      "epoch": 3.7048431917427553,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003681055015713217,
      "loss": 4.8661,
      "mean_token_accuracy": 0.20059096515178682,
      "num_tokens": 81776755.0,
      "step": 37330
    },
    {
      "entropy": 5.41259536743164,
      "epoch": 3.705339420404923,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003680732447592408,
      "loss": 4.9394,
      "mean_token_accuracy": 0.1882055416703224,
      "num_tokens": 81788088.0,
      "step": 37335
    },
    {
      "entropy": 5.346343946456909,
      "epoch": 3.7058356490670903,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036804098563914407,
      "loss": 5.0174,
      "mean_token_accuracy": 0.18598006665706635,
      "num_tokens": 81800114.0,
      "step": 37340
    },
    {
      "entropy": 5.382785940170288,
      "epoch": 3.7063318777292578,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003680087242118313,
      "loss": 4.9583,
      "mean_token_accuracy": 0.18436837792396546,
      "num_tokens": 81811309.0,
      "step": 37345
    },
    {
      "entropy": 5.4056027889251705,
      "epoch": 3.7068281063914252,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003679764604781026,
      "loss": 5.0195,
      "mean_token_accuracy": 0.1821539431810379,
      "num_tokens": 81821570.0,
      "step": 37350
    },
    {
      "entropy": 5.3976158618927,
      "epoch": 3.7073243350535927,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036794419443875805,
      "loss": 4.904,
      "mean_token_accuracy": 0.19326714426279068,
      "num_tokens": 81832674.0,
      "step": 37355
    },
    {
      "entropy": 5.409243249893189,
      "epoch": 3.70782056371576,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003679119260945977,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1796821117401123,
      "num_tokens": 81843463.0,
      "step": 37360
    },
    {
      "entropy": 5.371316242218017,
      "epoch": 3.7083167923779277,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003678796554464219,
      "loss": 4.9725,
      "mean_token_accuracy": 0.1932612404227257,
      "num_tokens": 81853512.0,
      "step": 37365
    },
    {
      "entropy": 5.4203479290008545,
      "epoch": 3.708813021040095,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036784738249503063,
      "loss": 5.0594,
      "mean_token_accuracy": 0.17642046511173248,
      "num_tokens": 81864703.0,
      "step": 37370
    },
    {
      "entropy": 5.300574207305909,
      "epoch": 3.7093092497022626,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003678151072412245,
      "loss": 4.8895,
      "mean_token_accuracy": 0.18994342386722565,
      "num_tokens": 81875387.0,
      "step": 37375
    },
    {
      "entropy": 5.348829936981201,
      "epoch": 3.70980547836443,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036778282968580355,
      "loss": 4.9029,
      "mean_token_accuracy": 0.1926938086748123,
      "num_tokens": 81885908.0,
      "step": 37380
    },
    {
      "entropy": 5.453210926055908,
      "epoch": 3.710301707026598,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003677505498295685,
      "loss": 4.972,
      "mean_token_accuracy": 0.18858874589204788,
      "num_tokens": 81898256.0,
      "step": 37385
    },
    {
      "entropy": 5.284591102600098,
      "epoch": 3.7107979356887655,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036771826767331955,
      "loss": 4.876,
      "mean_token_accuracy": 0.19636707603931428,
      "num_tokens": 81908993.0,
      "step": 37390
    },
    {
      "entropy": 5.302511215209961,
      "epoch": 3.711294164350933,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003676859832178575,
      "loss": 4.9304,
      "mean_token_accuracy": 0.1880905494093895,
      "num_tokens": 81919189.0,
      "step": 37395
    },
    {
      "entropy": 5.401530790328979,
      "epoch": 3.7117903930131004,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003676536964639827,
      "loss": 4.9172,
      "mean_token_accuracy": 0.18800496011972428,
      "num_tokens": 81929213.0,
      "step": 37400
    },
    {
      "entropy": 5.387616300582886,
      "epoch": 3.712286621675268,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003676214074124958,
      "loss": 4.9676,
      "mean_token_accuracy": 0.18315909653902054,
      "num_tokens": 81940735.0,
      "step": 37405
    },
    {
      "entropy": 5.316745948791504,
      "epoch": 3.7127828503374354,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036758911606419773,
      "loss": 4.8988,
      "mean_token_accuracy": 0.19003281593322754,
      "num_tokens": 81951962.0,
      "step": 37410
    },
    {
      "entropy": 5.403576517105103,
      "epoch": 3.713279078999603,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003675568224198889,
      "loss": 4.9166,
      "mean_token_accuracy": 0.19012386202812195,
      "num_tokens": 81963328.0,
      "step": 37415
    },
    {
      "entropy": 5.28456678390503,
      "epoch": 3.7137753076617708,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003675245264803705,
      "loss": 4.9341,
      "mean_token_accuracy": 0.18630936443805696,
      "num_tokens": 81975318.0,
      "step": 37420
    },
    {
      "entropy": 5.348439931869507,
      "epoch": 3.7142715363239382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003674922282464432,
      "loss": 4.9742,
      "mean_token_accuracy": 0.18871117234230042,
      "num_tokens": 81986036.0,
      "step": 37425
    },
    {
      "entropy": 5.489771986007691,
      "epoch": 3.7147677649861057,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003674599277189078,
      "loss": 5.0783,
      "mean_token_accuracy": 0.1771955206990242,
      "num_tokens": 81998709.0,
      "step": 37430
    },
    {
      "entropy": 5.45481858253479,
      "epoch": 3.715263993648273,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036742762489856554,
      "loss": 5.0363,
      "mean_token_accuracy": 0.17788662016391754,
      "num_tokens": 82010436.0,
      "step": 37435
    },
    {
      "entropy": 5.403902673721314,
      "epoch": 3.7157602223104407,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003673953197862173,
      "loss": 5.0028,
      "mean_token_accuracy": 0.1805103212594986,
      "num_tokens": 82022194.0,
      "step": 37440
    },
    {
      "entropy": 5.349641942977906,
      "epoch": 3.716256450972608,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036736301238266434,
      "loss": 4.9762,
      "mean_token_accuracy": 0.18735718876123428,
      "num_tokens": 82033105.0,
      "step": 37445
    },
    {
      "entropy": 5.3378438472747805,
      "epoch": 3.7167526796347756,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036733070268870754,
      "loss": 4.912,
      "mean_token_accuracy": 0.18129592537879943,
      "num_tokens": 82043161.0,
      "step": 37450
    },
    {
      "entropy": 5.316860055923462,
      "epoch": 3.717248908296943,
      "grad_norm": 1.0,
      "learning_rate": 0.0003672983907051483,
      "loss": 4.8882,
      "mean_token_accuracy": 0.192394258081913,
      "num_tokens": 82053889.0,
      "step": 37455
    },
    {
      "entropy": 5.39522123336792,
      "epoch": 3.7177451369591106,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036726607643278793,
      "loss": 4.9661,
      "mean_token_accuracy": 0.1829373061656952,
      "num_tokens": 82066230.0,
      "step": 37460
    },
    {
      "entropy": 5.393646478652954,
      "epoch": 3.718241365621278,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036723375987242763,
      "loss": 4.9388,
      "mean_token_accuracy": 0.18573978096246718,
      "num_tokens": 82077655.0,
      "step": 37465
    },
    {
      "entropy": 5.441417646408081,
      "epoch": 3.718737594283446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036720144102486886,
      "loss": 5.0051,
      "mean_token_accuracy": 0.18367393612861632,
      "num_tokens": 82089765.0,
      "step": 37470
    },
    {
      "entropy": 5.496977615356445,
      "epoch": 3.7192338229456134,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036716911989091304,
      "loss": 5.0723,
      "mean_token_accuracy": 0.17393119186162947,
      "num_tokens": 82101095.0,
      "step": 37475
    },
    {
      "entropy": 5.31279616355896,
      "epoch": 3.719730051607781,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036713679647136163,
      "loss": 4.8736,
      "mean_token_accuracy": 0.1898948773741722,
      "num_tokens": 82111741.0,
      "step": 37480
    },
    {
      "entropy": 5.364109849929809,
      "epoch": 3.7202262802699484,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003671044707670161,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18132892549037932,
      "num_tokens": 82122749.0,
      "step": 37485
    },
    {
      "entropy": 5.383986902236939,
      "epoch": 3.720722508932116,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036707214277867834,
      "loss": 4.9063,
      "mean_token_accuracy": 0.18613419383764268,
      "num_tokens": 82132032.0,
      "step": 37490
    },
    {
      "entropy": 5.36259970664978,
      "epoch": 3.7212187375942833,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003670398125071497,
      "loss": 4.9882,
      "mean_token_accuracy": 0.18190679252147673,
      "num_tokens": 82143169.0,
      "step": 37495
    },
    {
      "entropy": 5.446428918838501,
      "epoch": 3.721714966256451,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036700747995323214,
      "loss": 5.0302,
      "mean_token_accuracy": 0.1824725091457367,
      "num_tokens": 82154412.0,
      "step": 37500
    },
    {
      "entropy": 5.338373041152954,
      "epoch": 3.7222111949186187,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003669751451177273,
      "loss": 4.9409,
      "mean_token_accuracy": 0.19146450161933898,
      "num_tokens": 82165473.0,
      "step": 37505
    },
    {
      "entropy": 5.38622260093689,
      "epoch": 3.722707423580786,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000366942808001437,
      "loss": 4.969,
      "mean_token_accuracy": 0.1843883588910103,
      "num_tokens": 82175716.0,
      "step": 37510
    },
    {
      "entropy": 5.372196245193481,
      "epoch": 3.7232036522429537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036691046860516327,
      "loss": 5.0024,
      "mean_token_accuracy": 0.18265339434146882,
      "num_tokens": 82185981.0,
      "step": 37515
    },
    {
      "entropy": 5.409123516082763,
      "epoch": 3.723699880905121,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036687812692970796,
      "loss": 5.0057,
      "mean_token_accuracy": 0.1773834377527237,
      "num_tokens": 82197336.0,
      "step": 37520
    },
    {
      "entropy": 5.407315301895141,
      "epoch": 3.7241961095672886,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000366845782975873,
      "loss": 4.9922,
      "mean_token_accuracy": 0.1876524046063423,
      "num_tokens": 82207895.0,
      "step": 37525
    },
    {
      "entropy": 5.35406723022461,
      "epoch": 3.724692338229456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036681343674446054,
      "loss": 4.9046,
      "mean_token_accuracy": 0.19415030181407927,
      "num_tokens": 82218618.0,
      "step": 37530
    },
    {
      "entropy": 5.348025941848755,
      "epoch": 3.7251885668916236,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003667810882362728,
      "loss": 4.9598,
      "mean_token_accuracy": 0.18125088661909103,
      "num_tokens": 82229803.0,
      "step": 37535
    },
    {
      "entropy": 5.330456495285034,
