ita-latn-100mb-after-ppt-shuff-dyck-100mb-ckpt500_seed3407/checkpoint-33000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.2751091703056767,
  "eval_steps": 3000,
  "global_step": 33000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.8886785984039305,
      "epoch": 0.0004962286621675268,
      "grad_norm": 24.0,
      "learning_rate": 2e-06,
      "loss": 14.1768,
      "mean_token_accuracy": 0.0,
      "num_tokens": 11448.0,
      "step": 5
    },
    {
      "entropy": 4.918353462219239,
      "epoch": 0.0009924573243350536,
      "grad_norm": 27.125,
      "learning_rate": 4.5e-06,
      "loss": 14.1453,
      "mean_token_accuracy": 9.523809421807528e-05,
      "num_tokens": 22646.0,
      "step": 10
    },
    {
      "entropy": 5.069689464569092,
      "epoch": 0.0014886859865025804,
      "grad_norm": 36.75,
      "learning_rate": 7e-06,
      "loss": 13.8023,
      "mean_token_accuracy": 0.0,
      "num_tokens": 34011.0,
      "step": 15
    },
    {
      "entropy": 5.5015480518341064,
      "epoch": 0.001984914648670107,
      "grad_norm": 42.75,
      "learning_rate": 9.5e-06,
      "loss": 13.2381,
      "mean_token_accuracy": 0.0,
      "num_tokens": 44566.0,
      "step": 20
    },
    {
      "entropy": 7.525242757797241,
      "epoch": 0.002481143310837634,
      "grad_norm": 18.875,
      "learning_rate": 1.2e-05,
      "loss": 11.959,
      "mean_token_accuracy": 0.0,
      "num_tokens": 56274.0,
      "step": 25
    },
    {
      "entropy": 9.429654407501221,
      "epoch": 0.0029773719730051607,
      "grad_norm": 5.25,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 11.1919,
      "mean_token_accuracy": 0.0,
      "num_tokens": 66584.0,
      "step": 30
    },
    {
      "entropy": 10.561602783203124,
      "epoch": 0.0034736006351726877,
      "grad_norm": 2.40625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.7394,
      "mean_token_accuracy": 0.0,
      "num_tokens": 76948.0,
      "step": 35
    },
    {
      "entropy": 10.702620887756348,
      "epoch": 0.003969829297340214,
      "grad_norm": 2.3125,
      "learning_rate": 1.95e-05,
      "loss": 10.496,
      "mean_token_accuracy": 0.012860012706369162,
      "num_tokens": 87253.0,
      "step": 40
    },
    {
      "entropy": 10.697653579711915,
      "epoch": 0.004466057959507741,
      "grad_norm": 2.46875,
      "learning_rate": 2.2e-05,
      "loss": 10.1993,
      "mean_token_accuracy": 0.04175656381994486,
      "num_tokens": 98273.0,
      "step": 45
    },
    {
      "entropy": 10.658453273773194,
      "epoch": 0.004962286621675268,
      "grad_norm": 2.15625,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.9063,
      "mean_token_accuracy": 0.04210629928857088,
      "num_tokens": 107788.0,
      "step": 50
    },
    {
      "entropy": 10.590508460998535,
      "epoch": 0.0054585152838427945,
      "grad_norm": 2.078125,
      "learning_rate": 2.7e-05,
      "loss": 9.7831,
      "mean_token_accuracy": 0.03733267150819301,
      "num_tokens": 118632.0,
      "step": 55
    },
    {
      "entropy": 10.57080307006836,
      "epoch": 0.0059547439460103215,
      "grad_norm": 2.046875,
      "learning_rate": 2.95e-05,
      "loss": 9.6152,
      "mean_token_accuracy": 0.04101524651050568,
      "num_tokens": 128649.0,
      "step": 60
    },
    {
      "entropy": 10.565433502197266,
      "epoch": 0.0064509726081778485,
      "grad_norm": 1.6953125,
      "learning_rate": 3.2e-05,
      "loss": 9.5678,
      "mean_token_accuracy": 0.04273284543305635,
      "num_tokens": 138958.0,
      "step": 65
    },
    {
      "entropy": 10.547395515441895,
      "epoch": 0.0069472012703453755,
      "grad_norm": 1.828125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.5162,
      "mean_token_accuracy": 0.04243846274912357,
      "num_tokens": 151445.0,
      "step": 70
    },
    {
      "entropy": 10.541783046722411,
      "epoch": 0.007443429932512902,
      "grad_norm": 1.703125,
      "learning_rate": 3.7e-05,
      "loss": 9.4272,
      "mean_token_accuracy": 0.043674266710877416,
      "num_tokens": 162239.0,
      "step": 75
    },
    {
      "entropy": 10.524649047851563,
      "epoch": 0.007939658594680429,
      "grad_norm": 1.84375,
      "learning_rate": 3.95e-05,
      "loss": 9.3667,
      "mean_token_accuracy": 0.03882305920124054,
      "num_tokens": 173261.0,
      "step": 80
    },
    {
      "entropy": 10.50841827392578,
      "epoch": 0.008435887256847956,
      "grad_norm": 1.8125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.2444,
      "mean_token_accuracy": 0.037595166079699996,
      "num_tokens": 183857.0,
      "step": 85
    },
    {
      "entropy": 10.483632850646973,
      "epoch": 0.008932115919015483,
      "grad_norm": 1.6484375,
      "learning_rate": 4.45e-05,
      "loss": 9.2664,
      "mean_token_accuracy": 0.03864660933613777,
      "num_tokens": 196635.0,
      "step": 90
    },
    {
      "entropy": 10.438802337646484,
      "epoch": 0.00942834458118301,
      "grad_norm": 1.6875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.079,
      "mean_token_accuracy": 0.04349393919110298,
      "num_tokens": 208700.0,
      "step": 95
    },
    {
      "entropy": 10.382994365692138,
      "epoch": 0.009924573243350537,
      "grad_norm": 1.7578125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.9171,
      "mean_token_accuracy": 0.04082345124334097,
      "num_tokens": 218915.0,
      "step": 100
    },
    {
      "entropy": 10.337403869628906,
      "epoch": 0.010420801905518062,
      "grad_norm": 1.65625,
      "learning_rate": 5.2e-05,
      "loss": 8.7812,
      "mean_token_accuracy": 0.051039114221930505,
      "num_tokens": 230211.0,
      "step": 105
    },
    {
      "entropy": 10.211695384979247,
      "epoch": 0.010917030567685589,
      "grad_norm": 1.6015625,
      "learning_rate": 5.45e-05,
      "loss": 8.6614,
      "mean_token_accuracy": 0.05333258844912052,
      "num_tokens": 241174.0,
      "step": 110
    },
    {
      "entropy": 10.04779806137085,
      "epoch": 0.011413259229853116,
      "grad_norm": 1.4140625,
      "learning_rate": 5.7e-05,
      "loss": 8.5665,
      "mean_token_accuracy": 0.047243202850222585,
      "num_tokens": 251708.0,
      "step": 115
    },
    {
      "entropy": 9.996843624114991,
      "epoch": 0.011909487892020643,
      "grad_norm": 1.2421875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.585,
      "mean_token_accuracy": 0.04537748880684376,
      "num_tokens": 263287.0,
      "step": 120
    },
    {
      "entropy": 9.867857456207275,
      "epoch": 0.01240571655418817,
      "grad_norm": 1.1953125,
      "learning_rate": 6.2e-05,
      "loss": 8.3763,
      "mean_token_accuracy": 0.0451455045491457,
      "num_tokens": 275179.0,
      "step": 125
    },
    {
      "entropy": 9.665265655517578,
      "epoch": 0.012901945216355697,
      "grad_norm": 1.2421875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.1667,
      "mean_token_accuracy": 0.04741326570510864,
      "num_tokens": 287223.0,
      "step": 130
    },
    {
      "entropy": 9.438035774230958,
      "epoch": 0.013398173878523224,
      "grad_norm": 1.1015625,
      "learning_rate": 6.7e-05,
      "loss": 8.0475,
      "mean_token_accuracy": 0.04950449950993061,
      "num_tokens": 297868.0,
      "step": 135
    },
    {
      "entropy": 9.164513778686523,
      "epoch": 0.013894402540690751,
      "grad_norm": 0.83984375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.0005,
      "mean_token_accuracy": 0.05104476250708103,
      "num_tokens": 309702.0,
      "step": 140
    },
    {
      "entropy": 8.994746875762939,
      "epoch": 0.014390631202858276,
      "grad_norm": 0.87890625,
      "learning_rate": 7.2e-05,
      "loss": 8.0216,
      "mean_token_accuracy": 0.048004313558340075,
      "num_tokens": 321613.0,
      "step": 145
    },
    {
      "entropy": 8.776610565185546,
      "epoch": 0.014886859865025803,
      "grad_norm": 0.9375,
      "learning_rate": 7.45e-05,
      "loss": 7.8727,
      "mean_token_accuracy": 0.051788559183478355,
      "num_tokens": 332668.0,
      "step": 150
    },
    {
      "entropy": 8.660905170440675,
      "epoch": 0.01538308852719333,
      "grad_norm": 0.8828125,
      "learning_rate": 7.7e-05,
      "loss": 7.8823,
      "mean_token_accuracy": 0.049248311296105386,
      "num_tokens": 342579.0,
      "step": 155
    },
    {
      "entropy": 8.499152851104736,
      "epoch": 0.015879317189360857,
      "grad_norm": 0.98046875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.759,
      "mean_token_accuracy": 0.0496104896068573,
      "num_tokens": 352271.0,
      "step": 160
    },
    {
      "entropy": 8.3534499168396,
      "epoch": 0.016375545851528384,
      "grad_norm": 1.0078125,
      "learning_rate": 8.2e-05,
      "loss": 7.7938,
      "mean_token_accuracy": 0.061972716078162196,
      "num_tokens": 364033.0,
      "step": 165
    },
    {
      "entropy": 8.28783130645752,
      "epoch": 0.01687177451369591,
      "grad_norm": 1.0078125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.7301,
      "mean_token_accuracy": 0.05461362563073635,
      "num_tokens": 373788.0,
      "step": 170
    },
    {
      "entropy": 8.24440793991089,
      "epoch": 0.01736800317586344,
      "grad_norm": 1.1875,
      "learning_rate": 8.7e-05,
      "loss": 7.7277,
      "mean_token_accuracy": 0.057403456419706345,
      "num_tokens": 384883.0,
      "step": 175
    },
    {
      "entropy": 8.245327854156494,
      "epoch": 0.017864231838030965,
      "grad_norm": 1.0625,
      "learning_rate": 8.95e-05,
      "loss": 7.6288,
      "mean_token_accuracy": 0.06058619767427444,
      "num_tokens": 396044.0,
      "step": 180
    },
    {
      "entropy": 8.175503873825074,
      "epoch": 0.018360460500198492,
      "grad_norm": 1.359375,
      "learning_rate": 9.2e-05,
      "loss": 7.733,
      "mean_token_accuracy": 0.06297681033611298,
      "num_tokens": 406995.0,
      "step": 185
    },
    {
      "entropy": 8.033704376220703,
      "epoch": 0.01885668916236602,
      "grad_norm": 0.953125,
      "learning_rate": 9.45e-05,
      "loss": 7.5683,
      "mean_token_accuracy": 0.0688964381814003,
      "num_tokens": 417238.0,
      "step": 190
    },
    {
      "entropy": 8.177022457122803,
      "epoch": 0.019352917824533546,
      "grad_norm": 1.109375,
      "learning_rate": 9.7e-05,
      "loss": 7.645,
      "mean_token_accuracy": 0.06311573572456837,
      "num_tokens": 427263.0,
      "step": 195
    },
    {
      "entropy": 8.155741834640503,
      "epoch": 0.019849146486701073,
      "grad_norm": 1.015625,
      "learning_rate": 9.95e-05,
      "loss": 7.5372,
      "mean_token_accuracy": 0.0640363771468401,
      "num_tokens": 438357.0,
      "step": 200
    },
    {
      "entropy": 8.039666271209716,
      "epoch": 0.0203453751488686,
      "grad_norm": 1.25,
      "learning_rate": 0.000102,
      "loss": 7.5516,
      "mean_token_accuracy": 0.06006042025983334,
      "num_tokens": 448738.0,
      "step": 205
    },
    {
      "entropy": 7.96295280456543,
      "epoch": 0.020841603811036124,
      "grad_norm": 0.984375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.5624,
      "mean_token_accuracy": 0.07530630342662334,
      "num_tokens": 457873.0,
      "step": 210
    },
    {
      "entropy": 8.063592433929443,
      "epoch": 0.02133783247320365,
      "grad_norm": 1.046875,
      "learning_rate": 0.000107,
      "loss": 7.4987,
      "mean_token_accuracy": 0.06970696784555912,
      "num_tokens": 468640.0,
      "step": 215
    },
    {
      "entropy": 8.009668874740601,
      "epoch": 0.021834061135371178,
      "grad_norm": 1.4375,
      "learning_rate": 0.0001095,
      "loss": 7.5142,
      "mean_token_accuracy": 0.07091238908469677,
      "num_tokens": 479922.0,
      "step": 220
    },
    {
      "entropy": 7.893428993225098,
      "epoch": 0.022330289797538705,
      "grad_norm": 0.921875,
      "learning_rate": 0.000112,
      "loss": 7.4717,
      "mean_token_accuracy": 0.06831742338836193,
      "num_tokens": 491056.0,
      "step": 225
    },
    {
      "entropy": 7.929876375198364,
      "epoch": 0.022826518459706232,
      "grad_norm": 1.0625,
      "learning_rate": 0.0001145,
      "loss": 7.5665,
      "mean_token_accuracy": 0.06847823970019817,
      "num_tokens": 504365.0,
      "step": 230
    },
    {
      "entropy": 7.870587635040283,
      "epoch": 0.02332274712187376,
      "grad_norm": 1.28125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.437,
      "mean_token_accuracy": 0.06928115114569663,
      "num_tokens": 514347.0,
      "step": 235
    },
    {
      "entropy": 7.846385335922241,
      "epoch": 0.023818975784041286,
      "grad_norm": 1.46875,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.4702,
      "mean_token_accuracy": 0.06925889551639557,
      "num_tokens": 525912.0,
      "step": 240
    },
    {
      "entropy": 7.909278726577758,
      "epoch": 0.024315204446208813,
      "grad_norm": 1.03125,
      "learning_rate": 0.000122,
      "loss": 7.3455,
      "mean_token_accuracy": 0.07115834765136242,
      "num_tokens": 535566.0,
      "step": 245
    },
    {
      "entropy": 7.8972766399383545,
      "epoch": 0.02481143310837634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0001245,
      "loss": 7.3889,
      "mean_token_accuracy": 0.07348600402474403,
      "num_tokens": 546474.0,
      "step": 250
    },
    {
      "entropy": 7.799334192276001,
      "epoch": 0.025307661770543867,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000127,
      "loss": 7.4794,
      "mean_token_accuracy": 0.06563120521605015,
      "num_tokens": 558132.0,
      "step": 255
    },
    {
      "entropy": 7.913891267776489,
      "epoch": 0.025803890432711394,
      "grad_norm": 1.265625,
      "learning_rate": 0.0001295,
      "loss": 7.4623,
      "mean_token_accuracy": 0.06633393727242946,
      "num_tokens": 568309.0,
      "step": 260
    },
    {
      "entropy": 7.81136999130249,
      "epoch": 0.02630011909487892,
      "grad_norm": 1.125,
      "learning_rate": 0.000132,
      "loss": 7.372,
      "mean_token_accuracy": 0.06931877210736274,
      "num_tokens": 579701.0,
      "step": 265
    },
    {
      "entropy": 7.7742400646209715,
      "epoch": 0.026796347757046448,
      "grad_norm": 0.953125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.33,
      "mean_token_accuracy": 0.07372162342071534,
      "num_tokens": 590832.0,
      "step": 270
    },
    {
      "entropy": 7.759802675247192,
      "epoch": 0.027292576419213975,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.274,
      "mean_token_accuracy": 0.07079541869461536,
      "num_tokens": 601094.0,
      "step": 275
    },
    {
      "entropy": 7.690725088119507,
      "epoch": 0.027788805081381502,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001395,
      "loss": 7.2879,
      "mean_token_accuracy": 0.07795346416532993,
      "num_tokens": 612027.0,
      "step": 280
    },
    {
      "entropy": 7.692393255233765,
      "epoch": 0.02828503374354903,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.33,
      "mean_token_accuracy": 0.07078218050301074,
      "num_tokens": 623698.0,
      "step": 285
    },
    {
      "entropy": 7.724156379699707,
      "epoch": 0.028781262405716553,
      "grad_norm": 0.828125,
      "learning_rate": 0.0001445,
      "loss": 7.3473,
      "mean_token_accuracy": 0.07386260889470578,
      "num_tokens": 635631.0,
      "step": 290
    },
    {
      "entropy": 7.633536434173584,
      "epoch": 0.02927749106788408,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000147,
      "loss": 7.3238,
      "mean_token_accuracy": 0.07232154980301857,
      "num_tokens": 647663.0,
      "step": 295
    },
    {
      "entropy": 7.68490948677063,
      "epoch": 0.029773719730051607,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001495,
      "loss": 7.2869,
      "mean_token_accuracy": 0.0713068351149559,
      "num_tokens": 659352.0,
      "step": 300
    },
    {
      "entropy": 7.628432846069336,
      "epoch": 0.030269948392219134,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000152,
      "loss": 7.2491,
      "mean_token_accuracy": 0.07672590613365174,
      "num_tokens": 669379.0,
      "step": 305
    },
    {
      "entropy": 7.652841281890869,
      "epoch": 0.03076617705438666,
      "grad_norm": 1.203125,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.241,
      "mean_token_accuracy": 0.07731729820370674,
      "num_tokens": 680144.0,
      "step": 310
    },
    {
      "entropy": 7.682145023345948,
      "epoch": 0.03126240571655419,
      "grad_norm": 1.53125,
      "learning_rate": 0.000157,
      "loss": 7.1947,
      "mean_token_accuracy": 0.06403898783028125,
      "num_tokens": 691165.0,
      "step": 315
    },
    {
      "entropy": 7.572655248641968,
      "epoch": 0.031758634378721715,
      "grad_norm": 1.109375,
      "learning_rate": 0.0001595,
      "loss": 7.2853,
      "mean_token_accuracy": 0.07224688902497292,
      "num_tokens": 703767.0,
      "step": 320
    },
    {
      "entropy": 7.581577396392822,
      "epoch": 0.032254863040889245,
      "grad_norm": 1.46875,
      "learning_rate": 0.000162,
      "loss": 7.2793,
      "mean_token_accuracy": 0.07694573625922203,
      "num_tokens": 714298.0,
      "step": 325
    },
    {
      "entropy": 7.558791637420654,
      "epoch": 0.03275109170305677,
      "grad_norm": 1.25,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.2543,
      "mean_token_accuracy": 0.08345745950937271,
      "num_tokens": 724135.0,
      "step": 330
    },
    {
      "entropy": 7.579706907272339,
      "epoch": 0.03324732036522429,
      "grad_norm": 0.96875,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.3114,
      "mean_token_accuracy": 0.07194749340415,
      "num_tokens": 735050.0,
      "step": 335
    },
    {
      "entropy": 7.66069598197937,
      "epoch": 0.03374354902739182,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.2156,
      "mean_token_accuracy": 0.07318415641784667,
      "num_tokens": 746073.0,
      "step": 340
    },
    {
      "entropy": 7.529136419296265,
      "epoch": 0.034239777689559346,
      "grad_norm": 1.078125,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.2125,
      "mean_token_accuracy": 0.0734445221722126,
      "num_tokens": 758998.0,
      "step": 345
    },
    {
      "entropy": 7.582425689697265,
      "epoch": 0.03473600635172688,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.1518,
      "mean_token_accuracy": 0.07465869188308716,
      "num_tokens": 771608.0,
      "step": 350
    },
    {
      "entropy": 7.522878408432007,
      "epoch": 0.0352322350138944,
      "grad_norm": 0.984375,
      "learning_rate": 0.000177,
      "loss": 7.2138,
      "mean_token_accuracy": 0.07425806559622287,
      "num_tokens": 782394.0,
      "step": 355
    },
    {
      "entropy": 7.542528247833252,
      "epoch": 0.03572846367606193,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001795,
      "loss": 7.1216,
      "mean_token_accuracy": 0.07903320342302322,
      "num_tokens": 792389.0,
      "step": 360
    },
    {
      "entropy": 7.50439190864563,
      "epoch": 0.036224692338229454,
      "grad_norm": 1.59375,
      "learning_rate": 0.000182,
      "loss": 7.1662,
      "mean_token_accuracy": 0.07648928835988045,
      "num_tokens": 803219.0,
      "step": 365
    },
    {
      "entropy": 7.523057889938355,
      "epoch": 0.036720921000396985,
      "grad_norm": 1.078125,
      "learning_rate": 0.0001845,
      "loss": 7.235,
      "mean_token_accuracy": 0.07408236041665077,
      "num_tokens": 814249.0,
      "step": 370
    },
    {
      "entropy": 7.520695543289184,
      "epoch": 0.03721714966256451,
      "grad_norm": 1.09375,
      "learning_rate": 0.000187,
      "loss": 7.1213,
      "mean_token_accuracy": 0.08493848741054535,
      "num_tokens": 825138.0,
      "step": 375
    },
    {
      "entropy": 7.437196350097656,
      "epoch": 0.03771337832473204,
      "grad_norm": 1.125,
      "learning_rate": 0.0001895,
      "loss": 7.1724,
      "mean_token_accuracy": 0.07618945986032485,
      "num_tokens": 836024.0,
      "step": 380
    },
    {
      "entropy": 7.505164337158203,
      "epoch": 0.03820960698689956,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000192,
      "loss": 7.1619,
      "mean_token_accuracy": 0.07474682033061981,
      "num_tokens": 847765.0,
      "step": 385
    },
    {
      "entropy": 7.485037040710449,
      "epoch": 0.03870583564906709,
      "grad_norm": 1.296875,
      "learning_rate": 0.0001945,
      "loss": 7.1428,
      "mean_token_accuracy": 0.07660850211977958,
      "num_tokens": 859980.0,
      "step": 390
    },
    {
      "entropy": 7.523992443084717,
      "epoch": 0.039202064311234616,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.1394,
      "mean_token_accuracy": 0.07903257459402084,
      "num_tokens": 871514.0,
      "step": 395
    },
    {
      "entropy": 7.4822288990020756,
      "epoch": 0.03969829297340215,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.1353,
      "mean_token_accuracy": 0.07928794771432876,
      "num_tokens": 881984.0,
      "step": 400
    },
    {
      "entropy": 7.439731025695801,
      "epoch": 0.04019452163556967,
      "grad_norm": 1.09375,
      "learning_rate": 0.000202,
      "loss": 7.098,
      "mean_token_accuracy": 0.08132969811558724,
      "num_tokens": 894198.0,
      "step": 405
    },
    {
      "entropy": 7.445980262756348,
      "epoch": 0.0406907502977372,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.1122,
      "mean_token_accuracy": 0.08235671296715737,
      "num_tokens": 905394.0,
      "step": 410
    },
    {
      "entropy": 7.525100374221802,
      "epoch": 0.041186978959904724,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000207,
      "loss": 7.1626,
      "mean_token_accuracy": 0.07699285373091698,
      "num_tokens": 917010.0,
      "step": 415
    },
    {
      "entropy": 7.316656255722046,
      "epoch": 0.04168320762207225,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002095,
      "loss": 7.0005,
      "mean_token_accuracy": 0.08732899129390717,
      "num_tokens": 926761.0,
      "step": 420
    },
    {
      "entropy": 7.296533203125,
      "epoch": 0.04217943628423978,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000212,
      "loss": 7.0074,
      "mean_token_accuracy": 0.08514706641435624,
      "num_tokens": 938385.0,
      "step": 425
    },
    {
      "entropy": 7.430690574645996,
      "epoch": 0.0426756649464073,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002145,
      "loss": 7.0374,
      "mean_token_accuracy": 0.08442675322294235,
      "num_tokens": 949709.0,
      "step": 430
    },
    {
      "entropy": 7.3450085639953615,
      "epoch": 0.04317189360857483,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.0699,
      "mean_token_accuracy": 0.07798392251133919,
      "num_tokens": 960837.0,
      "step": 435
    },
    {
      "entropy": 7.343694639205933,
      "epoch": 0.043668122270742356,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002195,
      "loss": 6.903,
      "mean_token_accuracy": 0.08706051260232925,
      "num_tokens": 971036.0,
      "step": 440
    },
    {
      "entropy": 7.371674299240112,
      "epoch": 0.044164350932909886,
      "grad_norm": 0.984375,
      "learning_rate": 0.000222,
      "loss": 7.0502,
      "mean_token_accuracy": 0.0812618188560009,
      "num_tokens": 983057.0,
      "step": 445
    },
    {
      "entropy": 7.4037517547607425,
      "epoch": 0.04466057959507741,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002245,
      "loss": 7.0069,
      "mean_token_accuracy": 0.08837616667151452,
      "num_tokens": 994303.0,
      "step": 450
    },
    {
      "entropy": 7.2432009220123295,
      "epoch": 0.04515680825724494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.0026,
      "mean_token_accuracy": 0.0851232349872589,
      "num_tokens": 1005033.0,
      "step": 455
    },
    {
      "entropy": 7.39683952331543,
      "epoch": 0.045653036919412464,
      "grad_norm": 0.90625,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.0681,
      "mean_token_accuracy": 0.08551012352108955,
      "num_tokens": 1015940.0,
      "step": 460
    },
    {
      "entropy": 7.263912868499756,
      "epoch": 0.046149265581579994,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.0144,
      "mean_token_accuracy": 0.08879487365484237,
      "num_tokens": 1027582.0,
      "step": 465
    },
    {
      "entropy": 7.333393192291259,
      "epoch": 0.04664549424374752,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.9848,
      "mean_token_accuracy": 0.08102063313126565,
      "num_tokens": 1040046.0,
      "step": 470
    },
    {
      "entropy": 7.31667799949646,
      "epoch": 0.04714172290591505,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000237,
      "loss": 7.0628,
      "mean_token_accuracy": 0.07443273328244686,
      "num_tokens": 1050636.0,
      "step": 475
    },
    {
      "entropy": 7.372964000701904,
      "epoch": 0.04763795156808257,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002395,
      "loss": 6.9388,
      "mean_token_accuracy": 0.08691781386733055,
      "num_tokens": 1060748.0,
      "step": 480
    },
    {
      "entropy": 7.3284331321716305,
      "epoch": 0.0481341802302501,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000242,
      "loss": 6.9896,
      "mean_token_accuracy": 0.07861718982458114,
      "num_tokens": 1071940.0,
      "step": 485
    },
    {
      "entropy": 7.240336561203003,
      "epoch": 0.048630408892417626,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002445,
      "loss": 6.995,
      "mean_token_accuracy": 0.08761729970574379,
      "num_tokens": 1082962.0,
      "step": 490
    },
    {
      "entropy": 7.282444810867309,
      "epoch": 0.04912663755458515,
      "grad_norm": 1.125,
      "learning_rate": 0.000247,
      "loss": 6.9754,
      "mean_token_accuracy": 0.09364678412675857,
      "num_tokens": 1093082.0,
      "step": 495
    },
    {
      "entropy": 7.348704719543457,
      "epoch": 0.04962286621675268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002495,
      "loss": 7.049,
      "mean_token_accuracy": 0.08253686502575874,
      "num_tokens": 1103528.0,
      "step": 500
    },
    {
      "entropy": 7.214434576034546,
      "epoch": 0.0501190948789202,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000252,
      "loss": 6.9393,
      "mean_token_accuracy": 0.08427478075027466,
      "num_tokens": 1114813.0,
      "step": 505
    },
    {
      "entropy": 7.229425621032715,
      "epoch": 0.050615323541087734,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002545,
      "loss": 6.9222,
      "mean_token_accuracy": 0.08125569894909859,
      "num_tokens": 1125345.0,
      "step": 510
    },
    {
      "entropy": 7.092966032028198,
      "epoch": 0.05111155220325526,
      "grad_norm": 1.09375,
      "learning_rate": 0.000257,
      "loss": 6.9414,
      "mean_token_accuracy": 0.08359554558992385,
      "num_tokens": 1134985.0,
      "step": 515
    },
    {
      "entropy": 7.2688844203948975,
      "epoch": 0.05160778086542279,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002595,
      "loss": 6.9457,
      "mean_token_accuracy": 0.08630074635148048,
      "num_tokens": 1145796.0,
      "step": 520
    },
    {
      "entropy": 7.270513343811035,
      "epoch": 0.05210400952759031,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000262,
      "loss": 7.0875,
      "mean_token_accuracy": 0.07355156876146793,
      "num_tokens": 1156936.0,
      "step": 525
    },
    {
      "entropy": 7.258271265029907,
      "epoch": 0.05260023818975784,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.96,
      "mean_token_accuracy": 0.08982882723212242,
      "num_tokens": 1167526.0,
      "step": 530
    },
    {
      "entropy": 7.198262119293213,
      "epoch": 0.053096466851925365,
      "grad_norm": 1.625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.9928,
      "mean_token_accuracy": 0.08096232041716575,
      "num_tokens": 1177647.0,
      "step": 535
    },
    {
      "entropy": 7.275991725921631,
      "epoch": 0.053592695514092896,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.9641,
      "mean_token_accuracy": 0.08726390451192856,
      "num_tokens": 1188255.0,
      "step": 540
    },
    {
      "entropy": 7.114196634292602,
      "epoch": 0.05408892417626042,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.8869,
      "mean_token_accuracy": 0.09180611073970794,
      "num_tokens": 1199433.0,
      "step": 545
    },
    {
      "entropy": 7.082750701904297,
      "epoch": 0.05458515283842795,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002745,
      "loss": 6.9654,
      "mean_token_accuracy": 0.08378833159804344,
      "num_tokens": 1210067.0,
      "step": 550
    },
    {
      "entropy": 7.160634613037109,
      "epoch": 0.055081381500595473,
      "grad_norm": 1.234375,
      "learning_rate": 0.000277,
      "loss": 6.8504,
      "mean_token_accuracy": 0.08845426812767983,
      "num_tokens": 1220639.0,
      "step": 555
    },
    {
      "entropy": 7.191676330566406,
      "epoch": 0.055577610162763004,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002795,
      "loss": 6.8845,
      "mean_token_accuracy": 0.09016633853316307,
      "num_tokens": 1233158.0,
      "step": 560
    },
    {
      "entropy": 7.2028326988220215,
      "epoch": 0.05607383882493053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.8726,
      "mean_token_accuracy": 0.08301621749997139,
      "num_tokens": 1244552.0,
      "step": 565
    },
    {
      "entropy": 7.165884399414063,
      "epoch": 0.05657006748709806,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002845,
      "loss": 6.9698,
      "mean_token_accuracy": 0.0811222843825817,
      "num_tokens": 1255478.0,
      "step": 570
    },
    {
      "entropy": 7.039954090118409,
      "epoch": 0.05706629614926558,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000287,
      "loss": 6.7999,
      "mean_token_accuracy": 0.09229947701096534,
      "num_tokens": 1267507.0,
      "step": 575
    },
    {
      "entropy": 7.070255470275879,
      "epoch": 0.057562524811433105,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002895,
      "loss": 6.8057,
      "mean_token_accuracy": 0.0856806755065918,
      "num_tokens": 1279377.0,
      "step": 580
    },
    {
      "entropy": 7.210440063476563,
      "epoch": 0.058058753473600636,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000292,
      "loss": 6.8795,
      "mean_token_accuracy": 0.09145521745085716,
      "num_tokens": 1290671.0,
      "step": 585
    },
    {
      "entropy": 6.9877747058868405,
      "epoch": 0.05855498213576816,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002945,
      "loss": 6.8592,
      "mean_token_accuracy": 0.0912296824157238,
      "num_tokens": 1302792.0,
      "step": 590
    },
    {
      "entropy": 7.064874172210693,
      "epoch": 0.05905121079793569,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000297,
      "loss": 6.8074,
      "mean_token_accuracy": 0.08353102207183838,
      "num_tokens": 1313315.0,
      "step": 595
    },
    {
      "entropy": 7.0152363777160645,
      "epoch": 0.05954743946010321,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002995,
      "loss": 6.881,
      "mean_token_accuracy": 0.0832609310746193,
      "num_tokens": 1326358.0,
      "step": 600
    },
    {
      "entropy": 7.107955121994019,
      "epoch": 0.060043668122270744,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000302,
      "loss": 6.7556,
      "mean_token_accuracy": 0.0891527384519577,
      "num_tokens": 1336266.0,
      "step": 605
    },
    {
      "entropy": 6.989856767654419,
      "epoch": 0.06053989678443827,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003045,
      "loss": 6.7067,
      "mean_token_accuracy": 0.09343992173671722,
      "num_tokens": 1347610.0,
      "step": 610
    },
    {
      "entropy": 6.953636312484742,
      "epoch": 0.0610361254466058,
      "grad_norm": 1.109375,
      "learning_rate": 0.000307,