      "epoch": 3.725684795553791,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003667487374521117,
      "loss": 4.9313,
      "mean_token_accuracy": 0.18500450104475022,
      "num_tokens": 82239448.0,
      "step": 37540
    },
    {
      "entropy": 5.369283962249756,
      "epoch": 3.7261810242159585,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003667163843927797,
      "loss": 4.9766,
      "mean_token_accuracy": 0.18538911044597625,
      "num_tokens": 82250093.0,
      "step": 37545
    },
    {
      "entropy": 5.332195472717285,
      "epoch": 3.726677252878126,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036668402905907897,
      "loss": 4.9367,
      "mean_token_accuracy": 0.18216508030891418,
      "num_tokens": 82260479.0,
      "step": 37550
    },
    {
      "entropy": 5.381354856491089,
      "epoch": 3.727173481540294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036665167145181186,
      "loss": 5.0187,
      "mean_token_accuracy": 0.18341847956180574,
      "num_tokens": 82272482.0,
      "step": 37555
    },
    {
      "entropy": 5.379727220535278,
      "epoch": 3.7276697102024614,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036661931157178085,
      "loss": 4.9481,
      "mean_token_accuracy": 0.18318957090377808,
      "num_tokens": 82282456.0,
      "step": 37560
    },
    {
      "entropy": 5.482713317871093,
      "epoch": 3.728165938864629,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036658694941978825,
      "loss": 5.0411,
      "mean_token_accuracy": 0.18742189407348633,
      "num_tokens": 82292832.0,
      "step": 37565
    },
    {
      "entropy": 5.391187238693237,
      "epoch": 3.7286621675267964,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036655458499663675,
      "loss": 4.9457,
      "mean_token_accuracy": 0.1842226728796959,
      "num_tokens": 82302841.0,
      "step": 37570
    },
    {
      "entropy": 5.361298942565918,
      "epoch": 3.729158396188964,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003665222183031288,
      "loss": 4.9666,
      "mean_token_accuracy": 0.184636427462101,
      "num_tokens": 82314092.0,
      "step": 37575
    },
    {
      "entropy": 5.452130889892578,
      "epoch": 3.7296546248511313,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000366489849340067,
      "loss": 5.073,
      "mean_token_accuracy": 0.17501142472028733,
      "num_tokens": 82325153.0,
      "step": 37580
    },
    {
      "entropy": 5.478820085525513,
      "epoch": 3.730150853513299,
      "grad_norm": 1.0,
      "learning_rate": 0.00036645747810825415,
      "loss": 4.9998,
      "mean_token_accuracy": 0.18275833278894424,
      "num_tokens": 82336463.0,
      "step": 37585
    },
    {
      "entropy": 5.405705547332763,
      "epoch": 3.7306470821754667,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036642510460849285,
      "loss": 4.9623,
      "mean_token_accuracy": 0.18969677984714509,
      "num_tokens": 82347314.0,
      "step": 37590
    },
    {
      "entropy": 5.3953594207763675,
      "epoch": 3.731143310837634,
      "grad_norm": 1.109375,
      "learning_rate": 0.000366392728841586,
      "loss": 4.9851,
      "mean_token_accuracy": 0.18892952054738998,
      "num_tokens": 82358582.0,
      "step": 37595
    },
    {
      "entropy": 5.402613353729248,
      "epoch": 3.7316395394998016,
      "grad_norm": 1.0,
      "learning_rate": 0.0003663603508083364,
      "loss": 4.9999,
      "mean_token_accuracy": 0.1812979757785797,
      "num_tokens": 82369132.0,
      "step": 37600
    },
    {
      "entropy": 5.403618478775025,
      "epoch": 3.732135768161969,
      "grad_norm": 1.0625,
      "learning_rate": 0.000366327970509547,
      "loss": 5.0362,
      "mean_token_accuracy": 0.18143713623285293,
      "num_tokens": 82380011.0,
      "step": 37605
    },
    {
      "entropy": 5.384805727005005,
      "epoch": 3.7326319968241366,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036629558794602063,
      "loss": 5.0344,
      "mean_token_accuracy": 0.1845942720770836,
      "num_tokens": 82391113.0,
      "step": 37610
    },
    {
      "entropy": 5.411979866027832,
      "epoch": 3.733128225486304,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036626320311856046,
      "loss": 4.9677,
      "mean_token_accuracy": 0.1826810657978058,
      "num_tokens": 82402780.0,
      "step": 37615
    },
    {
      "entropy": 5.357946538925171,
      "epoch": 3.7336244541484715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036623081602796953,
      "loss": 4.9665,
      "mean_token_accuracy": 0.1813036546111107,
      "num_tokens": 82412005.0,
      "step": 37620
    },
    {
      "entropy": 5.402330636978149,
      "epoch": 3.734120682810639,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036619842667505096,
      "loss": 4.9353,
      "mean_token_accuracy": 0.1931862100958824,
      "num_tokens": 82420868.0,
      "step": 37625
    },
    {
      "entropy": 5.459146785736084,
      "epoch": 3.7346169114728065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036616603506060786,
      "loss": 4.9519,
      "mean_token_accuracy": 0.18084315359592437,
      "num_tokens": 82430397.0,
      "step": 37630
    },
    {
      "entropy": 5.407284545898437,
      "epoch": 3.735113140134974,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003661336411854436,
      "loss": 4.9466,
      "mean_token_accuracy": 0.19108105897903443,
      "num_tokens": 82440567.0,
      "step": 37635
    },
    {
      "entropy": 5.41730489730835,
      "epoch": 3.7356093687971414,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003661012450503613,
      "loss": 4.9601,
      "mean_token_accuracy": 0.18293903023004532,
      "num_tokens": 82450920.0,
      "step": 37640
    },
    {
      "entropy": 5.3469236373901365,
      "epoch": 3.7361055974593094,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036606884665616455,
      "loss": 4.9715,
      "mean_token_accuracy": 0.1870715320110321,
      "num_tokens": 82461741.0,
      "step": 37645
    },
    {
      "entropy": 5.383629560470581,
      "epoch": 3.736601826121477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003660364460036566,
      "loss": 4.9813,
      "mean_token_accuracy": 0.1828901872038841,
      "num_tokens": 82472360.0,
      "step": 37650
    },
    {
      "entropy": 5.301816606521607,
      "epoch": 3.7370980547836443,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003660040430936409,
      "loss": 4.9172,
      "mean_token_accuracy": 0.18494182378053664,
      "num_tokens": 82482758.0,
      "step": 37655
    },
    {
      "entropy": 5.350799798965454,
      "epoch": 3.737594283445812,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036597163792692107,
      "loss": 4.883,
      "mean_token_accuracy": 0.18817923069000245,
      "num_tokens": 82494059.0,
      "step": 37660
    },
    {
      "entropy": 5.409857940673828,
      "epoch": 3.7380905121079793,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036593923050430067,
      "loss": 5.046,
      "mean_token_accuracy": 0.18027021288871764,
      "num_tokens": 82505503.0,
      "step": 37665
    },
    {
      "entropy": 5.454179143905639,
      "epoch": 3.7385867407701467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036590682082658334,
      "loss": 5.0722,
      "mean_token_accuracy": 0.1792944237589836,
      "num_tokens": 82516669.0,
      "step": 37670
    },
    {
      "entropy": 5.394580459594726,
      "epoch": 3.7390829694323147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036587440889457274,
      "loss": 4.9251,
      "mean_token_accuracy": 0.18172682225704193,
      "num_tokens": 82526717.0,
      "step": 37675
    },
    {
      "entropy": 5.427966117858887,
      "epoch": 3.739579198094482,
      "grad_norm": 1.0,
      "learning_rate": 0.00036584199470907254,
      "loss": 5.0251,
      "mean_token_accuracy": 0.1781385973095894,
      "num_tokens": 82536617.0,
      "step": 37680
    },
    {
      "entropy": 5.4289182186126705,
      "epoch": 3.7400754267566496,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003658095782708866,
      "loss": 5.0292,
      "mean_token_accuracy": 0.17678293734788894,
      "num_tokens": 82547858.0,
      "step": 37685
    },
    {
      "entropy": 5.378513813018799,
      "epoch": 3.740571655418817,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036577715958081885,
      "loss": 4.9748,
      "mean_token_accuracy": 0.1819063365459442,
      "num_tokens": 82559334.0,
      "step": 37690
    },
    {
      "entropy": 5.395077037811279,
      "epoch": 3.7410678840809846,
      "grad_norm": 1.0,
      "learning_rate": 0.0003657447386396732,
      "loss": 4.9548,
      "mean_token_accuracy": 0.18555330932140351,
      "num_tokens": 82570687.0,
      "step": 37695
    },
    {
      "entropy": 5.467540454864502,
      "epoch": 3.741564112743152,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003657123154482534,
      "loss": 5.0544,
      "mean_token_accuracy": 0.18380136787891388,
      "num_tokens": 82582030.0,
      "step": 37700
    },
    {
      "entropy": 5.3322240829467775,
      "epoch": 3.7420603414053195,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036567989000736386,
      "loss": 4.8707,
      "mean_token_accuracy": 0.19031217843294143,
      "num_tokens": 82592828.0,
      "step": 37705
    },
    {
      "entropy": 5.367533349990845,
      "epoch": 3.742556570067487,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036564746231780836,
      "loss": 4.9293,
      "mean_token_accuracy": 0.19152417182922363,
      "num_tokens": 82603133.0,
      "step": 37710
    },
    {
      "entropy": 5.3464357376098635,
      "epoch": 3.7430527987296545,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003656150323803911,
      "loss": 4.9071,
      "mean_token_accuracy": 0.18921670615673064,
      "num_tokens": 82612889.0,
      "step": 37715
    },
    {
      "entropy": 5.362599182128906,
      "epoch": 3.743549027391822,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036558260019591633,
      "loss": 4.964,
      "mean_token_accuracy": 0.18747362941503526,
      "num_tokens": 82623647.0,
      "step": 37720
    },
    {
      "entropy": 5.388889265060425,
      "epoch": 3.7440452560539894,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036555016576518816,
      "loss": 4.979,
      "mean_token_accuracy": 0.1841151759028435,
      "num_tokens": 82633875.0,
      "step": 37725
    },
    {
      "entropy": 5.426928758621216,
      "epoch": 3.7445414847161573,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003655177290890111,
      "loss": 4.9825,
      "mean_token_accuracy": 0.1848501071333885,
      "num_tokens": 82644596.0,
      "step": 37730
    },
    {
      "entropy": 5.4497167587280275,
      "epoch": 3.745037713378325,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036548529016818936,
      "loss": 4.982,
      "mean_token_accuracy": 0.19004455655813218,
      "num_tokens": 82656073.0,
      "step": 37735
    },
    {
      "entropy": 5.467433738708496,
      "epoch": 3.7455339420404923,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003654528490035274,
      "loss": 5.0248,
      "mean_token_accuracy": 0.1816495567560196,
      "num_tokens": 82666644.0,
      "step": 37740
    },
    {
      "entropy": 5.377786159515381,
      "epoch": 3.7460301707026598,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003654204055958298,
      "loss": 4.9112,
      "mean_token_accuracy": 0.18611085116863252,
      "num_tokens": 82677812.0,
      "step": 37745
    },
    {
      "entropy": 5.3803300857543945,