      "loss": 6.676,
      "mean_token_accuracy": 0.09296446591615677,
      "num_tokens": 1358317.0,
      "step": 615
    },
    {
      "entropy": 6.972737407684326,
      "epoch": 0.06153235410877332,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003095,
      "loss": 6.7807,
      "mean_token_accuracy": 0.08731090128421784,
      "num_tokens": 1370126.0,
      "step": 620
    },
    {
      "entropy": 7.019057750701904,
      "epoch": 0.06202858277094085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000312,
      "loss": 6.7463,
      "mean_token_accuracy": 0.09216577932238579,
      "num_tokens": 1381571.0,
      "step": 625
    },
    {
      "entropy": 6.846717262268067,
      "epoch": 0.06252481143310838,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003145,
      "loss": 6.773,
      "mean_token_accuracy": 0.09080302193760872,
      "num_tokens": 1392486.0,
      "step": 630
    },
    {
      "entropy": 7.055567216873169,
      "epoch": 0.0630210400952759,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000317,
      "loss": 6.7758,
      "mean_token_accuracy": 0.09265405684709549,
      "num_tokens": 1403086.0,
      "step": 635
    },
    {
      "entropy": 6.926388502120972,
      "epoch": 0.06351726875744343,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003195,
      "loss": 6.7449,
      "mean_token_accuracy": 0.09237656593322754,
      "num_tokens": 1414537.0,
      "step": 640
    },
    {
      "entropy": 6.949403572082519,
      "epoch": 0.06401349741961096,
      "grad_norm": 1.125,
      "learning_rate": 0.000322,
      "loss": 6.7572,
      "mean_token_accuracy": 0.09533382952213287,
      "num_tokens": 1426998.0,
      "step": 645
    },
    {
      "entropy": 6.947071886062622,
      "epoch": 0.06450972608177849,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.7806,
      "mean_token_accuracy": 0.09227192178368568,
      "num_tokens": 1437667.0,
      "step": 650
    },
    {
      "entropy": 6.933514261245728,
      "epoch": 0.065005954743946,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.652,
      "mean_token_accuracy": 0.09583644419908524,
      "num_tokens": 1448280.0,
      "step": 655
    },
    {
      "entropy": 6.916765022277832,
      "epoch": 0.06550218340611354,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.7201,
      "mean_token_accuracy": 0.09147566556930542,
      "num_tokens": 1458608.0,
      "step": 660
    },
    {
      "entropy": 6.95277361869812,
      "epoch": 0.06599841206828107,
      "grad_norm": 1.484375,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.9062,
      "mean_token_accuracy": 0.08369888290762902,
      "num_tokens": 1470435.0,
      "step": 665
    },
    {
      "entropy": 6.9358902931213375,
      "epoch": 0.06649464073044858,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.7515,
      "mean_token_accuracy": 0.0901063933968544,
      "num_tokens": 1481959.0,
      "step": 670
    },
    {
      "entropy": 6.907770204544067,
      "epoch": 0.06699086939261611,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000337,
      "loss": 6.7936,
      "mean_token_accuracy": 0.08672575987875461,
      "num_tokens": 1493798.0,
      "step": 675
    },
    {
      "entropy": 6.924069547653199,
      "epoch": 0.06748709805478365,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003395,
      "loss": 6.6036,
      "mean_token_accuracy": 0.09497400969266892,
      "num_tokens": 1502969.0,
      "step": 680
    },
    {
      "entropy": 6.881195592880249,
      "epoch": 0.06798332671695118,
      "grad_norm": 1.109375,
      "learning_rate": 0.000342,
      "loss": 6.7521,
      "mean_token_accuracy": 0.09069281220436096,
      "num_tokens": 1513825.0,
      "step": 685
    },
    {
      "entropy": 6.85332555770874,
      "epoch": 0.06847955537911869,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.6429,
      "mean_token_accuracy": 0.0923303596675396,
      "num_tokens": 1525328.0,
      "step": 690
    },
    {
      "entropy": 6.871306228637695,
      "epoch": 0.06897578404128622,
      "grad_norm": 0.96875,
      "learning_rate": 0.000347,
      "loss": 6.6631,
      "mean_token_accuracy": 0.09412776604294777,
      "num_tokens": 1537383.0,
      "step": 695
    },
    {
      "entropy": 6.971198177337646,
      "epoch": 0.06947201270345375,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003495,
      "loss": 6.7733,
      "mean_token_accuracy": 0.09241362139582635,
      "num_tokens": 1549671.0,
      "step": 700
    },
    {
      "entropy": 6.766777658462525,
      "epoch": 0.06996824136562128,
      "grad_norm": 1.046875,
      "learning_rate": 0.000352,
      "loss": 6.6454,
      "mean_token_accuracy": 0.09111724868416786,
      "num_tokens": 1559974.0,
      "step": 705
    },
    {
      "entropy": 6.925774097442627,
      "epoch": 0.0704644700277888,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003545,
      "loss": 6.7489,
      "mean_token_accuracy": 0.08958815187215804,
      "num_tokens": 1570890.0,
      "step": 710
    },
    {
      "entropy": 6.800432395935059,
      "epoch": 0.07096069868995633,
      "grad_norm": 1.140625,
      "learning_rate": 0.000357,
      "loss": 6.6629,
      "mean_token_accuracy": 0.09856393039226533,
      "num_tokens": 1581434.0,
      "step": 715
    },
    {
      "entropy": 6.8169127941131595,
      "epoch": 0.07145692735212386,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003595,
      "loss": 6.7193,
      "mean_token_accuracy": 0.09074361547827721,
      "num_tokens": 1593710.0,
      "step": 720
    },
    {
      "entropy": 6.93365249633789,
      "epoch": 0.07195315601429139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000362,
      "loss": 6.6931,
      "mean_token_accuracy": 0.09370483383536339,
      "num_tokens": 1604635.0,
      "step": 725
    },
    {
      "entropy": 6.909399318695068,
      "epoch": 0.07244938467645891,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003645,
      "loss": 6.7543,
      "mean_token_accuracy": 0.09272385239601136,
      "num_tokens": 1615258.0,
      "step": 730
    },
    {
      "entropy": 6.855506086349488,
      "epoch": 0.07294561333862644,
      "grad_norm": 1.015625,
      "learning_rate": 0.000367,
      "loss": 6.6587,
      "mean_token_accuracy": 0.0944454848766327,
      "num_tokens": 1625942.0,
      "step": 735
    },
    {
      "entropy": 6.828038263320923,
      "epoch": 0.07344184200079397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003695,
      "loss": 6.6292,
      "mean_token_accuracy": 0.10174066349864005,
      "num_tokens": 1636270.0,
      "step": 740
    },
    {
      "entropy": 6.729400539398194,
      "epoch": 0.07393807066296149,
      "grad_norm": 1.0625,
      "learning_rate": 0.000372,
      "loss": 6.6152,
      "mean_token_accuracy": 0.0955189272761345,
      "num_tokens": 1647029.0,
      "step": 745
    },
    {
      "entropy": 6.830958700180053,
      "epoch": 0.07443429932512902,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003745,
      "loss": 6.7986,
      "mean_token_accuracy": 0.0909816287457943,
      "num_tokens": 1659483.0,
      "step": 750
    },
    {
      "entropy": 6.791053581237793,
      "epoch": 0.07493052798729655,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000377,
      "loss": 6.638,
      "mean_token_accuracy": 0.09132714346051216,
      "num_tokens": 1669425.0,
      "step": 755
    },
    {
      "entropy": 6.777761650085449,
      "epoch": 0.07542675664946408,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003795,
      "loss": 6.6407,
      "mean_token_accuracy": 0.09279207661747932,
      "num_tokens": 1680426.0,
      "step": 760
    },
    {
      "entropy": 6.760220146179199,
      "epoch": 0.0759229853116316,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000382,
      "loss": 6.6565,
      "mean_token_accuracy": 0.09227121248841286,
      "num_tokens": 1692803.0,
      "step": 765
    },
    {
      "entropy": 6.808647537231446,
      "epoch": 0.07641921397379912,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003845,
      "loss": 6.6722,
      "mean_token_accuracy": 0.09433563277125359,
      "num_tokens": 1703919.0,
      "step": 770
    },
    {
      "entropy": 6.796703290939331,
      "epoch": 0.07691544263596665,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.6536,
      "mean_token_accuracy": 0.10081230103969574,
      "num_tokens": 1714487.0,
      "step": 775
    },
    {
      "entropy": 6.826533603668213,
      "epoch": 0.07741167129813419,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.6979,
      "mean_token_accuracy": 0.09503390416502952,
      "num_tokens": 1726601.0,
      "step": 780
    },
    {
      "entropy": 6.7289519786834715,
      "epoch": 0.0779078999603017,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.5999,
      "mean_token_accuracy": 0.09352485463023186,
      "num_tokens": 1738743.0,
      "step": 785
    },
    {
      "entropy": 6.731641817092895,
      "epoch": 0.07840412862246923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.6684,
      "mean_token_accuracy": 0.09545368328690529,
      "num_tokens": 1748286.0,
      "step": 790
    },
    {
      "entropy": 6.755449342727661,
      "epoch": 0.07890035728463676,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.7012,
      "mean_token_accuracy": 0.10006179735064506,
      "num_tokens": 1758917.0,
      "step": 795
    },
    {
      "entropy": 6.707924747467041,
      "epoch": 0.0793965859468043,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003995,
      "loss": 6.5299,
      "mean_token_accuracy": 0.09730249717831611,
      "num_tokens": 1769179.0,
      "step": 800
    },
    {
      "entropy": 6.734688758850098,
      "epoch": 0.07989281460897181,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000402,
      "loss": 6.6672,
      "mean_token_accuracy": 0.09266628846526145,
      "num_tokens": 1782369.0,
      "step": 805
    },
    {
      "entropy": 6.783080816268921,
      "epoch": 0.08038904327113934,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004045,
      "loss": 6.6539,
      "mean_token_accuracy": 0.1000369630753994,
      "num_tokens": 1795992.0,
      "step": 810
    },
    {
      "entropy": 6.690567588806152,
      "epoch": 0.08088527193330687,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.5647,
      "mean_token_accuracy": 0.10151613354682923,
      "num_tokens": 1806938.0,
      "step": 815
    },
    {
      "entropy": 6.62420883178711,
      "epoch": 0.0813815005954744,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004095,
      "loss": 6.489,
      "mean_token_accuracy": 0.10247653797268867,
      "num_tokens": 1817663.0,
      "step": 820
    },
    {
      "entropy": 6.7051942348480225,
      "epoch": 0.08187772925764192,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000412,
      "loss": 6.6431,
      "mean_token_accuracy": 0.09830244034528732,
      "num_tokens": 1828254.0,
      "step": 825
    },
    {
      "entropy": 6.740552711486816,
      "epoch": 0.08237395791980945,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004145,
      "loss": 6.5703,
      "mean_token_accuracy": 0.09990052878856659,
      "num_tokens": 1839001.0,
      "step": 830
    },
    {
      "entropy": 6.675515031814575,
      "epoch": 0.08287018658197698,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000417,
      "loss": 6.6802,
      "mean_token_accuracy": 0.09807795733213424,
      "num_tokens": 1849906.0,
      "step": 835
    },
    {
      "entropy": 6.5840569019317625,
      "epoch": 0.0833664152441445,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004195,
      "loss": 6.5699,
      "mean_token_accuracy": 0.09285617843270302,
      "num_tokens": 1860982.0,
      "step": 840
    },
    {
      "entropy": 6.794158124923706,
      "epoch": 0.08386264390631203,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000422,
      "loss": 6.5139,
      "mean_token_accuracy": 0.10143777877092361,
      "num_tokens": 1872177.0,
      "step": 845
    },
    {
      "entropy": 6.596243000030517,
      "epoch": 0.08435887256847956,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004245,
      "loss": 6.4961,
      "mean_token_accuracy": 0.09911343604326248,
      "num_tokens": 1882783.0,
      "step": 850
    },
    {
      "entropy": 6.771530103683472,
      "epoch": 0.08485510123064709,
      "grad_norm": 1.0625,
      "learning_rate": 0.000427,
      "loss": 6.6692,
      "mean_token_accuracy": 0.09356241896748543,
      "num_tokens": 1893706.0,
      "step": 855
    },
    {
      "entropy": 6.628360414505005,
      "epoch": 0.0853513298928146,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004295,
      "loss": 6.545,
      "mean_token_accuracy": 0.09712332859635353,
      "num_tokens": 1903852.0,
      "step": 860
    },
    {
      "entropy": 6.61391053199768,
      "epoch": 0.08584755855498213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000432,
      "loss": 6.5803,
      "mean_token_accuracy": 0.10353554785251617,
      "num_tokens": 1914157.0,
      "step": 865
    },
    {
      "entropy": 6.660257244110108,
      "epoch": 0.08634378721714966,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004345,
      "loss": 6.507,
      "mean_token_accuracy": 0.09838124290108681,
      "num_tokens": 1924684.0,
      "step": 870
    },
    {
      "entropy": 6.690595579147339,
      "epoch": 0.0868400158793172,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000437,
      "loss": 6.6703,
      "mean_token_accuracy": 0.09410161599516868,
      "num_tokens": 1935657.0,
      "step": 875
    },
    {
      "entropy": 6.611545896530151,
      "epoch": 0.08733624454148471,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004395,
      "loss": 6.5409,
      "mean_token_accuracy": 0.09920520260930062,
      "num_tokens": 1946735.0,
      "step": 880
    },
    {
      "entropy": 6.714548015594483,
      "epoch": 0.08783247320365224,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000442,
      "loss": 6.7169,
      "mean_token_accuracy": 0.09320614784955979,
      "num_tokens": 1957871.0,
      "step": 885
    },
    {
      "entropy": 6.615920686721802,
      "epoch": 0.08832870186581977,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004445,
      "loss": 6.5327,
      "mean_token_accuracy": 0.09827413335442543,
      "num_tokens": 1969032.0,
      "step": 890
    },
    {
      "entropy": 6.704651021957398,
      "epoch": 0.0888249305279873,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000447,
      "loss": 6.5488,
      "mean_token_accuracy": 0.09838233441114426,
      "num_tokens": 1980212.0,
      "step": 895
    },
    {
      "entropy": 6.558094882965088,
      "epoch": 0.08932115919015482,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.4855,
      "mean_token_accuracy": 0.09865308403968812,
      "num_tokens": 1990333.0,
      "step": 900
    },
    {
      "entropy": 6.647093868255615,
      "epoch": 0.08981738785232235,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.5654,
      "mean_token_accuracy": 0.10231334716081619,
      "num_tokens": 2001486.0,
      "step": 905
    },
    {
      "entropy": 6.630073118209839,
      "epoch": 0.09031361651448988,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.5384,
      "mean_token_accuracy": 0.10137662813067436,
      "num_tokens": 2012535.0,
      "step": 910
    },
    {
      "entropy": 6.546298170089722,
      "epoch": 0.0908098451766574,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.4936,
      "mean_token_accuracy": 0.10514995977282524,
      "num_tokens": 2024116.0,
      "step": 915
    },
    {
      "entropy": 6.470860958099365,
      "epoch": 0.09130607383882493,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.5403,
      "mean_token_accuracy": 0.09997001364827156,
      "num_tokens": 2036341.0,
      "step": 920
    },
    {
      "entropy": 6.682841682434082,
      "epoch": 0.09180230250099246,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000462,
      "loss": 6.5173,
      "mean_token_accuracy": 0.09831630811095238,
      "num_tokens": 2046768.0,
      "step": 925
    },
    {