      "epoch": 3.7465263993648272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036538795994590093,
      "loss": 4.9746,
      "mean_token_accuracy": 0.18596852719783782,
      "num_tokens": 82689215.0,
      "step": 37750
    },
    {
      "entropy": 5.4268693923950195,
      "epoch": 3.7470226280269947,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036535551205454535,
      "loss": 5.0276,
      "mean_token_accuracy": 0.1817925125360489,
      "num_tokens": 82701115.0,
      "step": 37755
    },
    {
      "entropy": 5.432186222076416,
      "epoch": 3.7475188566891626,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036532306192256774,
      "loss": 5.0004,
      "mean_token_accuracy": 0.18725785613059998,
      "num_tokens": 82712064.0,
      "step": 37760
    },
    {
      "entropy": 5.41868953704834,
      "epoch": 3.74801508535133,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003652906095507729,
      "loss": 4.9981,
      "mean_token_accuracy": 0.18288055658340455,
      "num_tokens": 82721515.0,
      "step": 37765
    },
    {
      "entropy": 5.424613189697266,
      "epoch": 3.7485113140134976,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036525815493996557,
      "loss": 5.0554,
      "mean_token_accuracy": 0.17731202244758607,
      "num_tokens": 82731185.0,
      "step": 37770
    },
    {
      "entropy": 5.37995400428772,
      "epoch": 3.749007542675665,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036522569809095034,
      "loss": 4.922,
      "mean_token_accuracy": 0.1873130679130554,
      "num_tokens": 82740609.0,
      "step": 37775
    },
    {
      "entropy": 5.543717575073242,
      "epoch": 3.7495037713378325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036519323900453235,
      "loss": 5.1674,
      "mean_token_accuracy": 0.17738570868968964,
      "num_tokens": 82752737.0,
      "step": 37780
    },
    {
      "entropy": 5.471823120117188,
      "epoch": 3.75,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036516077768151636,
      "loss": 5.0496,
      "mean_token_accuracy": 0.17778754681348802,
      "num_tokens": 82763585.0,
      "step": 37785
    },
    {
      "entropy": 5.352301549911499,
      "epoch": 3.7504962286621675,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003651283141227073,
      "loss": 4.9402,
      "mean_token_accuracy": 0.18879782259464264,
      "num_tokens": 82774173.0,
      "step": 37790
    },
    {
      "entropy": 5.35998797416687,
      "epoch": 3.750992457324335,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003650958483289102,
      "loss": 4.9989,
      "mean_token_accuracy": 0.1878360241651535,
      "num_tokens": 82786279.0,
      "step": 37795
    },
    {
      "entropy": 5.408396291732788,
      "epoch": 3.7514886859865024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003650633803009303,
      "loss": 4.9153,
      "mean_token_accuracy": 0.18307312726974487,
      "num_tokens": 82798087.0,
      "step": 37800
    },
    {
      "entropy": 5.444722843170166,
      "epoch": 3.75198491464867,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036503091003957264,
      "loss": 4.9754,
      "mean_token_accuracy": 0.19041615575551987,
      "num_tokens": 82809712.0,
      "step": 37805
    },
    {
      "entropy": 5.334803056716919,
      "epoch": 3.7524811433108374,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036499843754564235,
      "loss": 4.8966,
      "mean_token_accuracy": 0.19494491219520568,
      "num_tokens": 82821894.0,
      "step": 37810
    },
    {
      "entropy": 5.395246171951294,
      "epoch": 3.7529773719730053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003649659628199448,
      "loss": 4.9654,
      "mean_token_accuracy": 0.186614927649498,
      "num_tokens": 82833094.0,
      "step": 37815
    },
    {
      "entropy": 5.345225381851196,
      "epoch": 3.7534736006351728,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003649334858632852,
      "loss": 4.9222,
      "mean_token_accuracy": 0.1901118353009224,
      "num_tokens": 82843690.0,
      "step": 37820
    },
    {
      "entropy": 5.391059064865113,
      "epoch": 3.7539698292973402,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000364901006676469,
      "loss": 4.9899,
      "mean_token_accuracy": 0.1854412943124771,
      "num_tokens": 82855810.0,
      "step": 37825
    },
    {
      "entropy": 5.375262928009033,
      "epoch": 3.7544660579595077,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036486852526030147,
      "loss": 4.8761,
      "mean_token_accuracy": 0.19300684034824372,
      "num_tokens": 82866836.0,
      "step": 37830
    },
    {
      "entropy": 5.399049568176269,
      "epoch": 3.754962286621675,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003648360416155882,
      "loss": 4.9973,
      "mean_token_accuracy": 0.18762501329183578,
      "num_tokens": 82877645.0,
      "step": 37835
    },
    {
      "entropy": 5.401464891433716,
      "epoch": 3.7554585152838427,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003648035557431347,
      "loss": 4.9894,
      "mean_token_accuracy": 0.18654358386993408,
      "num_tokens": 82888638.0,
      "step": 37840
    },
    {
      "entropy": 5.379884815216064,
      "epoch": 3.7559547439460106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036477106764374646,
      "loss": 4.9313,
      "mean_token_accuracy": 0.19433115273714066,
      "num_tokens": 82898550.0,
      "step": 37845
    },
    {
      "entropy": 5.434292507171631,
      "epoch": 3.756450972608178,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003647385773182292,
      "loss": 4.9984,
      "mean_token_accuracy": 0.19052371233701706,
      "num_tokens": 82908926.0,
      "step": 37850
    },
    {
      "entropy": 5.490703439712524,
      "epoch": 3.7569472012703455,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036470608476738865,
      "loss": 5.0631,
      "mean_token_accuracy": 0.18833167850971222,
      "num_tokens": 82921294.0,
      "step": 37855
    },
    {
      "entropy": 5.435345840454102,
      "epoch": 3.757443429932513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003646735899920304,
      "loss": 5.0092,
      "mean_token_accuracy": 0.18308591991662979,
      "num_tokens": 82931315.0,
      "step": 37860
    },
    {
      "entropy": 5.379985237121582,
      "epoch": 3.7579396585946805,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036464109299296047,
      "loss": 4.9738,
      "mean_token_accuracy": 0.18248577266931534,
      "num_tokens": 82942385.0,
      "step": 37865
    },
    {
      "entropy": 5.392311668395996,
      "epoch": 3.758435887256848,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003646085937709845,
      "loss": 4.9905,
      "mean_token_accuracy": 0.18895552903413773,
      "num_tokens": 82952803.0,
      "step": 37870
    },
    {
      "entropy": 5.40113320350647,
      "epoch": 3.7589321159190154,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003645760923269085,
      "loss": 4.9648,
      "mean_token_accuracy": 0.17973899841308594,
      "num_tokens": 82964143.0,
      "step": 37875
    },
    {
      "entropy": 5.386876106262207,
      "epoch": 3.759428344581183,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003645435886615385,
      "loss": 4.9225,
      "mean_token_accuracy": 0.18294892311096192,
      "num_tokens": 82975800.0,
      "step": 37880
    },
    {
      "entropy": 5.422968244552612,
      "epoch": 3.7599245732433504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036451108277568044,
      "loss": 5.0119,
      "mean_token_accuracy": 0.18378704339265822,
      "num_tokens": 82987462.0,
      "step": 37885
    },
    {
      "entropy": 5.326201057434082,
      "epoch": 3.760420801905518,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036447857467014036,
      "loss": 4.8842,
      "mean_token_accuracy": 0.19065327942371368,
      "num_tokens": 82996754.0,
      "step": 37890
    },
    {
      "entropy": 5.353768253326416,
      "epoch": 3.7609170305676853,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003644460643457245,
      "loss": 4.9368,
      "mean_token_accuracy": 0.1846108227968216,
      "num_tokens": 83009026.0,
      "step": 37895
    },
    {
      "entropy": 5.478455591201782,
      "epoch": 3.7614132592298533,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00036441355180323885,
      "loss": 5.0398,
      "mean_token_accuracy": 0.17704234570264815,
      "num_tokens": 83020451.0,
      "step": 37900
    },
    {
      "entropy": 5.366844749450683,
      "epoch": 3.7619094878920207,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003643810370434899,
      "loss": 4.9012,
      "mean_token_accuracy": 0.18939607441425324,
      "num_tokens": 83030822.0,
      "step": 37905
    },
    {
      "entropy": 5.38420786857605,
      "epoch": 3.762405716554188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003643485200672838,
      "loss": 4.9402,
      "mean_token_accuracy": 0.19590241611003875,
      "num_tokens": 83042441.0,
      "step": 37910
    },
    {
      "entropy": 5.403197479248047,
      "epoch": 3.7629019452163557,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036431600087542695,
      "loss": 4.9655,
      "mean_token_accuracy": 0.18446830958127974,
      "num_tokens": 83054386.0,
      "step": 37915
    },
    {
      "entropy": 5.346181774139405,
      "epoch": 3.763398173878523,
      "grad_norm": 1.0,
      "learning_rate": 0.0003642834794687257,
      "loss": 4.966,
      "mean_token_accuracy": 0.18294676691293715,
      "num_tokens": 83066221.0,
      "step": 37920
    },
    {
      "entropy": 5.381076002120972,
      "epoch": 3.7638944025406906,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003642509558479865,
      "loss": 5.021,
      "mean_token_accuracy": 0.18600305914878845,
      "num_tokens": 83078080.0,
      "step": 37925
    },
    {
      "entropy": 5.436972713470459,
      "epoch": 3.7643906312028586,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036421843001401607,
      "loss": 4.9628,
      "mean_token_accuracy": 0.19622029662132262,
      "num_tokens": 83089486.0,
      "step": 37930
    },
    {
      "entropy": 5.4109580516815186,
      "epoch": 3.764886859865026,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036418590196762076,
      "loss": 5.0038,
      "mean_token_accuracy": 0.18097843527793883,
      "num_tokens": 83101253.0,
      "step": 37935
    },
    {
      "entropy": 5.375190114974975,
      "epoch": 3.7653830885271935,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036415337170960714,
      "loss": 4.9132,
      "mean_token_accuracy": 0.1933341234922409,
      "num_tokens": 83112098.0,
      "step": 37940
    },
    {
      "entropy": 5.379444169998169,
      "epoch": 3.765879317189361,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036412083924078223,
      "loss": 4.9447,
      "mean_token_accuracy": 0.18382408767938613,
      "num_tokens": 83122047.0,
      "step": 37945
    },
    {
      "entropy": 5.4087036609649655,
      "epoch": 3.7663755458515285,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003640883045619523,
      "loss": 4.9718,
      "mean_token_accuracy": 0.18289670199155808,
      "num_tokens": 83132525.0,
      "step": 37950
    },
    {
      "entropy": 5.373443222045898,
      "epoch": 3.766871774513696,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003640557676739245,
      "loss": 5.0256,
      "mean_token_accuracy": 0.1774016335606575,
      "num_tokens": 83143274.0,
      "step": 37955
    },
    {
      "entropy": 5.447248744964599,
      "epoch": 3.7673680031758634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003640232285775056,
      "loss": 5.0353,
      "mean_token_accuracy": 0.18868474513292313,
      "num_tokens": 83154451.0,
      "step": 37960
    },
    {
      "entropy": 5.399449872970581,
      "epoch": 3.767864231838031,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036399068727350245,
      "loss": 4.9536,
      "mean_token_accuracy": 0.17672431468963623,
      "num_tokens": 83165535.0,
      "step": 37965
    },
    {
      "entropy": 5.319287729263306,
      "epoch": 3.7683604605001984,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036395814376272194,
      "loss": 4.8613,
      "mean_token_accuracy": 0.1934925839304924,
      "num_tokens": 83175955.0,
      "step": 37970
    },
    {
      "entropy": 5.350748825073242,
      "epoch": 3.768856689162366,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003639255980459713,
      "loss": 5.0308,
      "mean_token_accuracy": 0.1783616304397583,
      "num_tokens": 83186200.0,
      "step": 37975
    },
    {
      "entropy": 5.43886513710022,
      "epoch": 3.7693529178245333,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003638930501240572,
      "loss": 4.9836,
      "mean_token_accuracy": 0.18172026574611663,
      "num_tokens": 83197121.0,
      "step": 37980
    },
    {
      "entropy": 5.390499019622803,
      "epoch": 3.769849146486701,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036386049999778724,
      "loss": 4.9191,
      "mean_token_accuracy": 0.19163231700658798,
      "num_tokens": 83208580.0,
      "step": 37985
    },
    {
      "entropy": 5.377774381637574,
      "epoch": 3.7703453751488687,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036382794766796824,
      "loss": 4.8812,
      "mean_token_accuracy": 0.19145302027463912,
      "num_tokens": 83218589.0,
      "step": 37990
    },
    {
      "entropy": 5.3666153907775875,
      "epoch": 3.770841603811036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036379539313540757,
      "loss": 4.9695,
      "mean_token_accuracy": 0.18489098250865937,
      "num_tokens": 83229712.0,
      "step": 37995
    },
    {
      "entropy": 5.305514478683472,
      "epoch": 3.7713378324732036,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003637628364009125,
      "loss": 4.9149,
      "mean_token_accuracy": 0.19458130300045012,
      "num_tokens": 83240327.0,
      "step": 38000
    },
    {
      "entropy": 5.317814350128174,
      "epoch": 3.771834061135371,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036373027746529037,
      "loss": 4.9316,
      "mean_token_accuracy": 0.18837788999080657,
      "num_tokens": 83251966.0,
      "step": 38005
    },
    {
      "entropy": 5.364656209945679,
      "epoch": 3.7723302897975386,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003636977163293485,
      "loss": 4.8624,
      "mean_token_accuracy": 0.19161901175975798,
      "num_tokens": 83261948.0,
      "step": 38010
    },
    {
      "entropy": 5.397759246826172,
      "epoch": 3.772826518459706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003636651529938944,
      "loss": 4.982,
      "mean_token_accuracy": 0.18438658863306046,
      "num_tokens": 83273172.0,
      "step": 38015
    },
    {
      "entropy": 5.290429210662841,
      "epoch": 3.773322747121874,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003636325874597356,
      "loss": 4.8905,
      "mean_token_accuracy": 0.18819724768400192,
      "num_tokens": 83283836.0,
      "step": 38020
    },
    {
      "entropy": 5.337431478500366,
      "epoch": 3.7738189757840415,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036360001972767956,
      "loss": 4.984,
      "mean_token_accuracy": 0.1895502343773842,
      "num_tokens": 83294206.0,
      "step": 38025
    },
    {
      "entropy": 5.418859767913818,
      "epoch": 3.774315204446209,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036356744979853406,
      "loss": 4.9814,
      "mean_token_accuracy": 0.19162006229162215,
      "num_tokens": 83305066.0,
      "step": 38030
    },
    {
      "entropy": 5.334401035308838,
      "epoch": 3.7748114331083764,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036353487767310655,
      "loss": 4.9002,
      "mean_token_accuracy": 0.18896165937185289,
      "num_tokens": 83316169.0,
      "step": 38035
    },
    {
      "entropy": 5.409634399414062,
      "epoch": 3.775307661770544,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003635023033522049,
      "loss": 4.9891,
      "mean_token_accuracy": 0.1863592967391014,
      "num_tokens": 83327702.0,
      "step": 38040
    },
    {
      "entropy": 5.387862920761108,
      "epoch": 3.7758038904327114,
      "grad_norm": 1.0,
      "learning_rate": 0.0003634697268366368,
      "loss": 4.9537,
      "mean_token_accuracy": 0.1857646018266678,
      "num_tokens": 83339951.0,
      "step": 38045
    },
    {
      "entropy": 5.39029483795166,
      "epoch": 3.776300119094879,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003634371481272101,
      "loss": 4.927,
      "mean_token_accuracy": 0.1863989293575287,
      "num_tokens": 83351869.0,
      "step": 38050
    },
    {
      "entropy": 5.381684923171997,
      "epoch": 3.7767963477570463,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003634045672247327,
      "loss": 5.005,
      "mean_token_accuracy": 0.18684251755475997,
      "num_tokens": 83362715.0,
      "step": 38055
    },
    {
      "entropy": 5.380738925933838,
      "epoch": 3.777292576419214,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003633719841300126,
      "loss": 4.9262,
      "mean_token_accuracy": 0.18990847319364548,
      "num_tokens": 83372988.0,
      "step": 38060
    },
    {
      "entropy": 5.360979318618774,
      "epoch": 3.7777888050813813,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003633393988438576,
      "loss": 4.9112,
      "mean_token_accuracy": 0.18746692836284637,
      "num_tokens": 83383172.0,
      "step": 38065
    },
    {
      "entropy": 5.298733520507812,
      "epoch": 3.778285033743549,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003633068113670759,
      "loss": 4.8968,
      "mean_token_accuracy": 0.18699806481599807,
      "num_tokens": 83395402.0,
      "step": 38070
    },
    {
      "entropy": 5.396248149871826,
      "epoch": 3.7787812624057167,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036327422170047565,
      "loss": 5.0662,
      "mean_token_accuracy": 0.18648114055395126,
      "num_tokens": 83406936.0,
      "step": 38075
    },
    {
      "entropy": 5.457560110092163,
      "epoch": 3.779277491067884,
      "grad_norm": 1.0,
      "learning_rate": 0.0003632416298448648,
      "loss": 5.0519,
      "mean_token_accuracy": 0.17548438012599946,
      "num_tokens": 83420043.0,
      "step": 38080
    },
    {
      "entropy": 5.389977312088012,
      "epoch": 3.7797737197300516,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036320903580105176,
      "loss": 4.9395,
      "mean_token_accuracy": 0.18879646211862564,
      "num_tokens": 83430083.0,
      "step": 38085
    },
    {
      "entropy": 5.37856912612915,
      "epoch": 3.780269948392219,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036317643956984466,
      "loss": 4.8503,
      "mean_token_accuracy": 0.1949574574828148,
      "num_tokens": 83440992.0,
      "step": 38090
    },
    {
      "entropy": 5.4096753120422365,
      "epoch": 3.7807661770543866,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036314384115205184,
      "loss": 4.9669,
      "mean_token_accuracy": 0.18488237261772156,
      "num_tokens": 83452534.0,
      "step": 38095
    },
    {
      "entropy": 5.296305227279663,
      "epoch": 3.781262405716554,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003631112405484817,
      "loss": 4.8702,
      "mean_token_accuracy": 0.18937936425209045,
      "num_tokens": 83463064.0,
      "step": 38100
    },
    {
      "entropy": 5.407622909545898,
      "epoch": 3.781758634378722,
      "grad_norm": 1.0,
      "learning_rate": 0.00036307863775994275,
      "loss": 4.9897,
      "mean_token_accuracy": 0.1821557641029358,
      "num_tokens": 83474426.0,
      "step": 38105
    },
    {
      "entropy": 5.401681137084961,
      "epoch": 3.7822548630408894,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036304603278724325,
      "loss": 4.993,
      "mean_token_accuracy": 0.18067943602800368,
      "num_tokens": 83485896.0,
      "step": 38110
    },
    {
      "entropy": 5.412091970443726,
      "epoch": 3.782751091703057,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036301342563119194,
      "loss": 5.0238,
      "mean_token_accuracy": 0.18266811221837997,
      "num_tokens": 83496257.0,
      "step": 38115
    },
    {
      "entropy": 5.4113668441772464,
      "epoch": 3.7832473203652244,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003629808162925973,
      "loss": 4.9506,
      "mean_token_accuracy": 0.18926866948604584,
      "num_tokens": 83506500.0,
      "step": 38120
    },
    {
      "entropy": 5.4495916843414305,
      "epoch": 3.783743549027392,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036294820477226797,
      "loss": 4.9546,
      "mean_token_accuracy": 0.18409982770681382,
      "num_tokens": 83517226.0,
      "step": 38125
    },
    {
      "entropy": 5.333236646652222,
      "epoch": 3.7842397776895593,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003629155910710127,
      "loss": 4.9393,
      "mean_token_accuracy": 0.18857716917991638,
      "num_tokens": 83527307.0,
      "step": 38130
    },
    {
      "entropy": 5.375959396362305,
      "epoch": 3.784736006351727,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003628829751896403,
      "loss": 4.9534,
      "mean_token_accuracy": 0.18931064158678054,
      "num_tokens": 83539252.0,
      "step": 38135
    },
    {
      "entropy": 5.405652475357056,
      "epoch": 3.7852322350138943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036285035712895934,
      "loss": 4.9857,
      "mean_token_accuracy": 0.18175870478153228,
      "num_tokens": 83551791.0,
      "step": 38140
    },
    {
      "entropy": 5.363718318939209,
      "epoch": 3.7857284636760618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036281773688977894,
      "loss": 4.897,
      "mean_token_accuracy": 0.18593201637268067,
      "num_tokens": 83561545.0,
      "step": 38145
    },
    {
      "entropy": 5.328941011428833,
      "epoch": 3.7862246923382292,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003627851144729078,
      "loss": 4.8986,
      "mean_token_accuracy": 0.18724241703748704,
      "num_tokens": 83572050.0,
      "step": 38150
    },
    {
      "entropy": 5.365838861465454,
      "epoch": 3.786720921000397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036275248987915497,
      "loss": 4.9815,
      "mean_token_accuracy": 0.1899498388171196,
      "num_tokens": 83583177.0,
      "step": 38155
    },
    {
      "entropy": 5.375204229354859,
      "epoch": 3.7872171496625646,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036271986310932957,
      "loss": 4.9475,
      "mean_token_accuracy": 0.18561187535524368,
      "num_tokens": 83592906.0,
      "step": 38160
    },
    {
      "entropy": 5.376122999191284,
      "epoch": 3.787713378324732,
      "grad_norm": 1.0,
      "learning_rate": 0.0003626872341642406,
      "loss": 4.9937,
      "mean_token_accuracy": 0.18707952946424483,
      "num_tokens": 83602658.0,
      "step": 38165
    },
    {
      "entropy": 5.37216854095459,
      "epoch": 3.7882096069868996,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003626546030446971,