      "entropy": 6.5218424797058105,
      "epoch": 0.09229853116315999,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004645,
      "loss": 6.485,
      "mean_token_accuracy": 0.10012486949563026,
      "num_tokens": 2057315.0,
      "step": 930
    },
    {
      "entropy": 6.539970731735229,
      "epoch": 0.0927947598253275,
      "grad_norm": 1.0,
      "learning_rate": 0.000467,
      "loss": 6.4551,
      "mean_token_accuracy": 0.10452077984809875,
      "num_tokens": 2068267.0,
      "step": 935
    },
    {
      "entropy": 6.538546848297119,
      "epoch": 0.09329098848749504,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004695,
      "loss": 6.5244,
      "mean_token_accuracy": 0.10441852286458016,
      "num_tokens": 2078271.0,
      "step": 940
    },
    {
      "entropy": 6.543840551376343,
      "epoch": 0.09378721714966257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000472,
      "loss": 6.4694,
      "mean_token_accuracy": 0.09577009156346321,
      "num_tokens": 2088381.0,
      "step": 945
    },
    {
      "entropy": 6.56317892074585,
      "epoch": 0.0942834458118301,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004745,
      "loss": 6.4148,
      "mean_token_accuracy": 0.09971964061260223,
      "num_tokens": 2098899.0,
      "step": 950
    },
    {
      "entropy": 6.514262676239014,
      "epoch": 0.09477967447399761,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000477,
      "loss": 6.5697,
      "mean_token_accuracy": 0.09980119243264199,
      "num_tokens": 2110057.0,
      "step": 955
    },
    {
      "entropy": 6.484142780303955,
      "epoch": 0.09527590313616514,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004795,
      "loss": 6.4877,
      "mean_token_accuracy": 0.10196142494678498,
      "num_tokens": 2121758.0,
      "step": 960
    },
    {
      "entropy": 6.572415971755982,
      "epoch": 0.09577213179833267,
      "grad_norm": 0.921875,
      "learning_rate": 0.000482,
      "loss": 6.4947,
      "mean_token_accuracy": 0.10148481652140617,
      "num_tokens": 2131840.0,
      "step": 965
    },
    {
      "entropy": 6.479370641708374,
      "epoch": 0.0962683604605002,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004845,
      "loss": 6.4686,
      "mean_token_accuracy": 0.09988197311758995,
      "num_tokens": 2142084.0,
      "step": 970
    },
    {
      "entropy": 6.543777370452881,
      "epoch": 0.09676458912266772,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000487,
      "loss": 6.5201,
      "mean_token_accuracy": 0.10057111382484436,
      "num_tokens": 2152785.0,
      "step": 975
    },
    {
      "entropy": 6.58097357749939,
      "epoch": 0.09726081778483525,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004895,
      "loss": 6.5173,
      "mean_token_accuracy": 0.09812606722116471,
      "num_tokens": 2163774.0,
      "step": 980
    },
    {
      "entropy": 6.429722023010254,
      "epoch": 0.09775704644700278,
      "grad_norm": 0.859375,
      "learning_rate": 0.000492,
      "loss": 6.4215,
      "mean_token_accuracy": 0.10499785169959068,
      "num_tokens": 2175378.0,
      "step": 985
    },
    {
      "entropy": 6.654616022109986,
      "epoch": 0.0982532751091703,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004945,
      "loss": 6.6198,
      "mean_token_accuracy": 0.10243403911590576,
      "num_tokens": 2186510.0,
      "step": 990
    },
    {
      "entropy": 6.519903516769409,
      "epoch": 0.09874950377133783,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000497,
      "loss": 6.4311,
      "mean_token_accuracy": 0.10723361372947693,
      "num_tokens": 2197482.0,
      "step": 995
    },
    {
      "entropy": 6.421766948699951,
      "epoch": 0.09924573243350536,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995,
      "loss": 6.415,
      "mean_token_accuracy": 0.10275844857096672,
      "num_tokens": 2208329.0,
      "step": 1000
    },
    {
      "entropy": 6.386524295806884,
      "epoch": 0.09974196109567289,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999999982145551,
      "loss": 6.328,
      "mean_token_accuracy": 0.1056441530585289,
      "num_tokens": 2218632.0,
      "step": 1005
    },
    {
      "entropy": 6.481380987167358,
      "epoch": 0.1002381897578404,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999999909611855,
      "loss": 6.4444,
      "mean_token_accuracy": 0.1021676316857338,
      "num_tokens": 2231335.0,
      "step": 1010
    },
    {
      "entropy": 6.500467729568482,
      "epoch": 0.10073441842000794,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999999781283008,
      "loss": 6.4356,
      "mean_token_accuracy": 0.10310698971152306,
      "num_tokens": 2241704.0,
      "step": 1015
    },
    {
      "entropy": 6.412548065185547,
      "epoch": 0.10123064708217547,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999999597159017,
      "loss": 6.5081,
      "mean_token_accuracy": 0.09907391145825387,
      "num_tokens": 2252310.0,
      "step": 1020
    },
    {
      "entropy": 6.580806827545166,
      "epoch": 0.101726875744343,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999999357239882,
      "loss": 6.4046,
      "mean_token_accuracy": 0.10047375708818436,
      "num_tokens": 2262270.0,
      "step": 1025
    },
    {
      "entropy": 6.458594131469726,
      "epoch": 0.10222310440651051,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999999061525612,
      "loss": 6.5467,
      "mean_token_accuracy": 0.09957677125930786,
      "num_tokens": 2273986.0,
      "step": 1030
    },
    {
      "entropy": 6.567448663711548,
      "epoch": 0.10271933306867805,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999998710016214,
      "loss": 6.4326,
      "mean_token_accuracy": 0.10295439511537552,
      "num_tokens": 2283430.0,
      "step": 1035
    },
    {
      "entropy": 6.519222831726074,
      "epoch": 0.10321556173084558,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999998302711696,
      "loss": 6.446,
      "mean_token_accuracy": 0.09964360818266868,
      "num_tokens": 2294201.0,
      "step": 1040
    },
    {
      "entropy": 6.324606132507324,
      "epoch": 0.1037117903930131,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999997839612069,
      "loss": 6.3525,
      "mean_token_accuracy": 0.10616270750761032,
      "num_tokens": 2303876.0,
      "step": 1045
    },
    {
      "entropy": 6.41707067489624,
      "epoch": 0.10420801905518062,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999997320717342,
      "loss": 6.3981,
      "mean_token_accuracy": 0.10809943452477455,
      "num_tokens": 2313713.0,
      "step": 1050
    },
    {
      "entropy": 6.40936689376831,
      "epoch": 0.10470424771734815,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999996746027531,
      "loss": 6.4,
      "mean_token_accuracy": 0.10011846721172332,
      "num_tokens": 2324862.0,
      "step": 1055
    },
    {
      "entropy": 6.488155174255371,
      "epoch": 0.10520047637951568,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499999611554265,
      "loss": 6.4697,
      "mean_token_accuracy": 0.10407731533050538,
      "num_tokens": 2334974.0,
      "step": 1060
    },
    {
      "entropy": 6.491655540466309,
      "epoch": 0.10569670504168321,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999995429262712,
      "loss": 6.5141,
      "mean_token_accuracy": 0.10263812616467476,
      "num_tokens": 2345945.0,
      "step": 1065
    },
    {
      "entropy": 6.451990365982056,
      "epoch": 0.10619293370385073,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999994687187737,
      "loss": 6.424,
      "mean_token_accuracy": 0.10567846298217773,
      "num_tokens": 2356749.0,
      "step": 1070
    },
    {
      "entropy": 6.4968970775604244,
      "epoch": 0.10668916236601826,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999993889317742,
      "loss": 6.5025,
      "mean_token_accuracy": 0.09787843525409698,
      "num_tokens": 2368181.0,
      "step": 1075
    },
    {
      "entropy": 6.48430438041687,
      "epoch": 0.10718539102818579,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999993035652746,
      "loss": 6.4122,
      "mean_token_accuracy": 0.10403527170419694,
      "num_tokens": 2377737.0,
      "step": 1080
    },
    {
      "entropy": 6.356083869934082,
      "epoch": 0.10768161969035331,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999992126192771,
      "loss": 6.3292,
      "mean_token_accuracy": 0.10865161493420601,
      "num_tokens": 2388944.0,
      "step": 1085
    },
    {
      "entropy": 6.479707813262939,
      "epoch": 0.10817784835252084,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499999116093784,
      "loss": 6.4133,
      "mean_token_accuracy": 0.10108591020107269,
      "num_tokens": 2400758.0,
      "step": 1090
    },
    {
      "entropy": 6.363013744354248,
      "epoch": 0.10867407701468837,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999990139887976,
      "loss": 6.4044,
      "mean_token_accuracy": 0.10706018880009652,
      "num_tokens": 2411916.0,
      "step": 1095
    },
    {
      "entropy": 6.606807518005371,
      "epoch": 0.1091703056768559,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999989063043206,
      "loss": 6.4428,
      "mean_token_accuracy": 0.10081739947199822,
      "num_tokens": 2423310.0,
      "step": 1100
    },
    {
      "entropy": 6.337709760665893,
      "epoch": 0.10966653433902342,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999987930403555,
      "loss": 6.3665,
      "mean_token_accuracy": 0.10392144694924355,
      "num_tokens": 2432100.0,
      "step": 1105
    },
    {
      "entropy": 6.414704656600952,
      "epoch": 0.11016276300119095,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999986741969051,
      "loss": 6.4419,
      "mean_token_accuracy": 0.106266438215971,
      "num_tokens": 2443563.0,
      "step": 1110
    },
    {
      "entropy": 6.537985706329346,
      "epoch": 0.11065899166335848,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999985497739724,
      "loss": 6.4454,
      "mean_token_accuracy": 0.10320524126291275,
      "num_tokens": 2455435.0,
      "step": 1115
    },
    {
      "entropy": 6.321274328231811,
      "epoch": 0.11115522032552601,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999984197715606,
      "loss": 6.3804,
      "mean_token_accuracy": 0.10974763482809066,
      "num_tokens": 2466173.0,
      "step": 1120
    },
    {
      "entropy": 6.357246780395508,
      "epoch": 0.11165144898769352,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999982841896728,
      "loss": 6.3293,
      "mean_token_accuracy": 0.10533915609121322,
      "num_tokens": 2476349.0,
      "step": 1125
    },
    {
      "entropy": 6.488060379028321,
      "epoch": 0.11214767764986106,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999981430283123,
      "loss": 6.4613,
      "mean_token_accuracy": 0.10952543765306473,
      "num_tokens": 2487864.0,
      "step": 1130
    },
    {
      "entropy": 6.366047811508179,
      "epoch": 0.11264390631202859,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999979962874827,
      "loss": 6.3898,
      "mean_token_accuracy": 0.10598978772759438,
      "num_tokens": 2499188.0,
      "step": 1135
    },
    {
      "entropy": 6.48383674621582,
      "epoch": 0.11314013497419612,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999978439671877,
      "loss": 6.3782,
      "mean_token_accuracy": 0.10689363330602646,
      "num_tokens": 2509813.0,
      "step": 1140
    },
    {
      "entropy": 6.394137144088745,
      "epoch": 0.11363636363636363,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999976860674309,
      "loss": 6.4071,
      "mean_token_accuracy": 0.10305101722478867,
      "num_tokens": 2520813.0,
      "step": 1145
    },
    {
      "entropy": 6.387377500534058,
      "epoch": 0.11413259229853116,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999975225882164,
      "loss": 6.3967,
      "mean_token_accuracy": 0.10229461118578911,
      "num_tokens": 2531802.0,
      "step": 1150
    },
    {
      "entropy": 6.323769569396973,
      "epoch": 0.1146288209606987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499997353529548,
      "loss": 6.2047,
      "mean_token_accuracy": 0.12049669027328491,
      "num_tokens": 2541935.0,
      "step": 1155
    },
    {
      "entropy": 6.31162691116333,
      "epoch": 0.11512504962286621,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999971788914303,
      "loss": 6.3047,
      "mean_token_accuracy": 0.1079749420285225,
      "num_tokens": 2553224.0,
      "step": 1160
    },
    {
      "entropy": 6.3400167465209964,
      "epoch": 0.11562127828503374,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004999969986738672,
      "loss": 6.3378,
      "mean_token_accuracy": 0.10177241489291192,
      "num_tokens": 2565355.0,
      "step": 1165
    },
    {
      "entropy": 6.436748456954956,
      "epoch": 0.11611750694720127,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999968128768634,
      "loss": 6.4301,
      "mean_token_accuracy": 0.1053300179541111,
      "num_tokens": 2575962.0,
      "step": 1170
    },
    {
      "entropy": 6.377365303039551,
      "epoch": 0.1166137356093688,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999966215004236,
      "loss": 6.4384,
      "mean_token_accuracy": 0.11097546443343162,
      "num_tokens": 2588136.0,
      "step": 1175
    },
    {
      "entropy": 6.446149015426636,
      "epoch": 0.11710996427153632,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999964245445522,
      "loss": 6.3112,
      "mean_token_accuracy": 0.10942729488015175,
      "num_tokens": 2597785.0,
      "step": 1180
    },
    {
      "entropy": 6.339339637756348,
      "epoch": 0.11760619293370385,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999962220092545,
      "loss": 6.3702,
      "mean_token_accuracy": 0.11220366284251213,
      "num_tokens": 2609738.0,
      "step": 1185
    },
    {
      "entropy": 6.273840665817261,
      "epoch": 0.11810242159587138,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999960138945351,
      "loss": 6.2457,
      "mean_token_accuracy": 0.11884041801095009,
      "num_tokens": 2620179.0,
      "step": 1190
    },
    {
      "entropy": 6.364072942733765,
      "epoch": 0.11859865025803891,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999958002003996,
      "loss": 6.2789,
      "mean_token_accuracy": 0.10949316993355751,
      "num_tokens": 2631550.0,
      "step": 1195
    },
    {
      "entropy": 6.356981039047241,
      "epoch": 0.11909487892020643,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999955809268529,
      "loss": 6.33,
      "mean_token_accuracy": 0.10574184879660606,
      "num_tokens": 2642467.0,
      "step": 1200
    },
    {
      "entropy": 6.3026384830474855,
      "epoch": 0.11959110758237396,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999953560739008,
      "loss": 6.3069,
      "mean_token_accuracy": 0.10857978910207748,
      "num_tokens": 2652132.0,
      "step": 1205
    },
    {
      "entropy": 6.351331615447998,
      "epoch": 0.12008733624454149,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999951256415485,
      "loss": 6.3573,
      "mean_token_accuracy": 0.10829652771353722,
      "num_tokens": 2663336.0,
      "step": 1210
    },
    {
      "entropy": 6.465686893463134,
      "epoch": 0.12058356490670902,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499994889629802,
      "loss": 6.4458,
      "mean_token_accuracy": 0.09753695875406265,
      "num_tokens": 2674201.0,
      "step": 1215
    },
    {
      "entropy": 6.390116167068482,
      "epoch": 0.12107979356887653,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499994648038667,
      "loss": 6.3737,
      "mean_token_accuracy": 0.11085178926587105,
      "num_tokens": 2685073.0,
      "step": 1220
    },
    {
      "entropy": 6.4054559707641605,
      "epoch": 0.12157602223104406,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999944008681496,