      "loss": 4.9857,
      "mean_token_accuracy": 0.1891256794333458,
      "num_tokens": 83612774.0,
      "step": 38170
    },
    {
      "entropy": 5.402672195434571,
      "epoch": 3.788705835649067,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036262196975150835,
      "loss": 4.9088,
      "mean_token_accuracy": 0.19435665160417556,
      "num_tokens": 83624284.0,
      "step": 38175
    },
    {
      "entropy": 5.388205194473267,
      "epoch": 3.7892020643112345,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003625893342854836,
      "loss": 5.029,
      "mean_token_accuracy": 0.17964451760053635,
      "num_tokens": 83636850.0,
      "step": 38180
    },
    {
      "entropy": 5.411356449127197,
      "epoch": 3.789698292973402,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036255669664743195,
      "loss": 4.9798,
      "mean_token_accuracy": 0.18449868112802506,
      "num_tokens": 83647562.0,
      "step": 38185
    },
    {
      "entropy": 5.36938328742981,
      "epoch": 3.79019452163557,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036252405683816306,
      "loss": 4.9368,
      "mean_token_accuracy": 0.18656123727560042,
      "num_tokens": 83659020.0,
      "step": 38190
    },
    {
      "entropy": 5.38705153465271,
      "epoch": 3.7906907502977374,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003624914148584861,
      "loss": 5.0071,
      "mean_token_accuracy": 0.18367520570755005,
      "num_tokens": 83670162.0,
      "step": 38195
    },
    {
      "entropy": 5.342919826507568,
      "epoch": 3.791186978959905,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003624587707092105,
      "loss": 4.878,
      "mean_token_accuracy": 0.18772806525230407,
      "num_tokens": 83680609.0,
      "step": 38200
    },
    {
      "entropy": 5.393730735778808,
      "epoch": 3.7916832076220723,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036242612439114593,
      "loss": 4.9185,
      "mean_token_accuracy": 0.19098495095968246,
      "num_tokens": 83690912.0,
      "step": 38205
    },
    {
      "entropy": 5.359207057952881,
      "epoch": 3.79217943628424,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036239347590510186,
      "loss": 5.0091,
      "mean_token_accuracy": 0.18213339298963546,
      "num_tokens": 83702022.0,
      "step": 38210
    },
    {
      "entropy": 5.424046611785888,
      "epoch": 3.7926756649464073,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003623608252518879,
      "loss": 5.0612,
      "mean_token_accuracy": 0.17626082748174668,
      "num_tokens": 83713836.0,
      "step": 38215
    },
    {
      "entropy": 5.427211284637451,
      "epoch": 3.7931718936085748,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003623281724323137,
      "loss": 5.0198,
      "mean_token_accuracy": 0.1878461703658104,
      "num_tokens": 83725038.0,
      "step": 38220
    },
    {
      "entropy": 5.360574245452881,
      "epoch": 3.7936681222707422,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000362295517447189,
      "loss": 4.8981,
      "mean_token_accuracy": 0.19058598279953004,
      "num_tokens": 83734984.0,
      "step": 38225
    },
    {
      "entropy": 5.427775526046753,
      "epoch": 3.7941643509329097,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003622628602973236,
      "loss": 4.9705,
      "mean_token_accuracy": 0.1869811400771141,
      "num_tokens": 83746323.0,
      "step": 38230
    },
    {
      "entropy": 5.303826808929443,
      "epoch": 3.794660579595077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036223020098352724,
      "loss": 4.8898,
      "mean_token_accuracy": 0.1864023819565773,
      "num_tokens": 83758334.0,
      "step": 38235
    },
    {
      "entropy": 5.389777755737304,
      "epoch": 3.7951568082572447,
      "grad_norm": 1.0,
      "learning_rate": 0.0003621975395066099,
      "loss": 4.9087,
      "mean_token_accuracy": 0.19062846451997756,
      "num_tokens": 83769022.0,
      "step": 38240
    },
    {
      "entropy": 5.438600492477417,
      "epoch": 3.7956530369194126,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036216487586738144,
      "loss": 5.0019,
      "mean_token_accuracy": 0.18464328050613404,
      "num_tokens": 83780682.0,
      "step": 38245
    },
    {
      "entropy": 5.415435647964477,
      "epoch": 3.79614926558158,
      "grad_norm": 1.046875,
      "learning_rate": 0.000362132210066652,
      "loss": 4.987,
      "mean_token_accuracy": 0.18633269071578978,
      "num_tokens": 83790936.0,
      "step": 38250
    },
    {
      "entropy": 5.400527286529541,
      "epoch": 3.7966454942437475,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036209954210523137,
      "loss": 5.0316,
      "mean_token_accuracy": 0.1767904654145241,
      "num_tokens": 83802498.0,
      "step": 38255
    },
    {
      "entropy": 5.357082271575928,
      "epoch": 3.797141722905915,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003620668719839298,
      "loss": 4.9297,
      "mean_token_accuracy": 0.18576789200305938,
      "num_tokens": 83813269.0,
      "step": 38260
    },
    {
      "entropy": 5.351670455932617,
      "epoch": 3.7976379515680825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036203419970355747,
      "loss": 4.9171,
      "mean_token_accuracy": 0.1860394597053528,
      "num_tokens": 83822692.0,
      "step": 38265
    },
    {
      "entropy": 5.391479206085205,
      "epoch": 3.79813418023025,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003620015252649245,
      "loss": 5.0237,
      "mean_token_accuracy": 0.1860877051949501,
      "num_tokens": 83834375.0,
      "step": 38270
    },
    {
      "entropy": 5.483149623870849,
      "epoch": 3.798630408892418,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036196884866884115,
      "loss": 5.0931,
      "mean_token_accuracy": 0.17969442158937454,
      "num_tokens": 83846328.0,
      "step": 38275
    },
    {
      "entropy": 5.426376676559448,
      "epoch": 3.7991266375545854,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036193616991611784,
      "loss": 4.9499,
      "mean_token_accuracy": 0.18097348511219025,
      "num_tokens": 83857086.0,
      "step": 38280
    },
    {
      "entropy": 5.434071636199951,
      "epoch": 3.799622866216753,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036190348900756475,
      "loss": 5.0353,
      "mean_token_accuracy": 0.18182770609855653,
      "num_tokens": 83867152.0,
      "step": 38285
    },
    {
      "entropy": 5.393623018264771,
      "epoch": 3.8001190948789203,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003618708059439924,
      "loss": 4.95,
      "mean_token_accuracy": 0.1923008069396019,
      "num_tokens": 83878960.0,
      "step": 38290
    },
    {
      "entropy": 5.348230171203613,
      "epoch": 3.8006153235410878,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003618381207262113,
      "loss": 4.8917,
      "mean_token_accuracy": 0.1911143407225609,
      "num_tokens": 83890656.0,
      "step": 38295
    },
    {
      "entropy": 5.363742208480835,
      "epoch": 3.8011115522032553,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003618054333550319,
      "loss": 4.9546,
      "mean_token_accuracy": 0.1889301836490631,
      "num_tokens": 83901099.0,
      "step": 38300
    },
    {
      "entropy": 5.357969951629639,
      "epoch": 3.8016077808654227,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036177274383126487,
      "loss": 4.9786,
      "mean_token_accuracy": 0.18802209794521332,
      "num_tokens": 83912016.0,
      "step": 38305
    },
    {
      "entropy": 5.433476829528809,
      "epoch": 3.80210400952759,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003617400521557207,
      "loss": 5.0088,
      "mean_token_accuracy": 0.17673120200634002,
      "num_tokens": 83923198.0,
      "step": 38310
    },
    {
      "entropy": 5.4200469017028805,
      "epoch": 3.8026002381897577,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003617073583292101,
      "loss": 5.0367,
      "mean_token_accuracy": 0.181343774497509,
      "num_tokens": 83935310.0,
      "step": 38315
    },
    {
      "entropy": 5.360188293457031,
      "epoch": 3.803096466851925,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036167466235254394,
      "loss": 4.9362,
      "mean_token_accuracy": 0.18506519943475724,
      "num_tokens": 83947104.0,
      "step": 38320
    },
    {
      "entropy": 5.34461555480957,
      "epoch": 3.8035926955140926,
      "grad_norm": 1.0,
      "learning_rate": 0.00036164196422653293,
      "loss": 4.9386,
      "mean_token_accuracy": 0.18566972464323045,
      "num_tokens": 83957263.0,
      "step": 38325
    },
    {
      "entropy": 5.303035497665405,
      "epoch": 3.8040889241762605,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036160926395198786,
      "loss": 4.8835,
      "mean_token_accuracy": 0.19697186797857286,
      "num_tokens": 83970680.0,
      "step": 38330
    },
    {
      "entropy": 5.415306234359742,
      "epoch": 3.804585152838428,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003615765615297197,
      "loss": 5.014,
      "mean_token_accuracy": 0.1825781375169754,
      "num_tokens": 83982456.0,
      "step": 38335
    },
    {
      "entropy": 5.404288578033447,
      "epoch": 3.8050813815005955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003615438569605394,
      "loss": 5.0298,
      "mean_token_accuracy": 0.18567270487546922,
      "num_tokens": 83993303.0,
      "step": 38340
    },
    {
      "entropy": 5.428073215484619,
      "epoch": 3.805577610162763,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00036151115024525777,
      "loss": 4.9746,
      "mean_token_accuracy": 0.18167368024587632,
      "num_tokens": 84004701.0,
      "step": 38345
    },
    {
      "entropy": 5.433213090896606,
      "epoch": 3.8060738388249304,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036147844138468616,
      "loss": 5.0254,
      "mean_token_accuracy": 0.18761799186468125,
      "num_tokens": 84014779.0,
      "step": 38350
    },
    {
      "entropy": 5.407609558105468,
      "epoch": 3.806570067487098,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003614457303796356,
      "loss": 5.0334,
      "mean_token_accuracy": 0.1807498648762703,
      "num_tokens": 84025309.0,
      "step": 38355
    },
    {
      "entropy": 5.4826795101165775,
      "epoch": 3.807066296149266,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003614130172309171,
      "loss": 5.1066,
      "mean_token_accuracy": 0.17940421849489213,
      "num_tokens": 84037406.0,
      "step": 38360
    },
    {
      "entropy": 5.431571006774902,
      "epoch": 3.8075625248114333,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003613803019393419,
      "loss": 5.0143,
      "mean_token_accuracy": 0.187004853785038,
      "num_tokens": 84048439.0,
      "step": 38365
    },
    {
      "entropy": 5.413782262802124,
      "epoch": 3.808058753473601,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036134758450572153,
      "loss": 4.9952,
      "mean_token_accuracy": 0.18169749677181243,
      "num_tokens": 84059681.0,
      "step": 38370
    },
    {
      "entropy": 5.5062848091125485,
      "epoch": 3.8085549821357683,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000361314864930867,
      "loss": 5.0534,
      "mean_token_accuracy": 0.18025970011949538,
      "num_tokens": 84069945.0,
      "step": 38375
    },
    {
      "entropy": 5.535307455062866,
      "epoch": 3.8090512107979357,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003612821432155898,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1738048419356346,
      "num_tokens": 84082440.0,
      "step": 38380
    },
    {
      "entropy": 5.376684188842773,
      "epoch": 3.809547439460103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003612494193607015,
      "loss": 4.9247,
      "mean_token_accuracy": 0.19897170066833497,
      "num_tokens": 84093431.0,
      "step": 38385
    },
    {
      "entropy": 5.3939775943756105,
      "epoch": 3.8100436681222707,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003612166933670134,
      "loss": 5.05,
      "mean_token_accuracy": 0.18118272423744203,
      "num_tokens": 84105904.0,
      "step": 38390
    },
    {
      "entropy": 5.418429183959961,
      "epoch": 3.810539896784438,
      "grad_norm": 0.921875,
      "learning_rate": 0.000361183965235337,
      "loss": 5.0626,
      "mean_token_accuracy": 0.17860884219408035,
      "num_tokens": 84117967.0,
      "step": 38395
    },
    {
      "entropy": 5.416527366638183,
      "epoch": 3.8110361254466056,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003611512349664841,
      "loss": 4.8844,
      "mean_token_accuracy": 0.1909754142165184,
      "num_tokens": 84129287.0,
      "step": 38400
    },
    {
      "entropy": 5.47270770072937,
      "epoch": 3.811532354108773,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003611185025612661,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18608985841274261,
      "num_tokens": 84140291.0,
      "step": 38405
    },
    {
      "entropy": 5.479383277893066,
      "epoch": 3.8120285827709406,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003610857680204949,
      "loss": 5.0195,
      "mean_token_accuracy": 0.1849560633301735,
      "num_tokens": 84150955.0,
      "step": 38410
    },
    {
      "entropy": 5.3728643417358395,
      "epoch": 3.8125248114331085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036105303134498207,
      "loss": 4.9008,
      "mean_token_accuracy": 0.1931607335805893,
      "num_tokens": 84161665.0,
      "step": 38415
    },
    {
      "entropy": 5.374589157104492,
      "epoch": 3.813021040095276,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036102029253553945,
      "loss": 4.9934,
      "mean_token_accuracy": 0.18340943455696107,
      "num_tokens": 84172225.0,
      "step": 38420
    },
    {
      "entropy": 5.385047245025635,
      "epoch": 3.8135172687574435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003609875515929791,
      "loss": 4.9373,
      "mean_token_accuracy": 0.18159808069467545,
      "num_tokens": 84181877.0,
      "step": 38425
    },
    {
      "entropy": 5.343411445617676,
      "epoch": 3.814013497419611,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003609548085181126,
      "loss": 4.9334,
      "mean_token_accuracy": 0.18633129298686982,
      "num_tokens": 84193342.0,
      "step": 38430
    },
    {
      "entropy": 5.361592197418213,
      "epoch": 3.8145097260817784,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003609220633117521,
      "loss": 4.9117,
      "mean_token_accuracy": 0.18666685819625856,
      "num_tokens": 84204719.0,
      "step": 38435
    },
    {
      "entropy": 5.367474174499511,
      "epoch": 3.815005954743946,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003608893159747097,
      "loss": 4.8921,
      "mean_token_accuracy": 0.18506485670804979,
      "num_tokens": 84214931.0,
      "step": 38440
    },
    {
      "entropy": 5.340617656707764,
      "epoch": 3.815502183406114,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036085656650779716,
      "loss": 4.9567,
      "mean_token_accuracy": 0.19176491498947143,
      "num_tokens": 84227429.0,
      "step": 38445
    },
    {
      "entropy": 5.345865678787232,
      "epoch": 3.8159984120682813,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036082381491182685,
      "loss": 4.9511,
      "mean_token_accuracy": 0.1858614295721054,
      "num_tokens": 84236917.0,
      "step": 38450
    },
    {
      "entropy": 5.359819173812866,
      "epoch": 3.8164946407304488,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003607910611876109,
      "loss": 4.9408,
      "mean_token_accuracy": 0.18840288668870925,
      "num_tokens": 84247964.0,
      "step": 38455
    },
    {
      "entropy": 5.409671306610107,
      "epoch": 3.8169908693926162,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003607583053359615,
      "loss": 5.0489,
      "mean_token_accuracy": 0.18191129863262176,
      "num_tokens": 84259559.0,
      "step": 38460
    },
    {
      "entropy": 5.40087103843689,
      "epoch": 3.8174870980547837,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003607255473576908,
      "loss": 4.9093,
      "mean_token_accuracy": 0.18809125423431397,
      "num_tokens": 84270849.0,
      "step": 38465
    },
    {
      "entropy": 5.405011796951294,
      "epoch": 3.817983326716951,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036069278725361144,
      "loss": 4.9092,
      "mean_token_accuracy": 0.19039339423179627,
      "num_tokens": 84281977.0,
      "step": 38470
    },
    {
      "entropy": 5.371596574783325,
      "epoch": 3.8184795553791187,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036066002502453556,
      "loss": 4.9567,
      "mean_token_accuracy": 0.18602424561977388,
      "num_tokens": 84294289.0,
      "step": 38475
    },
    {
      "entropy": 5.3550048828125,
      "epoch": 3.818975784041286,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003606272606712755,
      "loss": 4.9435,
      "mean_token_accuracy": 0.1911015674471855,
      "num_tokens": 84304114.0,
      "step": 38480
    },
    {
      "entropy": 5.321315574645996,
      "epoch": 3.8194720127034536,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036059449419464397,
      "loss": 4.8791,
      "mean_token_accuracy": 0.18631520569324495,
      "num_tokens": 84314467.0,
      "step": 38485
    },
    {
      "entropy": 5.4240429401397705,
      "epoch": 3.819968241365621,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003605617255954535,
      "loss": 5.0605,
      "mean_token_accuracy": 0.18236940652132033,
      "num_tokens": 84325565.0,
      "step": 38490
    },
    {
      "entropy": 5.3507835388183596,
      "epoch": 3.8204644700277886,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003605289548745164,
      "loss": 4.9329,
      "mean_token_accuracy": 0.18550778478384017,
      "num_tokens": 84336578.0,
      "step": 38495
    },
    {
      "entropy": 5.39247875213623,
      "epoch": 3.8209606986899565,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003604961820326457,
      "loss": 4.9365,
      "mean_token_accuracy": 0.18318533301353454,
      "num_tokens": 84347447.0,
      "step": 38500
    },
    {
      "entropy": 5.318336057662964,
      "epoch": 3.821456927352124,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036046340707065387,
      "loss": 4.9277,
      "mean_token_accuracy": 0.18660268187522888,
      "num_tokens": 84359527.0,
      "step": 38505
    },
    {
      "entropy": 5.37781138420105,
      "epoch": 3.8219531560142914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003604306299893537,
      "loss": 4.9508,
      "mean_token_accuracy": 0.1804889976978302,
      "num_tokens": 84369653.0,
      "step": 38510
    },
    {
      "entropy": 5.354791688919067,
      "epoch": 3.822449384676459,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036039785078955793,
      "loss": 4.8558,
      "mean_token_accuracy": 0.19240472912788392,
      "num_tokens": 84378881.0,
      "step": 38515
    },
    {
      "entropy": 5.45207109451294,
      "epoch": 3.8229456133386264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003603650694720796,
      "loss": 5.0676,
      "mean_token_accuracy": 0.17979714870452881,
      "num_tokens": 84390382.0,
      "step": 38520
    },
    {
      "entropy": 5.405211114883423,
      "epoch": 3.823441842000794,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036033228603773136,
      "loss": 4.9597,
      "mean_token_accuracy": 0.1878790721297264,
      "num_tokens": 84401032.0,
      "step": 38525
    },
    {
      "entropy": 5.422819900512695,
      "epoch": 3.8239380706629618,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036029950048732635,
      "loss": 5.0916,
      "mean_token_accuracy": 0.18033552169799805,
      "num_tokens": 84411934.0,
      "step": 38530
    },
    {
      "entropy": 5.387335920333863,
      "epoch": 3.8244342993251292,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036026671282167753,
      "loss": 4.9373,
      "mean_token_accuracy": 0.18766212910413743,
      "num_tokens": 84423291.0,
      "step": 38535
    },
    {
      "entropy": 5.390517997741699,
      "epoch": 3.8249305279872967,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000360233923041598,
      "loss": 5.0287,
      "mean_token_accuracy": 0.17917223125696183,
      "num_tokens": 84436451.0,
      "step": 38540
    },
    {
      "entropy": 5.435557842254639,
      "epoch": 3.825426756649464,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036020113114790085,
      "loss": 4.9641,
      "mean_token_accuracy": 0.1872591882944107,
      "num_tokens": 84447933.0,
      "step": 38545
    },
    {
      "entropy": 5.385077857971192,
      "epoch": 3.8259229853116317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003601683371413992,
      "loss": 4.9338,
      "mean_token_accuracy": 0.18827173262834548,
      "num_tokens": 84458135.0,
      "step": 38550
    },
    {
      "entropy": 5.442798137664795,
      "epoch": 3.826419213973799,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036013554102290636,
      "loss": 5.0001,
      "mean_token_accuracy": 0.18371894359588622,
      "num_tokens": 84468620.0,
      "step": 38555
    },
    {
      "entropy": 5.3387367725372314,
      "epoch": 3.8269154426359666,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003601027427932355,
      "loss": 4.8788,
      "mean_token_accuracy": 0.18751574009656907,
      "num_tokens": 84479891.0,
      "step": 38560
    },
    {
      "entropy": 5.367966651916504,
      "epoch": 3.827411671298134,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003600699424532,
      "loss": 4.9948,
      "mean_token_accuracy": 0.18879065662622452,
      "num_tokens": 84490621.0,
      "step": 38565
    },
    {
      "entropy": 5.403881978988648,
      "epoch": 3.8279078999603016,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036003714000361326,
      "loss": 4.9577,
      "mean_token_accuracy": 0.1931892603635788,
      "num_tokens": 84500940.0,
      "step": 38570
    },
    {
      "entropy": 5.351820993423462,
      "epoch": 3.828404128622469,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036000433544528866,
      "loss": 4.9794,
      "mean_token_accuracy": 0.18868977427482606,
      "num_tokens": 84511891.0,
      "step": 38575
    },
    {
      "entropy": 5.416730356216431,
      "epoch": 3.8289003572846365,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003599715287790397,
      "loss": 4.9836,
      "mean_token_accuracy": 0.1851874530315399,
      "num_tokens": 84522546.0,
      "step": 38580
    },
    {
      "entropy": 5.457096433639526,
      "epoch": 3.8293965859468044,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035993872000567994,
      "loss": 4.9664,
      "mean_token_accuracy": 0.18188048750162125,
      "num_tokens": 84532281.0,
      "step": 38585
    },
    {
      "entropy": 5.380301141738892,
      "epoch": 3.829892814608972,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035990590912602304,
      "loss": 4.9268,