      "loss": 6.3104,
      "mean_token_accuracy": 0.10459840670228004,
      "num_tokens": 2695024.0,
      "step": 1225
    },
    {
      "entropy": 6.346751165390015,
      "epoch": 0.1220722508932116,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999941481182559,
      "loss": 6.3983,
      "mean_token_accuracy": 0.10288520082831383,
      "num_tokens": 2706232.0,
      "step": 1230
    },
    {
      "entropy": 6.386320161819458,
      "epoch": 0.12256847955537911,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999938897889921,
      "loss": 6.3744,
      "mean_token_accuracy": 0.11175505593419074,
      "num_tokens": 2715874.0,
      "step": 1235
    },
    {
      "entropy": 6.317949056625366,
      "epoch": 0.12306470821754664,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999936258803647,
      "loss": 6.3121,
      "mean_token_accuracy": 0.11142380386590958,
      "num_tokens": 2727409.0,
      "step": 1240
    },
    {
      "entropy": 6.273382472991943,
      "epoch": 0.12356093687971417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999933563923802,
      "loss": 6.2154,
      "mean_token_accuracy": 0.1114815816283226,
      "num_tokens": 2737259.0,
      "step": 1245
    },
    {
      "entropy": 6.341649723052979,
      "epoch": 0.1240571655418817,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999930813250451,
      "loss": 6.3557,
      "mean_token_accuracy": 0.10958174094557763,
      "num_tokens": 2748416.0,
      "step": 1250
    },
    {
      "entropy": 6.3774512767791744,
      "epoch": 0.12455339420404922,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999928006783667,
      "loss": 6.3204,
      "mean_token_accuracy": 0.11053804978728295,
      "num_tokens": 2760470.0,
      "step": 1255
    },
    {
      "entropy": 6.337265825271606,
      "epoch": 0.12504962286621676,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999925144523515,
      "loss": 6.3438,
      "mean_token_accuracy": 0.104658143222332,
      "num_tokens": 2770222.0,
      "step": 1260
    },
    {
      "entropy": 6.334999036788941,
      "epoch": 0.12554585152838427,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999922226470066,
      "loss": 6.2795,
      "mean_token_accuracy": 0.10962152928113937,
      "num_tokens": 2780212.0,
      "step": 1265
    },
    {
      "entropy": 6.251826810836792,
      "epoch": 0.1260420801905518,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999919252623397,
      "loss": 6.2525,
      "mean_token_accuracy": 0.11011324599385261,
      "num_tokens": 2790239.0,
      "step": 1270
    },
    {
      "entropy": 6.402917814254761,
      "epoch": 0.12653830885271933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999916222983577,
      "loss": 6.2774,
      "mean_token_accuracy": 0.1086140252649784,
      "num_tokens": 2801902.0,
      "step": 1275
    },
    {
      "entropy": 6.204992580413818,
      "epoch": 0.12703453751488686,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999913137550682,
      "loss": 6.2721,
      "mean_token_accuracy": 0.10906686112284661,
      "num_tokens": 2813268.0,
      "step": 1280
    },
    {
      "entropy": 6.3671046733856205,
      "epoch": 0.1275307661770544,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999909996324791,
      "loss": 6.3598,
      "mean_token_accuracy": 0.10677680447697639,
      "num_tokens": 2824506.0,
      "step": 1285
    },
    {
      "entropy": 6.3029297351837155,
      "epoch": 0.12802699483922192,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499990679930598,
      "loss": 6.2556,
      "mean_token_accuracy": 0.11228155717253685,
      "num_tokens": 2835144.0,
      "step": 1290
    },
    {
      "entropy": 6.21924409866333,
      "epoch": 0.12852322350138945,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999903546494327,
      "loss": 6.2655,
      "mean_token_accuracy": 0.10710735470056534,
      "num_tokens": 2845382.0,
      "step": 1295
    },
    {
      "entropy": 6.328607368469238,
      "epoch": 0.12901945216355698,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999900237889916,
      "loss": 6.2964,
      "mean_token_accuracy": 0.1106351412832737,
      "num_tokens": 2854697.0,
      "step": 1300
    },
    {
      "entropy": 6.344019556045533,
      "epoch": 0.12951568082572448,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999896873492827,
      "loss": 6.2941,
      "mean_token_accuracy": 0.10326201841235161,
      "num_tokens": 2866216.0,
      "step": 1305
    },
    {
      "entropy": 6.258399963378906,
      "epoch": 0.130011909487892,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999893453303143,
      "loss": 6.2558,
      "mean_token_accuracy": 0.11178297996520996,
      "num_tokens": 2877734.0,
      "step": 1310
    },
    {
      "entropy": 6.344228315353393,
      "epoch": 0.13050813815005954,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999889977320949,
      "loss": 6.3055,
      "mean_token_accuracy": 0.11218280494213104,
      "num_tokens": 2889445.0,
      "step": 1315
    },
    {
      "entropy": 6.236292695999145,
      "epoch": 0.13100436681222707,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999886445546332,
      "loss": 6.2455,
      "mean_token_accuracy": 0.11572465375065803,
      "num_tokens": 2900485.0,
      "step": 1320
    },
    {
      "entropy": 6.3338706493377686,
      "epoch": 0.1315005954743946,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499988285797938,
      "loss": 6.2335,
      "mean_token_accuracy": 0.11599084883928298,
      "num_tokens": 2909795.0,
      "step": 1325
    },
    {
      "entropy": 6.189287090301514,
      "epoch": 0.13199682413656214,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499987921462018,
      "loss": 6.2602,
      "mean_token_accuracy": 0.11903260201215744,
      "num_tokens": 2920153.0,
      "step": 1330
    },
    {
      "entropy": 6.2777505874633786,
      "epoch": 0.13249305279872967,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999875515468825,
      "loss": 6.2512,
      "mean_token_accuracy": 0.1119612142443657,
      "num_tokens": 2930457.0,
      "step": 1335
    },
    {
      "entropy": 6.312501382827759,
      "epoch": 0.13298928146089717,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999871760525403,
      "loss": 6.3165,
      "mean_token_accuracy": 0.10896633416414261,
      "num_tokens": 2941305.0,
      "step": 1340
    },
    {
      "entropy": 6.297947740554809,
      "epoch": 0.1334855101230647,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999867949790011,
      "loss": 6.2476,
      "mean_token_accuracy": 0.11246784254908562,
      "num_tokens": 2951798.0,
      "step": 1345
    },
    {
      "entropy": 6.243113613128662,
      "epoch": 0.13398173878523223,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999864083262743,
      "loss": 6.162,
      "mean_token_accuracy": 0.11877489164471626,
      "num_tokens": 2962486.0,
      "step": 1350
    },
    {
      "entropy": 6.228711986541748,
      "epoch": 0.13447796744739976,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999860160943691,
      "loss": 6.2658,
      "mean_token_accuracy": 0.11053448170423508,
      "num_tokens": 2974317.0,
      "step": 1355
    },
    {
      "entropy": 6.215711355209351,
      "epoch": 0.1349741961095673,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999856182832958,
      "loss": 6.1817,
      "mean_token_accuracy": 0.11494843661785126,
      "num_tokens": 2985440.0,
      "step": 1360
    },
    {
      "entropy": 6.291191291809082,
      "epoch": 0.13547042477173482,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999852148930637,
      "loss": 6.2546,
      "mean_token_accuracy": 0.11355492547154426,
      "num_tokens": 2997044.0,
      "step": 1365
    },
    {
      "entropy": 6.23797197341919,
      "epoch": 0.13596665343390235,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999848059236832,
      "loss": 6.262,
      "mean_token_accuracy": 0.1076361782848835,
      "num_tokens": 3005965.0,
      "step": 1370
    },
    {
      "entropy": 6.239419984817505,
      "epoch": 0.13646288209606988,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999843913751643,
      "loss": 6.1175,
      "mean_token_accuracy": 0.11854560673236847,
      "num_tokens": 3015031.0,
      "step": 1375
    },
    {
      "entropy": 6.195667743682861,
      "epoch": 0.13695911075823738,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999839712475174,
      "loss": 6.1909,
      "mean_token_accuracy": 0.11440791487693787,
      "num_tokens": 3026275.0,
      "step": 1380
    },
    {
      "entropy": 6.2530198097229,
      "epoch": 0.13745533942040492,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999835455407526,
      "loss": 6.2354,
      "mean_token_accuracy": 0.1116865485906601,
      "num_tokens": 3036778.0,
      "step": 1385
    },
    {
      "entropy": 6.314652347564698,
      "epoch": 0.13795156808257245,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999831142548808,
      "loss": 6.3486,
      "mean_token_accuracy": 0.11005391925573349,
      "num_tokens": 3048149.0,
      "step": 1390
    },
    {
      "entropy": 6.2538410186767575,
      "epoch": 0.13844779674473998,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999826773899124,
      "loss": 6.214,
      "mean_token_accuracy": 0.11115290969610214,
      "num_tokens": 3059149.0,
      "step": 1395
    },
    {
      "entropy": 6.2944536209106445,
      "epoch": 0.1389440254069075,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004999822349458586,
      "loss": 6.226,
      "mean_token_accuracy": 0.11450750976800919,
      "num_tokens": 3071613.0,
      "step": 1400
    },
    {
      "entropy": 6.322410726547242,
      "epoch": 0.13944025406907504,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999817869227301,
      "loss": 6.3583,
      "mean_token_accuracy": 0.10745183452963829,
      "num_tokens": 3084742.0,
      "step": 1405
    },
    {
      "entropy": 6.289934778213501,
      "epoch": 0.13993648273124257,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999813333205381,
      "loss": 6.2773,
      "mean_token_accuracy": 0.11367975026369095,
      "num_tokens": 3095356.0,
      "step": 1410
    },
    {
      "entropy": 6.247778844833374,
      "epoch": 0.14043271139341007,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999808741392937,
      "loss": 6.2522,
      "mean_token_accuracy": 0.10875962376594543,
      "num_tokens": 3107717.0,
      "step": 1415
    },
    {
      "entropy": 6.226774549484253,
      "epoch": 0.1409289400555776,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999804093790086,
      "loss": 6.251,
      "mean_token_accuracy": 0.10959596037864686,
      "num_tokens": 3119205.0,
      "step": 1420
    },
    {
      "entropy": 6.309685230255127,
      "epoch": 0.14142516871774513,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499979939039694,
      "loss": 6.2812,
      "mean_token_accuracy": 0.11436648070812225,
      "num_tokens": 3130564.0,
      "step": 1425
    },
    {
      "entropy": 6.275003099441529,
      "epoch": 0.14192139737991266,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999794631213618,
      "loss": 6.22,
      "mean_token_accuracy": 0.11114049553871155,
      "num_tokens": 3141883.0,
      "step": 1430
    },
    {
      "entropy": 6.281518220901489,
      "epoch": 0.1424176260420802,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999789816240236,
      "loss": 6.2302,
      "mean_token_accuracy": 0.11397273391485215,
      "num_tokens": 3152817.0,
      "step": 1435
    },
    {
      "entropy": 6.17591028213501,
      "epoch": 0.14291385470424772,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999784945476915,
      "loss": 6.1486,
      "mean_token_accuracy": 0.11776702702045441,
      "num_tokens": 3163546.0,
      "step": 1440
    },
    {
      "entropy": 6.216448545455933,
      "epoch": 0.14341008336641525,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999780018923775,
      "loss": 6.1949,
      "mean_token_accuracy": 0.10940220728516578,
      "num_tokens": 3173899.0,
      "step": 1445
    },
    {
      "entropy": 6.15954270362854,
      "epoch": 0.14390631202858278,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999775036580939,
      "loss": 6.0969,
      "mean_token_accuracy": 0.11880587413907051,
      "num_tokens": 3184433.0,
      "step": 1450
    },
    {
      "entropy": 6.242514944076538,
      "epoch": 0.1444025406907503,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999769998448529,
      "loss": 6.2142,
      "mean_token_accuracy": 0.10818950086832047,
      "num_tokens": 3196077.0,
      "step": 1455
    },
    {
      "entropy": 6.138770914077758,
      "epoch": 0.14489876935291782,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999764904526672,
      "loss": 6.1403,
      "mean_token_accuracy": 0.11515245363116264,
      "num_tokens": 3206645.0,
      "step": 1460
    },
    {
      "entropy": 6.3275079250335695,
      "epoch": 0.14539499801508535,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999759754815492,
      "loss": 6.2513,
      "mean_token_accuracy": 0.11938706040382385,
      "num_tokens": 3216786.0,
      "step": 1465
    },
    {
      "entropy": 6.199605226516724,
      "epoch": 0.14589122667725288,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999754549315118,
      "loss": 6.1823,
      "mean_token_accuracy": 0.11293254643678666,
      "num_tokens": 3227384.0,
      "step": 1470
    },
    {
      "entropy": 6.108070659637451,
      "epoch": 0.1463874553394204,
      "grad_norm": 0.828125,
      "learning_rate": 0.000499974928802568,
      "loss": 6.0334,
      "mean_token_accuracy": 0.11813693642616271,
      "num_tokens": 3238800.0,
      "step": 1475
    },
    {
      "entropy": 6.180989933013916,
      "epoch": 0.14688368400158794,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999743970947306,
      "loss": 6.2034,
      "mean_token_accuracy": 0.11198758184909821,
      "num_tokens": 3248627.0,
      "step": 1480
    },
    {
      "entropy": 6.197469997406006,
      "epoch": 0.14737991266375547,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499973859808013,
      "loss": 6.2012,
      "mean_token_accuracy": 0.11548065319657326,
      "num_tokens": 3258989.0,
      "step": 1485
    },
    {
      "entropy": 6.222706890106201,
      "epoch": 0.14787614132592297,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999733169424284,
      "loss": 6.1897,
      "mean_token_accuracy": 0.11579868942499161,
      "num_tokens": 3269474.0,
      "step": 1490
    },
    {
      "entropy": 6.1594078540802,
      "epoch": 0.1483723699880905,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999727684979904,
      "loss": 6.1833,
      "mean_token_accuracy": 0.11907703801989555,
      "num_tokens": 3280453.0,
      "step": 1495
    },
    {
      "entropy": 6.195897006988526,
      "epoch": 0.14886859865025803,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999722144747125,
      "loss": 6.1237,
      "mean_token_accuracy": 0.11352067068219185,
      "num_tokens": 3290234.0,
      "step": 1500
    },
    {
      "entropy": 6.077924489974976,
      "epoch": 0.14936482731242556,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999716548726085,
      "loss": 6.1053,
      "mean_token_accuracy": 0.1176950380206108,
      "num_tokens": 3300943.0,
      "step": 1505
    },
    {
      "entropy": 6.199636650085449,
      "epoch": 0.1498610559745931,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999710896916921,
      "loss": 6.0607,
      "mean_token_accuracy": 0.11784961298108101,
      "num_tokens": 3311296.0,
      "step": 1510
    },
    {
      "entropy": 6.159174585342408,
      "epoch": 0.15035728463676062,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999705189319774,
      "loss": 6.2248,
      "mean_token_accuracy": 0.10928900018334389,
      "num_tokens": 3322310.0,
      "step": 1515
    },
    {
      "entropy": 6.187262296676636,
      "epoch": 0.15085351329892815,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999699425934788,