      "mean_token_accuracy": 0.18530546128749847,
      "num_tokens": 84542477.0,
      "step": 38590
    },
    {
      "entropy": 5.40141019821167,
      "epoch": 3.8303890432711394,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035987309614088245,
      "loss": 4.9429,
      "mean_token_accuracy": 0.18691912591457366,
      "num_tokens": 84552780.0,
      "step": 38595
    },
    {
      "entropy": 5.43554048538208,
      "epoch": 3.830885271933307,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035984028105107195,
      "loss": 5.0164,
      "mean_token_accuracy": 0.1813918322324753,
      "num_tokens": 84564895.0,
      "step": 38600
    },
    {
      "entropy": 5.353679847717285,
      "epoch": 3.8313815005954743,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035980746385740536,
      "loss": 4.8941,
      "mean_token_accuracy": 0.19075107276439668,
      "num_tokens": 84575680.0,
      "step": 38605
    },
    {
      "entropy": 5.329762983322143,
      "epoch": 3.831877729257642,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035977464456069636,
      "loss": 4.9406,
      "mean_token_accuracy": 0.18733202815055847,
      "num_tokens": 84586865.0,
      "step": 38610
    },
    {
      "entropy": 5.34686770439148,
      "epoch": 3.8323739579198093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035974182316175887,
      "loss": 4.9538,
      "mean_token_accuracy": 0.19021493643522264,
      "num_tokens": 84598285.0,
      "step": 38615
    },
    {
      "entropy": 5.432721090316773,
      "epoch": 3.832870186581977,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003597089996614068,
      "loss": 5.0145,
      "mean_token_accuracy": 0.18151868283748626,
      "num_tokens": 84609481.0,
      "step": 38620
    },
    {
      "entropy": 5.507374858856201,
      "epoch": 3.8333664152441447,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035967617406045406,
      "loss": 5.0751,
      "mean_token_accuracy": 0.17914764434099198,
      "num_tokens": 84619719.0,
      "step": 38625
    },
    {
      "entropy": 5.392676830291748,
      "epoch": 3.833862643906312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035964334635971455,
      "loss": 4.9331,
      "mean_token_accuracy": 0.18127639293670655,
      "num_tokens": 84630074.0,
      "step": 38630
    },
    {
      "entropy": 5.320416450500488,
      "epoch": 3.8343588725684796,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003596105165600025,
      "loss": 4.8814,
      "mean_token_accuracy": 0.19372988045215606,
      "num_tokens": 84640508.0,
      "step": 38635
    },
    {
      "entropy": 5.374908876419068,
      "epoch": 3.834855101230647,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035957768466213203,
      "loss": 4.9408,
      "mean_token_accuracy": 0.19038447141647338,
      "num_tokens": 84651824.0,
      "step": 38640
    },
    {
      "entropy": 5.3919134616851805,
      "epoch": 3.8353513298928146,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035954485066691717,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18965740799903869,
      "num_tokens": 84662107.0,
      "step": 38645
    },
    {
      "entropy": 5.3496589183807375,
      "epoch": 3.835847558554982,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003595120145751723,
      "loss": 4.8917,
      "mean_token_accuracy": 0.19463938772678374,
      "num_tokens": 84674192.0,
      "step": 38650
    },
    {
      "entropy": 5.335631656646728,
      "epoch": 3.8363437872171495,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035947917638771146,
      "loss": 4.9322,
      "mean_token_accuracy": 0.1878965198993683,
      "num_tokens": 84684294.0,
      "step": 38655
    },
    {
      "entropy": 5.350260066986084,
      "epoch": 3.836840015879317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035944633610534904,
      "loss": 4.9518,
      "mean_token_accuracy": 0.18646268993616105,
      "num_tokens": 84695934.0,
      "step": 38660
    },
    {
      "entropy": 5.450834798812866,
      "epoch": 3.8373362445414845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035941349372889956,
      "loss": 5.0054,
      "mean_token_accuracy": 0.182110732793808,
      "num_tokens": 84707549.0,
      "step": 38665
    },
    {
      "entropy": 5.379143619537354,
      "epoch": 3.8378324732036524,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035938064925917726,
      "loss": 4.8815,
      "mean_token_accuracy": 0.19715082347393037,
      "num_tokens": 84717864.0,
      "step": 38670
    },
    {
      "entropy": 5.3343743801116945,
      "epoch": 3.83832870186582,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003593478026969967,
      "loss": 4.9142,
      "mean_token_accuracy": 0.18281013667583465,
      "num_tokens": 84728328.0,
      "step": 38675
    },
    {
      "entropy": 5.4428245544433596,
      "epoch": 3.8388249305279873,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035931495404317243,
      "loss": 5.0545,
      "mean_token_accuracy": 0.1745957300066948,
      "num_tokens": 84739037.0,
      "step": 38680
    },
    {
      "entropy": 5.323485422134399,
      "epoch": 3.839321159190155,
      "grad_norm": 1.046875,
      "learning_rate": 0.000359282103298519,
      "loss": 4.9676,
      "mean_token_accuracy": 0.18527508527040482,
      "num_tokens": 84750356.0,
      "step": 38685
    },
    {
      "entropy": 5.396051359176636,
      "epoch": 3.8398173878523223,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035924925046385085,
      "loss": 4.9045,
      "mean_token_accuracy": 0.18959163278341293,
      "num_tokens": 84759622.0,
      "step": 38690
    },
    {
      "entropy": 5.330963754653931,
      "epoch": 3.8403136165144898,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035921639553998297,
      "loss": 4.9554,
      "mean_token_accuracy": 0.18477234840393067,
      "num_tokens": 84771602.0,
      "step": 38695
    },
    {
      "entropy": 5.350813007354736,
      "epoch": 3.8408098451766572,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035918353852772996,
      "loss": 4.938,
      "mean_token_accuracy": 0.1865801304578781,
      "num_tokens": 84781730.0,
      "step": 38700
    },
    {
      "entropy": 5.4214160442352295,
      "epoch": 3.841306073838825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003591506794279066,
      "loss": 4.9244,
      "mean_token_accuracy": 0.19279717206954955,
      "num_tokens": 84793846.0,
      "step": 38705
    },
    {
      "entropy": 5.473769855499268,
      "epoch": 3.8418023025009926,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003591178182413276,
      "loss": 5.0853,
      "mean_token_accuracy": 0.18205033987760544,
      "num_tokens": 84804704.0,
      "step": 38710
    },
    {
      "entropy": 5.390202856063842,
      "epoch": 3.84229853116316,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035908495496880804,
      "loss": 4.9042,
      "mean_token_accuracy": 0.18948998153209687,
      "num_tokens": 84816052.0,
      "step": 38715
    },
    {
      "entropy": 5.392240524291992,
      "epoch": 3.8427947598253276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003590520896111628,
      "loss": 5.0235,
      "mean_token_accuracy": 0.1847312182188034,
      "num_tokens": 84827412.0,
      "step": 38720
    },
    {
      "entropy": 5.448624753952027,
      "epoch": 3.843290988487495,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035901922216920683,
      "loss": 4.9995,
      "mean_token_accuracy": 0.1823480248451233,
      "num_tokens": 84840003.0,
      "step": 38725
    },
    {
      "entropy": 5.42149248123169,
      "epoch": 3.8437872171496625,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003589863526437552,
      "loss": 4.8972,
      "mean_token_accuracy": 0.19568770676851271,
      "num_tokens": 84850385.0,
      "step": 38730
    },
    {
      "entropy": 5.363600873947144,
      "epoch": 3.84428344581183,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003589534810356231,
      "loss": 4.8946,
      "mean_token_accuracy": 0.18659417778253556,
      "num_tokens": 84860628.0,
      "step": 38735
    },
    {
      "entropy": 5.350324249267578,
      "epoch": 3.8447796744739975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035892060734562544,
      "loss": 4.8796,
      "mean_token_accuracy": 0.19282436966896058,
      "num_tokens": 84870654.0,
      "step": 38740
    },
    {
      "entropy": 5.299403381347656,
      "epoch": 3.845275903136165,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035888773157457756,
      "loss": 4.9079,
      "mean_token_accuracy": 0.18566543459892274,
      "num_tokens": 84882928.0,
      "step": 38745
    },
    {
      "entropy": 5.425803756713867,
      "epoch": 3.8457721317983324,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003588548537232947,
      "loss": 4.9805,
      "mean_token_accuracy": 0.18945468068122864,
      "num_tokens": 84893626.0,
      "step": 38750
    },
    {
      "entropy": 5.398520517349243,
      "epoch": 3.8462683604605004,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035882197379259224,
      "loss": 4.9975,
      "mean_token_accuracy": 0.18129572421312332,
      "num_tokens": 84904442.0,
      "step": 38755
    },
    {
      "entropy": 5.397961902618408,
      "epoch": 3.846764589122668,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003587890917832854,
      "loss": 4.983,
      "mean_token_accuracy": 0.18801040798425675,
      "num_tokens": 84914444.0,
      "step": 38760
    },
    {
      "entropy": 5.39901213645935,
      "epoch": 3.8472608177848353,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003587562076961897,
      "loss": 5.0279,
      "mean_token_accuracy": 0.18463262766599656,
      "num_tokens": 84925274.0,
      "step": 38765
    },
    {
      "entropy": 5.374049234390259,
      "epoch": 3.847757046447003,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035872332153212046,
      "loss": 4.9235,
      "mean_token_accuracy": 0.1875335842370987,
      "num_tokens": 84936236.0,
      "step": 38770
    },
    {
      "entropy": 5.441817665100098,
      "epoch": 3.8482532751091703,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035869043329189314,
      "loss": 5.0341,
      "mean_token_accuracy": 0.17754068821668625,
      "num_tokens": 84947895.0,
      "step": 38775
    },
    {
      "entropy": 5.3428243637084964,
      "epoch": 3.8487495037713377,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003586575429763236,
      "loss": 4.9311,
      "mean_token_accuracy": 0.19218553006649017,
      "num_tokens": 84958584.0,
      "step": 38780
    },
    {
      "entropy": 5.4272322177886965,
      "epoch": 3.849245732433505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003586246505862272,
      "loss": 5.0089,
      "mean_token_accuracy": 0.1882467657327652,
      "num_tokens": 84970135.0,
      "step": 38785
    },
    {
      "entropy": 5.520283937454224,
      "epoch": 3.849741961095673,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035859175612241957,
      "loss": 5.0337,
      "mean_token_accuracy": 0.1808510735630989,
      "num_tokens": 84982086.0,
      "step": 38790
    },
    {
      "entropy": 5.398874998092651,
      "epoch": 3.8502381897578406,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035855885958571666,
      "loss": 4.9438,
      "mean_token_accuracy": 0.1914091870188713,
      "num_tokens": 84994399.0,
      "step": 38795
    },
    {
      "entropy": 5.443240833282471,
      "epoch": 3.850734418420008,
      "grad_norm": 1.109375,
      "learning_rate": 0.000358525960976934,
      "loss": 5.0526,
      "mean_token_accuracy": 0.18332272469997407,
      "num_tokens": 85004248.0,
      "step": 38800
    },
    {
      "entropy": 5.4444108486175535,
      "epoch": 3.8512306470821756,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003584930602968875,
      "loss": 5.0693,
      "mean_token_accuracy": 0.17799494415521622,
      "num_tokens": 85015306.0,
      "step": 38805
    },
    {
      "entropy": 5.3613471508026125,
      "epoch": 3.851726875744343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035846015754639304,
      "loss": 4.8739,
      "mean_token_accuracy": 0.1920573815703392,
      "num_tokens": 85025566.0,
      "step": 38810
    },
    {
      "entropy": 5.373002719879151,
      "epoch": 3.8522231044065105,
      "grad_norm": 1.0,
      "learning_rate": 0.0003584272527262665,
      "loss": 4.9835,
      "mean_token_accuracy": 0.18812242448329924,
      "num_tokens": 85036692.0,
      "step": 38815
    },
    {
      "entropy": 5.442480039596558,
      "epoch": 3.852719333068678,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035839434583732384,
      "loss": 5.0236,
      "mean_token_accuracy": 0.18059470355510712,
      "num_tokens": 85046712.0,
      "step": 38820
    },
    {
      "entropy": 5.460082530975342,
      "epoch": 3.8532155617308455,
      "grad_norm": 1.0,
      "learning_rate": 0.00035836143688038107,
      "loss": 5.0031,
      "mean_token_accuracy": 0.18558603823184966,
      "num_tokens": 85056122.0,
      "step": 38825
    },
    {
      "entropy": 5.401173400878906,
      "epoch": 3.853711790393013,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035832852585625436,
      "loss": 5.0035,
      "mean_token_accuracy": 0.1794420063495636,
      "num_tokens": 85068201.0,
      "step": 38830
    },
    {
      "entropy": 5.445226383209229,
      "epoch": 3.8542080190551804,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003582956127657598,
      "loss": 4.9889,
      "mean_token_accuracy": 0.18897037953138351,
      "num_tokens": 85079248.0,
      "step": 38835
    },
    {
      "entropy": 5.337292766571045,
      "epoch": 3.8547042477173483,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0003582626976097134,
      "loss": 4.8565,
      "mean_token_accuracy": 0.18871454745531083,
      "num_tokens": 85091602.0,
      "step": 38840
    },
    {
      "entropy": 5.331934642791748,
      "epoch": 3.855200476379516,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003582297803889316,
      "loss": 4.8767,
      "mean_token_accuracy": 0.19425364583730698,
      "num_tokens": 85101540.0,
      "step": 38845
    },
    {
      "entropy": 5.324833583831787,
      "epoch": 3.8556967050416833,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003581968611042306,
      "loss": 4.9482,
      "mean_token_accuracy": 0.18672905415296553,
      "num_tokens": 85112856.0,
      "step": 38850
    },
    {
      "entropy": 5.387713479995727,
      "epoch": 3.8561929337038507,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035816393975642675,
      "loss": 4.9363,
      "mean_token_accuracy": 0.18733417242765427,
      "num_tokens": 85123269.0,
      "step": 38855
    },
    {
      "entropy": 5.301868867874146,
      "epoch": 3.856689162366018,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035813101634633626,
      "loss": 4.815,
      "mean_token_accuracy": 0.20223129242658616,
      "num_tokens": 85134175.0,
      "step": 38860
    },
    {
      "entropy": 5.38244857788086,
      "epoch": 3.8571853910281857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035809809087477586,
      "loss": 4.9903,
      "mean_token_accuracy": 0.18552082329988479,
      "num_tokens": 85146808.0,
      "step": 38865
    },
    {
      "entropy": 5.398805761337281,
      "epoch": 3.857681619690353,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003580651633425617,
      "loss": 4.9259,
      "mean_token_accuracy": 0.18102766275405885,
      "num_tokens": 85158651.0,
      "step": 38870
    },
    {
      "entropy": 5.454098749160766,
      "epoch": 3.858177848352521,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003580322337505106,
      "loss": 4.982,
      "mean_token_accuracy": 0.18477554768323898,
      "num_tokens": 85169152.0,
      "step": 38875
    },
    {
      "entropy": 5.361530876159668,
      "epoch": 3.8586740770146886,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035799930209943906,
      "loss": 4.991,
      "mean_token_accuracy": 0.17592110931873323,
      "num_tokens": 85180516.0,
      "step": 38880
    },
    {
      "entropy": 5.424444198608398,
      "epoch": 3.859170305676856,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003579663683901636,
      "loss": 4.9457,
      "mean_token_accuracy": 0.1841923028230667,
      "num_tokens": 85192109.0,
      "step": 38885
    },
    {
      "entropy": 5.416516447067261,
      "epoch": 3.8596665343390235,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035793343262350094,
      "loss": 4.9581,
      "mean_token_accuracy": 0.19018668979406356,
      "num_tokens": 85203005.0,
      "step": 38890
    },
    {
      "entropy": 5.37452507019043,
      "epoch": 3.860162763001191,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003579004948002679,
      "loss": 4.9523,
      "mean_token_accuracy": 0.1896603912115097,
      "num_tokens": 85214374.0,
      "step": 38895
    },
    {
      "entropy": 5.28033127784729,
      "epoch": 3.8606589916633585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035786755492128134,
      "loss": 4.8478,
      "mean_token_accuracy": 0.19545539915561677,
      "num_tokens": 85225056.0,
      "step": 38900
    },
    {
      "entropy": 5.440131378173828,
      "epoch": 3.861155220325526,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003578346129873578,
      "loss": 5.0878,
      "mean_token_accuracy": 0.1780263215303421,
      "num_tokens": 85236154.0,
      "step": 38905
    },
    {
      "entropy": 5.367788171768188,
      "epoch": 3.8616514489876934,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035780166899931437,
      "loss": 4.9382,
      "mean_token_accuracy": 0.18726292848587037,
      "num_tokens": 85246757.0,
      "step": 38910
    },
    {
      "entropy": 5.391908311843872,
      "epoch": 3.862147677649861,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003577687229579681,
      "loss": 4.8976,
      "mean_token_accuracy": 0.18914539217948914,
      "num_tokens": 85258066.0,
      "step": 38915
    },
    {
      "entropy": 5.407634353637695,
      "epoch": 3.8626439063120284,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003577357748641357,
      "loss": 4.9737,
      "mean_token_accuracy": 0.18129431158304216,
      "num_tokens": 85268204.0,
      "step": 38920
    },
    {
      "entropy": 5.302196931838989,
      "epoch": 3.863140134974196,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003577028247186344,
      "loss": 4.8914,
      "mean_token_accuracy": 0.18911400437355042,
      "num_tokens": 85278808.0,
      "step": 38925
    },
    {
      "entropy": 5.382884454727173,
      "epoch": 3.8636363636363638,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035766987252228133,
      "loss": 4.9549,
      "mean_token_accuracy": 0.1844101995229721,
      "num_tokens": 85290082.0,
      "step": 38930
    },
    {
      "entropy": 5.329162979125977,
      "epoch": 3.8641325922985312,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003576369182758935,
      "loss": 4.9533,
      "mean_token_accuracy": 0.18752709478139878,
      "num_tokens": 85300313.0,
      "step": 38935
    },
    {
      "entropy": 5.332443046569824,
      "epoch": 3.8646288209606987,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003576039619802882,
      "loss": 4.8295,
      "mean_token_accuracy": 0.2008573070168495,
      "num_tokens": 85310967.0,
      "step": 38940
    },
    {
      "entropy": 5.419860935211181,
      "epoch": 3.865125049622866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003575710036362826,
      "loss": 4.9362,
      "mean_token_accuracy": 0.18898520320653917,
      "num_tokens": 85321835.0,
      "step": 38945
    },
    {
      "entropy": 5.413885641098022,
      "epoch": 3.8656212782850337,
      "grad_norm": 1.125,
      "learning_rate": 0.0003575380432446941,
      "loss": 4.9832,
      "mean_token_accuracy": 0.1834860622882843,
      "num_tokens": 85334193.0,
      "step": 38950
    },
    {
      "entropy": 5.404535102844238,
      "epoch": 3.866117506947201,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003575050808063399,
      "loss": 4.977,
      "mean_token_accuracy": 0.18633776754140854,
      "num_tokens": 85344864.0,
      "step": 38955
    },
    {
      "entropy": 5.36099100112915,
      "epoch": 3.866613735609369,
      "grad_norm": 1.125,
      "learning_rate": 0.00035747211632203757,
      "loss": 4.8987,
      "mean_token_accuracy": 0.188039393723011,
      "num_tokens": 85356183.0,
      "step": 38960
    },
    {
      "entropy": 5.392890882492066,
      "epoch": 3.8671099642715365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003574391497926044,
      "loss": 4.8876,
      "mean_token_accuracy": 0.19414369016885757,
      "num_tokens": 85367157.0,
      "step": 38965
    },
    {
      "entropy": 5.410049247741699,
      "epoch": 3.867606192933704,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000357406181218858,
      "loss": 5.0008,
      "mean_token_accuracy": 0.18726975172758104,
      "num_tokens": 85378995.0,
      "step": 38970
    },
    {
      "entropy": 5.349441814422607,
      "epoch": 3.8681024215958715,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003573732106016158,
      "loss": 4.9452,
      "mean_token_accuracy": 0.1896394282579422,
      "num_tokens": 85389949.0,
      "step": 38975
    },
    {
      "entropy": 5.366569852828979,
      "epoch": 3.868598650258039,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003573402379416955,
      "loss": 4.987,
      "mean_token_accuracy": 0.18052351474761963,
      "num_tokens": 85401261.0,
      "step": 38980
    },
    {
      "entropy": 5.472564268112182,
      "epoch": 3.8690948789202064,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035730726323991474,
      "loss": 5.0118,
      "mean_token_accuracy": 0.19690193086862565,
      "num_tokens": 85413446.0,
      "step": 38985
    },
    {
      "entropy": 5.379844379425049,
      "epoch": 3.869591107582374,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003572742864970913,
      "loss": 4.9489,
      "mean_token_accuracy": 0.18358099460601807,
      "num_tokens": 85423819.0,
      "step": 38990
    },
    {
      "entropy": 5.387105798721313,
      "epoch": 3.8700873362445414,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003572413077140427,
      "loss": 4.9284,
      "mean_token_accuracy": 0.19047790467739106,
      "num_tokens": 85435129.0,
      "step": 38995
    },
    {
      "entropy": 5.413461065292358,
      "epoch": 3.870583564906709,
      "grad_norm": 1.015625,
      "learning_rate": 0.000357208326891587,
      "loss": 4.9439,
      "mean_token_accuracy": 0.19088121503591537,
      "num_tokens": 85445845.0,
      "step": 39000
    },
    {
      "epoch": 3.870583564906709,
      "eval_entropy": 5.207119879062115,
      "eval_loss": 5.109464645385742,
      "eval_mean_token_accuracy": 0.1848997473592825,
      "eval_num_tokens": 85445845.0,
      "eval_runtime": 26.3159,
      "eval_samples_per_second": 1279.415,
      "eval_steps_per_second": 159.941,
      "step": 39000
    }
  ],
  "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": 1.43515073175552e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}