      "loss": 6.2243,
      "mean_token_accuracy": 0.11323094964027405,
      "num_tokens": 3334176.0,
      "step": 1520
    },
    {
      "entropy": 6.178784799575806,
      "epoch": 0.15134974196109569,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999693606762103,
      "loss": 6.2073,
      "mean_token_accuracy": 0.10905725955963134,
      "num_tokens": 3344839.0,
      "step": 1525
    },
    {
      "entropy": 6.270957374572754,
      "epoch": 0.1518459706232632,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999687731801865,
      "loss": 6.1825,
      "mean_token_accuracy": 0.11094732880592346,
      "num_tokens": 3357046.0,
      "step": 1530
    },
    {
      "entropy": 6.150786876678467,
      "epoch": 0.15234219928543072,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999681801054217,
      "loss": 6.2501,
      "mean_token_accuracy": 0.11335301697254181,
      "num_tokens": 3367783.0,
      "step": 1535
    },
    {
      "entropy": 6.206290578842163,
      "epoch": 0.15283842794759825,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499967581451931,
      "loss": 6.1134,
      "mean_token_accuracy": 0.11155367642641068,
      "num_tokens": 3378529.0,
      "step": 1540
    },
    {
      "entropy": 6.179344654083252,
      "epoch": 0.15333465660976578,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499966977219729,
      "loss": 6.3094,
      "mean_token_accuracy": 0.11146700233221055,
      "num_tokens": 3389318.0,
      "step": 1545
    },
    {
      "entropy": 6.294518518447876,
      "epoch": 0.1538308852719333,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999663674088306,
      "loss": 6.313,
      "mean_token_accuracy": 0.10722307339310647,
      "num_tokens": 3400603.0,
      "step": 1550
    },
    {
      "entropy": 6.270757818222046,
      "epoch": 0.15432711393410084,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999657520192511,
      "loss": 6.1938,
      "mean_token_accuracy": 0.11140423566102982,
      "num_tokens": 3410316.0,
      "step": 1555
    },
    {
      "entropy": 6.242702484130859,
      "epoch": 0.15482334259626837,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999651310510058,
      "loss": 6.1446,
      "mean_token_accuracy": 0.11679755821824074,
      "num_tokens": 3422155.0,
      "step": 1560
    },
    {
      "entropy": 6.17017388343811,
      "epoch": 0.15531957125843587,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999645045041099,
      "loss": 6.0721,
      "mean_token_accuracy": 0.11942542716860771,
      "num_tokens": 3432892.0,
      "step": 1565
    },
    {
      "entropy": 6.174464702606201,
      "epoch": 0.1558157999206034,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999638723785791,
      "loss": 6.2059,
      "mean_token_accuracy": 0.11832942962646484,
      "num_tokens": 3443967.0,
      "step": 1570
    },
    {
      "entropy": 6.144176197052002,
      "epoch": 0.15631202858277093,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499963234674429,
      "loss": 6.0957,
      "mean_token_accuracy": 0.12020640000700951,
      "num_tokens": 3454354.0,
      "step": 1575
    },
    {
      "entropy": 6.038722705841065,
      "epoch": 0.15680825724493846,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999625913916754,
      "loss": 6.1151,
      "mean_token_accuracy": 0.11729008257389069,
      "num_tokens": 3465232.0,
      "step": 1580
    },
    {
      "entropy": 6.125705909729004,
      "epoch": 0.157304485907106,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999619425303343,
      "loss": 6.0879,
      "mean_token_accuracy": 0.12173460498452186,
      "num_tokens": 3475443.0,
      "step": 1585
    },
    {
      "entropy": 6.21633563041687,
      "epoch": 0.15780071456927353,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999612880904219,
      "loss": 6.1283,
      "mean_token_accuracy": 0.11958795338869095,
      "num_tokens": 3486643.0,
      "step": 1590
    },
    {
      "entropy": 6.145956325531006,
      "epoch": 0.15829694323144106,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999606280719542,
      "loss": 6.173,
      "mean_token_accuracy": 0.11473753675818443,
      "num_tokens": 3497557.0,
      "step": 1595
    },
    {
      "entropy": 6.26501202583313,
      "epoch": 0.1587931718936086,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999599624749475,
      "loss": 6.1633,
      "mean_token_accuracy": 0.1159546658396721,
      "num_tokens": 3509790.0,
      "step": 1600
    },
    {
      "entropy": 6.065554237365722,
      "epoch": 0.1592894005557761,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999592912994187,
      "loss": 6.0811,
      "mean_token_accuracy": 0.11343559995293617,
      "num_tokens": 3521577.0,
      "step": 1605
    },
    {
      "entropy": 6.207467794418335,
      "epoch": 0.15978562921794362,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999586145453842,
      "loss": 6.1295,
      "mean_token_accuracy": 0.10831035524606705,
      "num_tokens": 3532126.0,
      "step": 1610
    },
    {
      "entropy": 5.990396785736084,
      "epoch": 0.16028185788011115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999579322128608,
      "loss": 5.9618,
      "mean_token_accuracy": 0.11795824691653252,
      "num_tokens": 3542560.0,
      "step": 1615
    },
    {
      "entropy": 6.117848062515259,
      "epoch": 0.16077808654227868,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999572443018654,
      "loss": 6.137,
      "mean_token_accuracy": 0.11492020189762116,
      "num_tokens": 3553863.0,
      "step": 1620
    },
    {
      "entropy": 6.154970407485962,
      "epoch": 0.1612743152044462,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999565508124151,
      "loss": 6.1508,
      "mean_token_accuracy": 0.11818645223975181,
      "num_tokens": 3564953.0,
      "step": 1625
    },
    {
      "entropy": 6.148259401321411,
      "epoch": 0.16177054386661374,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499955851744527,
      "loss": 6.1469,
      "mean_token_accuracy": 0.11931650564074517,
      "num_tokens": 3576247.0,
      "step": 1630
    },
    {
      "entropy": 6.276931715011597,
      "epoch": 0.16226677252878127,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999551470982185,
      "loss": 6.1795,
      "mean_token_accuracy": 0.11549673676490783,
      "num_tokens": 3587600.0,
      "step": 1635
    },
    {
      "entropy": 6.206128883361816,
      "epoch": 0.1627630011909488,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999544368735073,
      "loss": 6.1665,
      "mean_token_accuracy": 0.11787421181797982,
      "num_tokens": 3597727.0,
      "step": 1640
    },
    {
      "entropy": 6.091890716552735,
      "epoch": 0.1632592298531163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999537210704106,
      "loss": 6.1341,
      "mean_token_accuracy": 0.1190246470272541,
      "num_tokens": 3608066.0,
      "step": 1645
    },
    {
      "entropy": 6.20475926399231,
      "epoch": 0.16375545851528384,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999529996889463,
      "loss": 6.1987,
      "mean_token_accuracy": 0.11030768677592277,
      "num_tokens": 3618263.0,
      "step": 1650
    },
    {
      "entropy": 6.184187507629394,
      "epoch": 0.16425168717745137,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999522727291325,
      "loss": 6.1268,
      "mean_token_accuracy": 0.11970260366797447,
      "num_tokens": 3629465.0,
      "step": 1655
    },
    {
      "entropy": 6.179236936569214,
      "epoch": 0.1647479158396189,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499951540190987,
      "loss": 6.154,
      "mean_token_accuracy": 0.11628529205918312,
      "num_tokens": 3639657.0,
      "step": 1660
    },
    {
      "entropy": 6.174856472015381,
      "epoch": 0.16524414450178643,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999508020745281,
      "loss": 6.1086,
      "mean_token_accuracy": 0.1154894195497036,
      "num_tokens": 3652374.0,
      "step": 1665
    },
    {
      "entropy": 6.104015922546386,
      "epoch": 0.16574037316395396,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999500583797739,
      "loss": 6.0969,
      "mean_token_accuracy": 0.11629031077027321,
      "num_tokens": 3664083.0,
      "step": 1670
    },
    {
      "entropy": 6.143244981765747,
      "epoch": 0.1662366018261215,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999493091067431,
      "loss": 6.1398,
      "mean_token_accuracy": 0.1213216483592987,
      "num_tokens": 3675314.0,
      "step": 1675
    },
    {
      "entropy": 6.131026697158814,
      "epoch": 0.166732830488289,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999485542554541,
      "loss": 6.0897,
      "mean_token_accuracy": 0.10728424638509751,
      "num_tokens": 3686019.0,
      "step": 1680
    },
    {
      "entropy": 6.1852058410644535,
      "epoch": 0.16722905915045652,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999477938259256,
      "loss": 6.196,
      "mean_token_accuracy": 0.11261390075087548,
      "num_tokens": 3698170.0,
      "step": 1685
    },
    {
      "entropy": 6.133717155456543,
      "epoch": 0.16772528781262405,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004999470278181766,
      "loss": 6.186,
      "mean_token_accuracy": 0.11304959654808044,
      "num_tokens": 3709243.0,
      "step": 1690
    },
    {
      "entropy": 6.20488805770874,
      "epoch": 0.16822151647479158,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999462562322261,
      "loss": 6.1447,
      "mean_token_accuracy": 0.11698395982384682,
      "num_tokens": 3720196.0,
      "step": 1695
    },
    {
      "entropy": 6.172057008743286,
      "epoch": 0.1687177451369591,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499945479068093,
      "loss": 6.1913,
      "mean_token_accuracy": 0.11760392040014267,
      "num_tokens": 3732714.0,
      "step": 1700
    },
    {
      "entropy": 6.189049911499024,
      "epoch": 0.16921397379912664,
      "grad_norm": 0.765625,
      "learning_rate": 0.0004999446963257969,
      "loss": 6.1714,
      "mean_token_accuracy": 0.11059175580739974,
      "num_tokens": 3746474.0,
      "step": 1705
    },
    {
      "entropy": 6.084878253936767,
      "epoch": 0.16971020246129417,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999439080053571,
      "loss": 6.0414,
      "mean_token_accuracy": 0.12301717177033425,
      "num_tokens": 3757528.0,
      "step": 1710
    },
    {
      "entropy": 6.24871335029602,
      "epoch": 0.1702064311234617,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999431141067928,
      "loss": 6.1689,
      "mean_token_accuracy": 0.11376198828220367,
      "num_tokens": 3768360.0,
      "step": 1715
    },
    {
      "entropy": 6.100267028808593,
      "epoch": 0.1707026597856292,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999423146301242,
      "loss": 6.0366,
      "mean_token_accuracy": 0.11526194885373116,
      "num_tokens": 3779809.0,
      "step": 1720
    },
    {
      "entropy": 6.038203620910645,
      "epoch": 0.17119888844779674,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999415095753709,
      "loss": 6.03,
      "mean_token_accuracy": 0.11550426036119461,
      "num_tokens": 3791455.0,
      "step": 1725
    },
    {
      "entropy": 6.056922388076782,
      "epoch": 0.17169511710996427,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999406989425527,
      "loss": 6.0831,
      "mean_token_accuracy": 0.11582357957959175,
      "num_tokens": 3803209.0,
      "step": 1730
    },
    {
      "entropy": 6.087919235229492,
      "epoch": 0.1721913457721318,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999398827316901,
      "loss": 6.0977,
      "mean_token_accuracy": 0.11691041514277459,
      "num_tokens": 3813553.0,
      "step": 1735
    },
    {
      "entropy": 6.157799959182739,
      "epoch": 0.17268757443429933,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499939060942803,
      "loss": 6.1439,
      "mean_token_accuracy": 0.10938044339418411,
      "num_tokens": 3824856.0,
      "step": 1740
    },
    {
      "entropy": 6.220856618881226,
      "epoch": 0.17318380309646686,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999382335759119,
      "loss": 6.2309,
      "mean_token_accuracy": 0.11000438928604125,
      "num_tokens": 3835489.0,
      "step": 1745
    },
    {
      "entropy": 6.134221601486206,
      "epoch": 0.1736800317586344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999374006310372,
      "loss": 6.0638,
      "mean_token_accuracy": 0.11413465365767479,
      "num_tokens": 3844838.0,
      "step": 1750
    },
    {
      "entropy": 6.067860126495361,
      "epoch": 0.1741762604208019,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999365621081999,
      "loss": 6.0341,
      "mean_token_accuracy": 0.11537155881524086,
      "num_tokens": 3855288.0,
      "step": 1755
    },
    {
      "entropy": 6.108763790130615,
      "epoch": 0.17467248908296942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999357180074204,
      "loss": 6.0853,
      "mean_token_accuracy": 0.12003114819526672,
      "num_tokens": 3865585.0,
      "step": 1760
    },
    {
      "entropy": 6.129130601882935,
      "epoch": 0.17516871774513695,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999348683287199,
      "loss": 6.0621,
      "mean_token_accuracy": 0.11469487771391869,
      "num_tokens": 3877694.0,
      "step": 1765
    },
    {
      "entropy": 6.094071865081787,
      "epoch": 0.17566494640730448,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999340130721192,
      "loss": 6.0386,
      "mean_token_accuracy": 0.12670635432004929,
      "num_tokens": 3888802.0,
      "step": 1770
    },
    {
      "entropy": 6.04920973777771,
      "epoch": 0.17616117506947201,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999331522376396,
      "loss": 6.0047,
      "mean_token_accuracy": 0.11465661451220513,
      "num_tokens": 3900294.0,
      "step": 1775
    },
    {
      "entropy": 6.030979776382447,
      "epoch": 0.17665740373163955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999322858253027,
      "loss": 6.0339,
      "mean_token_accuracy": 0.12006412371993065,
      "num_tokens": 3910865.0,
      "step": 1780
    },
    {
      "entropy": 6.070213460922242,
      "epoch": 0.17715363239380708,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999314138351297,
      "loss": 6.1034,
      "mean_token_accuracy": 0.1195277787744999,
      "num_tokens": 3921960.0,
      "step": 1785
    },
    {
      "entropy": 6.161261892318725,
      "epoch": 0.1776498610559746,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999305362671424,
      "loss": 6.1101,
      "mean_token_accuracy": 0.11471805498003959,
      "num_tokens": 3934085.0,
      "step": 1790
    },
    {
      "entropy": 6.052437019348145,
      "epoch": 0.1781460897181421,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999296531213624,
      "loss": 5.9506,
      "mean_token_accuracy": 0.1281500868499279,
      "num_tokens": 3944532.0,
      "step": 1795
    },
    {
      "entropy": 6.051175928115844,
      "epoch": 0.17864231838030964,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999287643978117,
      "loss": 5.9968,
      "mean_token_accuracy": 0.12282953560352325,
      "num_tokens": 3954995.0,
      "step": 1800
    },
    {
      "entropy": 5.990139293670654,
      "epoch": 0.17913854704247717,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999278700965124,
      "loss": 6.0559,
      "mean_token_accuracy": 0.12489112317562104,
      "num_tokens": 3966621.0,
      "step": 1805
    },
    {
      "entropy": 6.166021299362183,
      "epoch": 0.1796347757046447,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999269702174866,
      "loss": 6.119,
      "mean_token_accuracy": 0.11739438995718957,
      "num_tokens": 3977568.0,
      "step": 1810
    },
    {
      "entropy": 6.073067665100098,
      "epoch": 0.18013100436681223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999260647607565,
      "loss": 6.1373,
      "mean_token_accuracy": 0.11110868230462075,
      "num_tokens": 3988571.0,
      "step": 1815
    },
    {
      "entropy": 6.138613891601563,
      "epoch": 0.18062723302897976,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999251537263448,
      "loss": 5.9477,
      "mean_token_accuracy": 0.1258134886622429,
      "num_tokens": 3998380.0,
      "step": 1820
    },
    {
      "entropy": 6.0303796291351315,
      "epoch": 0.1811234616911473,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999242371142739,
      "loss": 6.1287,
      "mean_token_accuracy": 0.12175759598612786,
      "num_tokens": 4010017.0,
      "step": 1825
    },
    {
      "entropy": 6.189897441864014,
      "epoch": 0.1816196903533148,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999233149245665,
      "loss": 6.0939,
      "mean_token_accuracy": 0.1167963296175003,
      "num_tokens": 4020922.0,
      "step": 1830
    },
    {
      "entropy": 6.123414802551269,
      "epoch": 0.18211591901548232,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999223871572456,
      "loss": 6.0808,
      "mean_token_accuracy": 0.11856624335050583,
      "num_tokens": 4030927.0,
      "step": 1835
    },
    {
      "entropy": 6.089998006820679,
      "epoch": 0.18261214767764986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999214538123342,
      "loss": 6.1432,
      "mean_token_accuracy": 0.11269862428307534,
      "num_tokens": 4041375.0,
      "step": 1840
    },
    {
      "entropy": 6.0862884521484375,
      "epoch": 0.18310837633981739,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999205148898554,
      "loss": 6.1411,
      "mean_token_accuracy": 0.11809576898813248,
      "num_tokens": 4053213.0,
      "step": 1845
    },
    {
      "entropy": 6.142365455627441,
      "epoch": 0.18360460500198492,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999195703898325,
      "loss": 5.9824,
      "mean_token_accuracy": 0.12277794778347015,
      "num_tokens": 4064536.0,
      "step": 1850
    },
    {
      "entropy": 6.081491136550904,
      "epoch": 0.18410083366415245,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999186203122889,
      "loss": 6.0625,
      "mean_token_accuracy": 0.12416149452328681,
      "num_tokens": 4075282.0,
      "step": 1855
    },
    {
      "entropy": 6.013858938217163,
      "epoch": 0.18459706232631998,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999176646572482,
      "loss": 6.0563,
      "mean_token_accuracy": 0.1242367833852768,
      "num_tokens": 4085501.0,
      "step": 1860
    },
    {
      "entropy": 6.091145706176758,
      "epoch": 0.1850932909884875,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499916703424734,
      "loss": 6.0064,
      "mean_token_accuracy": 0.11573861911892891,
      "num_tokens": 4095697.0,
      "step": 1865
    },
    {
      "entropy": 6.072324752807617,
      "epoch": 0.185589519650655,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999157366147702,
      "loss": 6.0599,
      "mean_token_accuracy": 0.1197610542178154,
      "num_tokens": 4106459.0,
      "step": 1870
    },
    {
      "entropy": 6.090449953079224,
      "epoch": 0.18608574831282254,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999147642273809,
      "loss": 5.998,
      "mean_token_accuracy": 0.11759251281619072,
      "num_tokens": 4118320.0,
      "step": 1875
    },
    {
      "entropy": 5.991074848175049,
      "epoch": 0.18658197697499007,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049991378626259,
      "loss": 6.0245,
      "mean_token_accuracy": 0.12806328162550926,
      "num_tokens": 4131249.0,
      "step": 1880
    },
    {
      "entropy": 6.093191146850586,
      "epoch": 0.1870782056371576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999128027204219,
      "loss": 6.096,
      "mean_token_accuracy": 0.11801933944225311,
      "num_tokens": 4141480.0,
      "step": 1885
    },
    {
      "entropy": 6.126098394393921,
      "epoch": 0.18757443429932513,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499911813600901,
      "loss": 6.1558,
      "mean_token_accuracy": 0.11598219275474549,
      "num_tokens": 4152344.0,
      "step": 1890
    },
    {
      "entropy": 6.173122406005859,
      "epoch": 0.18807066296149266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999108189040516,
      "loss": 6.1105,
      "mean_token_accuracy": 0.1155496172606945,
      "num_tokens": 4164772.0,
      "step": 1895
    },
    {
      "entropy": 6.0153892040252686,
      "epoch": 0.1885668916236602,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999098186298986,
      "loss": 6.0434,
      "mean_token_accuracy": 0.11878785714507104,
      "num_tokens": 4175511.0,
      "step": 1900
    },
    {
      "entropy": 6.039416551589966,
      "epoch": 0.1890631202858277,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999088127784667,
      "loss": 6.0148,
      "mean_token_accuracy": 0.11949604898691177,
      "num_tokens": 4187445.0,
      "step": 1905
    },
    {
      "entropy": 6.069828367233276,
      "epoch": 0.18955934894799523,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499907801349781,
      "loss": 6.0138,
      "mean_token_accuracy": 0.11969439908862115,
      "num_tokens": 4198124.0,
      "step": 1910
    },
    {
      "entropy": 6.048256969451904,
      "epoch": 0.19005557761016276,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999067843438664,
      "loss": 6.0505,
      "mean_token_accuracy": 0.11893324404954911,
      "num_tokens": 4208954.0,
      "step": 1915
    },
    {
      "entropy": 6.1492753505706785,
      "epoch": 0.1905518062723303,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999057617607482,
      "loss": 6.1068,
      "mean_token_accuracy": 0.11971905007958412,
      "num_tokens": 4218896.0,
      "step": 1920
    },
    {
      "entropy": 6.014862298965454,
      "epoch": 0.19104803493449782,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999047336004517,
      "loss": 6.0095,
      "mean_token_accuracy": 0.11468009054660797,
      "num_tokens": 4228851.0,
      "step": 1925
    },
    {
      "entropy": 5.982360315322876,
      "epoch": 0.19154426359666535,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999036998630024,
      "loss": 5.9659,
      "mean_token_accuracy": 0.12004356682300568,
      "num_tokens": 4240377.0,
      "step": 1930
    },
    {
      "entropy": 5.99826283454895,
      "epoch": 0.19204049225883288,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499902660548426,
      "loss": 5.9635,
      "mean_token_accuracy": 0.1215371236205101,
      "num_tokens": 4251341.0,
      "step": 1935
    },
    {
      "entropy": 6.100321102142334,
      "epoch": 0.1925367209210004,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999016156567482,
      "loss": 6.0659,
      "mean_token_accuracy": 0.12046532705426216,
      "num_tokens": 4261206.0,
      "step": 1940
    },
    {
      "entropy": 6.080335712432861,
      "epoch": 0.1930329495831679,
      "grad_norm": 0.875,
      "learning_rate": 0.000499900565187995,
      "loss": 6.0723,
      "mean_token_accuracy": 0.11344361081719398,
      "num_tokens": 4272662.0,
      "step": 1945
    },
    {
      "entropy": 6.123919200897217,
      "epoch": 0.19352917824533544,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998995091421924,
      "loss": 6.0541,
      "mean_token_accuracy": 0.11754847764968872,
      "num_tokens": 4283302.0,
      "step": 1950
    },
    {
      "entropy": 5.979746294021607,
      "epoch": 0.19402540690750297,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998984475193665,
      "loss": 5.9978,
      "mean_token_accuracy": 0.12176699414849282,
      "num_tokens": 4296087.0,
      "step": 1955
    },
    {
      "entropy": 6.052849388122558,
      "epoch": 0.1945216355696705,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998973803195439,
      "loss": 6.0093,
      "mean_token_accuracy": 0.12155119329690933,
      "num_tokens": 4305918.0,
      "step": 1960
    },
    {
      "entropy": 6.0047338008880615,
      "epoch": 0.19501786423183803,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998963075427508,
      "loss": 6.0725,
      "mean_token_accuracy": 0.11939249560236931,
      "num_tokens": 4316507.0,
      "step": 1965
    },
    {
      "entropy": 6.079337644577026,
      "epoch": 0.19551409289400556,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004998952291890138,
      "loss": 5.9977,
      "mean_token_accuracy": 0.1253109961748123,
      "num_tokens": 4329091.0,
      "step": 1970
    },
    {
      "entropy": 6.064958190917968,
      "epoch": 0.1960103215561731,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998941452583599,
      "loss": 6.0762,
      "mean_token_accuracy": 0.12064750641584396,
      "num_tokens": 4339266.0,
      "step": 1975
    },
    {
      "entropy": 5.988303995132446,
      "epoch": 0.1965065502183406,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998930557508156,
      "loss": 5.9826,
      "mean_token_accuracy": 0.12658628076314926,
      "num_tokens": 4349300.0,
      "step": 1980
    },
    {
      "entropy": 5.995124006271363,
      "epoch": 0.19700277888050813,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998919606664082,
      "loss": 5.9143,
      "mean_token_accuracy": 0.13212018981575965,
      "num_tokens": 4360314.0,
      "step": 1985
    },
    {
      "entropy": 6.041628313064575,
      "epoch": 0.19749900754267566,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998908600051648,
      "loss": 5.9705,
      "mean_token_accuracy": 0.12074373364448547,
      "num_tokens": 4370077.0,
      "step": 1990
    },
    {
      "entropy": 5.971292448043823,
      "epoch": 0.1979952362048432,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998897537671126,
      "loss": 5.9705,
      "mean_token_accuracy": 0.12351988479495049,
      "num_tokens": 4381000.0,
      "step": 1995
    },
    {
      "entropy": 6.094802951812744,
      "epoch": 0.19849146486701072,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998886419522792,
      "loss": 6.1871,
      "mean_token_accuracy": 0.11414133310317993,
      "num_tokens": 4393485.0,
      "step": 2000
    },
    {
      "entropy": 6.160151147842408,
      "epoch": 0.19898769352917825,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998875245606919,
      "loss": 6.0252,
      "mean_token_accuracy": 0.12122476175427437,
      "num_tokens": 4404354.0,
      "step": 2005
    },
    {
      "entropy": 6.004584503173828,
      "epoch": 0.19948392219134578,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998864015923788,
      "loss": 6.1679,
      "mean_token_accuracy": 0.12271130830049515,
      "num_tokens": 4414818.0,
      "step": 2010
    },
    {
      "entropy": 6.125269937515259,
      "epoch": 0.1999801508535133,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998852730473674,
      "loss": 6.0353,
      "mean_token_accuracy": 0.11513861268758774,
      "num_tokens": 4426271.0,
      "step": 2015
    },
    {
      "entropy": 6.017019605636596,
      "epoch": 0.2004763795156808,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998841389256859,
      "loss": 6.1004,
      "mean_token_accuracy": 0.11922633796930313,
      "num_tokens": 4436981.0,
      "step": 2020
    },
    {
      "entropy": 6.022520351409912,
      "epoch": 0.20097260817784834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998829992273622,
      "loss": 5.9114,
      "mean_token_accuracy": 0.12870077788829803,
      "num_tokens": 4446494.0,
      "step": 2025
    },
    {
      "entropy": 5.996196699142456,
      "epoch": 0.20146883684001587,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998818539524248,
      "loss": 6.0074,
      "mean_token_accuracy": 0.1251174710690975,
      "num_tokens": 4458535.0,
      "step": 2030
    },
    {
      "entropy": 6.087194585800171,
      "epoch": 0.2019650655021834,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499880703100902,
      "loss": 6.0014,
      "mean_token_accuracy": 0.12391166165471076,
      "num_tokens": 4469025.0,
      "step": 2035
    },
    {
      "entropy": 6.010956525802612,
      "epoch": 0.20246129416435094,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998795466728223,
      "loss": 5.986,
      "mean_token_accuracy": 0.12361643686890603,
      "num_tokens": 4478431.0,
      "step": 2040
    },
    {
      "entropy": 5.994862222671509,
      "epoch": 0.20295752282651847,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998783846682144,
      "loss": 6.1286,
      "mean_token_accuracy": 0.11792842224240303,
      "num_tokens": 4490215.0,
      "step": 2045
    },
    {
      "entropy": 6.091261959075927,
      "epoch": 0.203453751488686,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004998772170871072,
      "loss": 5.9872,
      "mean_token_accuracy": 0.1232263281941414,
      "num_tokens": 4502895.0,
      "step": 2050
    },
    {
      "entropy": 6.012570142745972,
      "epoch": 0.20394998015085353,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998760439295295,
      "loss": 6.0185,
      "mean_token_accuracy": 0.12015356048941613,
      "num_tokens": 4514468.0,
      "step": 2055
    },
    {
      "entropy": 6.082999658584595,
      "epoch": 0.20444620881302103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998748651955105,
      "loss": 6.0319,
      "mean_token_accuracy": 0.12087535560131073,
      "num_tokens": 4525787.0,
      "step": 2060
    },
    {
      "entropy": 6.01540241241455,
      "epoch": 0.20494243747518856,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004998736808850795,
      "loss": 6.1173,
      "mean_token_accuracy": 0.11445825323462486,
      "num_tokens": 4537536.0,
      "step": 2065
    },
    {
      "entropy": 6.1401674270629885,
      "epoch": 0.2054386661373561,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998724909982656,
      "loss": 6.0429,
      "mean_token_accuracy": 0.12223877683281899,
      "num_tokens": 4549619.0,
      "step": 2070
    },
    {
      "entropy": 5.962413740158081,
      "epoch": 0.20593489479952362,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998712955350986,
      "loss": 5.9808,
      "mean_token_accuracy": 0.12422219887375832,
      "num_tokens": 4560248.0,
      "step": 2075
    },
    {
      "entropy": 6.061628007888794,
      "epoch": 0.20643112346169115,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998700944956081,
      "loss": 6.0998,
      "mean_token_accuracy": 0.12046362236142158,
      "num_tokens": 4571888.0,
      "step": 2080
    },
    {
      "entropy": 6.062523126602173,
      "epoch": 0.20692735212385868,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998688878798237,
      "loss": 5.9366,
      "mean_token_accuracy": 0.13008833229541777,
      "num_tokens": 4583234.0,
      "step": 2085
    },
    {
      "entropy": 6.068667030334472,
      "epoch": 0.2074235807860262,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998676756877753,
      "loss": 6.1538,
      "mean_token_accuracy": 0.11486675590276718,
      "num_tokens": 4595252.0,
      "step": 2090
    },
    {
      "entropy": 6.0270226955413815,
      "epoch": 0.20791980944819372,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998664579194934,
      "loss": 5.9196,
      "mean_token_accuracy": 0.12394649386405945,
      "num_tokens": 4607091.0,
      "step": 2095
    },
    {
      "entropy": 6.029794692993164,
      "epoch": 0.20841603811036125,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998652345750077,
      "loss": 5.8893,
      "mean_token_accuracy": 0.12757945880293847,
      "num_tokens": 4618003.0,
      "step": 2100
    },
    {
      "entropy": 5.918219375610351,
      "epoch": 0.20891226677252878,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998640056543487,
      "loss": 5.8908,
      "mean_token_accuracy": 0.1205370806157589,
      "num_tokens": 4628429.0,
      "step": 2105
    },
    {
      "entropy": 6.010912561416626,
      "epoch": 0.2094084954346963,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998627711575469,
      "loss": 5.9631,
      "mean_token_accuracy": 0.12769097983837127,
      "num_tokens": 4639414.0,
      "step": 2110
    },
    {
      "entropy": 5.969498491287231,
      "epoch": 0.20990472409686384,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000499861531084633,
      "loss": 6.0658,
      "mean_token_accuracy": 0.12155803292989731,
      "num_tokens": 4652594.0,
      "step": 2115
    },
    {
      "entropy": 5.99975962638855,
      "epoch": 0.21040095275903137,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998602854356376,
      "loss": 5.8884,
      "mean_token_accuracy": 0.12490163445472717,
      "num_tokens": 4662981.0,
      "step": 2120
    },
    {
      "entropy": 5.956492328643799,
      "epoch": 0.2108971814211989,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998590342105917,
      "loss": 5.9854,
      "mean_token_accuracy": 0.12499078437685966,
      "num_tokens": 4672469.0,
      "step": 2125
    },
    {
      "entropy": 6.015379524230957,
      "epoch": 0.21139341008336643,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998577774095262,
      "loss": 5.9799,
      "mean_token_accuracy": 0.11904595419764519,
      "num_tokens": 4683640.0,
      "step": 2130
    },
    {
      "entropy": 5.978434896469116,
      "epoch": 0.21188963874553393,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998565150324724,
      "loss": 5.9373,
      "mean_token_accuracy": 0.12495681792497634,
      "num_tokens": 4694032.0,
      "step": 2135
    },
    {
      "entropy": 5.911871337890625,
      "epoch": 0.21238586740770146,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998552470794616,
      "loss": 5.823,
      "mean_token_accuracy": 0.12693770080804825,
      "num_tokens": 4704253.0,
      "step": 2140
    },
    {
      "entropy": 5.94983696937561,
      "epoch": 0.212882096069869,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998539735505251,
      "loss": 5.9686,
      "mean_token_accuracy": 0.11972972378134727,
      "num_tokens": 4715327.0,
      "step": 2145
    },
    {
      "entropy": 6.121762084960937,
      "epoch": 0.21337832473203652,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998526944456946,
      "loss": 5.9843,
      "mean_token_accuracy": 0.11618738770484924,
      "num_tokens": 4726450.0,
      "step": 2150
    },
    {
      "entropy": 5.960577487945557,
      "epoch": 0.21387455339420405,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998514097650018,
      "loss": 6.0075,
      "mean_token_accuracy": 0.11687625348567962,
      "num_tokens": 4736630.0,
      "step": 2155
    },
    {
      "entropy": 6.02658486366272,
      "epoch": 0.21437078205637158,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998501195084786,
      "loss": 5.8837,
      "mean_token_accuracy": 0.12827948555350305,
      "num_tokens": 4746592.0,
      "step": 2160
    },
    {
      "entropy": 5.982325458526612,
      "epoch": 0.21486701071853911,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998488236761569,
      "loss": 6.0762,
      "mean_token_accuracy": 0.1284046731889248,
      "num_tokens": 4757824.0,
      "step": 2165
    },
    {
      "entropy": 6.027835941314697,
      "epoch": 0.21536323938070662,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998475222680688,
      "loss": 6.0,
      "mean_token_accuracy": 0.12244343236088753,
      "num_tokens": 4768453.0,
      "step": 2170
    },
    {
      "entropy": 5.9532242774963375,
      "epoch": 0.21585946804287415,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998462152842468,
      "loss": 5.9826,
      "mean_token_accuracy": 0.11837821900844574,
      "num_tokens": 4779664.0,
      "step": 2175
    },
    {
      "entropy": 6.096589851379394,
      "epoch": 0.21635569670504168,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499844902724723,
      "loss": 6.0019,
      "mean_token_accuracy": 0.1223648227751255,
      "num_tokens": 4791012.0,
      "step": 2180
    },
    {
      "entropy": 5.902175521850586,
      "epoch": 0.2168519253672092,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998435845895302,
      "loss": 5.8794,
      "mean_token_accuracy": 0.12457833737134934,
      "num_tokens": 4800578.0,
      "step": 2185
    },
    {
      "entropy": 5.982421445846557,
      "epoch": 0.21734815402937674,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499842260878701,
      "loss": 5.8903,
      "mean_token_accuracy": 0.12047785967588424,
      "num_tokens": 4810940.0,
      "step": 2190
    },
    {
      "entropy": 5.987110233306884,
      "epoch": 0.21784438269154427,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004998409315922683,
      "loss": 6.0421,
      "mean_token_accuracy": 0.12021683305501937,
      "num_tokens": 4822696.0,
      "step": 2195
    },
    {
      "entropy": 6.000821590423584,
      "epoch": 0.2183406113537118,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998395967302649,
      "loss": 5.896,
      "mean_token_accuracy": 0.1261095680296421,
      "num_tokens": 4833669.0,
      "step": 2200
    },
    {
      "entropy": 6.080050802230835,
      "epoch": 0.21883684001587933,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499838256292724,
      "loss": 6.0807,
      "mean_token_accuracy": 0.12102781087160111,
      "num_tokens": 4844021.0,
      "step": 2205
    },
    {
      "entropy": 5.95439338684082,
      "epoch": 0.21933306867804683,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998369102796787,
      "loss": 5.9517,
      "mean_token_accuracy": 0.12605141550302507,
      "num_tokens": 4855336.0,
      "step": 2210
    },
    {
      "entropy": 6.0016449928283695,
      "epoch": 0.21982929734021436,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004998355586911627,
      "loss": 5.9464,
      "mean_token_accuracy": 0.12409858405590057,
      "num_tokens": 4867713.0,
      "step": 2215
    },
    {
      "entropy": 5.9480170726776125,
      "epoch": 0.2203255260023819,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998342015272093,
      "loss": 5.9325,
      "mean_token_accuracy": 0.12322600036859513,
      "num_tokens": 4879134.0,
      "step": 2220
    },
    {
      "entropy": 6.047569274902344,
      "epoch": 0.22082175466454942,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998328387878521,
      "loss": 6.0989,
      "mean_token_accuracy": 0.1215038277208805,
      "num_tokens": 4891755.0,
      "step": 2225
    },
    {
      "entropy": 6.152658557891845,
      "epoch": 0.22131798332671695,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998314704731249,
      "loss": 6.0836,
      "mean_token_accuracy": 0.11992574110627174,
      "num_tokens": 4903264.0,
      "step": 2230
    },
    {
      "entropy": 5.933616685867309,
      "epoch": 0.22181421198888449,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998300965830618,
      "loss": 5.9148,
      "mean_token_accuracy": 0.12403179481625556,
      "num_tokens": 4912533.0,
      "step": 2235
    },
    {
      "entropy": 5.878796815872192,
      "epoch": 0.22231044065105202,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998287171176968,
      "loss": 5.9132,
      "mean_token_accuracy": 0.12680847868323325,
      "num_tokens": 4923288.0,
      "step": 2240
    },
    {
      "entropy": 6.013169240951538,
      "epoch": 0.22280666931321952,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998273320770641,
      "loss": 5.9415,
      "mean_token_accuracy": 0.12333439588546753,
      "num_tokens": 4934759.0,
      "step": 2245
    },
    {
      "entropy": 5.932636260986328,
      "epoch": 0.22330289797538705,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499825941461198,
      "loss": 5.966,
      "mean_token_accuracy": 0.12490196079015732,
      "num_tokens": 4947024.0,
      "step": 2250
    },
    {
      "entropy": 5.971539068222046,
      "epoch": 0.22379912663755458,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998245452701329,
      "loss": 5.9222,
      "mean_token_accuracy": 0.12477421313524246,
      "num_tokens": 4959281.0,
      "step": 2255
    },
    {
      "entropy": 6.050222492218017,
      "epoch": 0.2242953552997221,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998231435039036,
      "loss": 5.9793,
      "mean_token_accuracy": 0.11924687549471855,
      "num_tokens": 4970466.0,
      "step": 2260
    },
    {
      "entropy": 5.949010181427002,
      "epoch": 0.22479158396188964,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998217361625449,
      "loss": 5.9377,
      "mean_token_accuracy": 0.12520918771624565,
      "num_tokens": 4981135.0,
      "step": 2265
    },
    {
      "entropy": 5.9944313049316404,
      "epoch": 0.22528781262405717,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998203232460915,
      "loss": 5.9933,
      "mean_token_accuracy": 0.12928032726049424,
      "num_tokens": 4991292.0,
      "step": 2270
    },
    {
      "entropy": 5.963805294036865,
      "epoch": 0.2257840412862247,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998189047545785,
      "loss": 5.9373,
      "mean_token_accuracy": 0.12618958726525306,
      "num_tokens": 5002806.0,
      "step": 2275
    },
    {
      "entropy": 5.945759534835815,
      "epoch": 0.22628026994839223,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998174806880411,
      "loss": 5.8989,
      "mean_token_accuracy": 0.13040453866124152,
      "num_tokens": 5013474.0,
      "step": 2280
    },
    {
      "entropy": 5.951687908172607,
      "epoch": 0.22677649861055973,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998160510465146,
      "loss": 5.9352,
      "mean_token_accuracy": 0.13000500723719596,
      "num_tokens": 5025151.0,
      "step": 2285
    },
    {
      "entropy": 5.926647806167603,
      "epoch": 0.22727272727272727,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998146158300345,
      "loss": 5.9359,
      "mean_token_accuracy": 0.1301646463572979,
      "num_tokens": 5037166.0,
      "step": 2290
    },
    {
      "entropy": 6.0174994468688965,
      "epoch": 0.2277689559348948,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998131750386364,
      "loss": 5.9849,
      "mean_token_accuracy": 0.12377216964960099,
      "num_tokens": 5047452.0,
      "step": 2295
    },
    {
      "entropy": 5.9414409637451175,
      "epoch": 0.22826518459706233,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998117286723559,
      "loss": 5.9811,
      "mean_token_accuracy": 0.12003129869699478,
      "num_tokens": 5058501.0,
      "step": 2300
    },
    {
      "entropy": 6.017748069763184,
      "epoch": 0.22876141325922986,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499810276731229,
      "loss": 5.9855,
      "mean_token_accuracy": 0.12389934659004212,
      "num_tokens": 5069716.0,
      "step": 2305
    },
    {
      "entropy": 6.036419868469238,
      "epoch": 0.2292576419213974,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998088192152918,
      "loss": 5.9543,
      "mean_token_accuracy": 0.12356158047914505,
      "num_tokens": 5078509.0,
      "step": 2310
    },
    {
      "entropy": 5.953884792327881,
      "epoch": 0.22975387058356492,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998073561245801,
      "loss": 6.0043,
      "mean_token_accuracy": 0.12320954129099845,
      "num_tokens": 5090355.0,
      "step": 2315
    },
    {
      "entropy": 6.02992992401123,
      "epoch": 0.23025009924573242,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998058874591305,
      "loss": 5.8885,
      "mean_token_accuracy": 0.12766256108880042,
      "num_tokens": 5100582.0,
      "step": 2320
    },
    {
      "entropy": 5.984819793701172,
      "epoch": 0.23074632790789995,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004998044132189793,
      "loss": 5.9563,
      "mean_token_accuracy": 0.12425923943519593,
      "num_tokens": 5112176.0,
      "step": 2325
    },
    {
      "entropy": 5.964567947387695,
      "epoch": 0.23124255657006748,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499802933404163,
      "loss": 5.8697,
      "mean_token_accuracy": 0.12405415698885917,
      "num_tokens": 5122840.0,
      "step": 2330
    },
    {
      "entropy": 5.942762327194214,
      "epoch": 0.231738785232235,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998014480147183,
      "loss": 5.8782,
      "mean_token_accuracy": 0.12970717400312423,
      "num_tokens": 5132926.0,
      "step": 2335
    },
    {
      "entropy": 5.883530950546264,
      "epoch": 0.23223501389440254,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997999570506823,
      "loss": 5.8536,
      "mean_token_accuracy": 0.12876308858394622,
      "num_tokens": 5143394.0,
      "step": 2340
    },
    {
      "entropy": 5.885425662994384,
      "epoch": 0.23273124255657007,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997984605120917,
      "loss": 5.8876,
      "mean_token_accuracy": 0.12427316978573799,
      "num_tokens": 5155710.0,
      "step": 2345
    },
    {
      "entropy": 5.837533855438233,
      "epoch": 0.2332274712187376,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997969583989835,
      "loss": 5.8163,
      "mean_token_accuracy": 0.12433635219931602,
      "num_tokens": 5167065.0,
      "step": 2350
    },
    {
      "entropy": 5.938689565658569,
      "epoch": 0.23372369988090513,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997954507113952,
      "loss": 5.9198,
      "mean_token_accuracy": 0.1236913837492466,
      "num_tokens": 5177535.0,
      "step": 2355
    },
    {
      "entropy": 5.9044514179229735,
      "epoch": 0.23421992854307264,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997939374493642,
      "loss": 5.8341,
      "mean_token_accuracy": 0.12915939837694168,
      "num_tokens": 5188324.0,
      "step": 2360
    },
    {
      "entropy": 5.918756198883057,
      "epoch": 0.23471615720524017,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499792418612928,
      "loss": 5.9219,
      "mean_token_accuracy": 0.12620971575379372,
      "num_tokens": 5199223.0,
      "step": 2365
    },
    {
      "entropy": 5.933010911941528,
      "epoch": 0.2352123858674077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997908942021241,
      "loss": 5.9016,
      "mean_token_accuracy": 0.12857112437486648,
      "num_tokens": 5210080.0,
      "step": 2370
    },
    {
      "entropy": 6.007422018051147,
      "epoch": 0.23570861452957523,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997893642169904,
      "loss": 6.0153,
      "mean_token_accuracy": 0.12386548593640327,
      "num_tokens": 5220867.0,
      "step": 2375
    },
    {
      "entropy": 5.9750877857208256,
      "epoch": 0.23620484319174276,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997878286575648,
      "loss": 5.8596,
      "mean_token_accuracy": 0.12206832617521286,
      "num_tokens": 5231451.0,
      "step": 2380
    },
    {
      "entropy": 5.932880973815918,
      "epoch": 0.2367010718539103,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997862875238855,
      "loss": 5.9662,
      "mean_token_accuracy": 0.12139840200543403,
      "num_tokens": 5244643.0,
      "step": 2385
    },
    {
      "entropy": 5.986804580688476,
      "epoch": 0.23719730051607782,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997847408159907,
      "loss": 5.9781,
      "mean_token_accuracy": 0.11682926490902901,
      "num_tokens": 5256279.0,
      "step": 2390
    },
    {
      "entropy": 5.988444232940674,
      "epoch": 0.23769352917824532,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997831885339186,
      "loss": 5.9486,
      "mean_token_accuracy": 0.12903917878866195,
      "num_tokens": 5268498.0,
      "step": 2395
    },
    {
      "entropy": 5.876736688613891,
      "epoch": 0.23818975784041285,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997816306777077,
      "loss": 5.9269,
      "mean_token_accuracy": 0.1273012116551399,
      "num_tokens": 5280869.0,
      "step": 2400
    },
    {
      "entropy": 6.107938194274903,
      "epoch": 0.23868598650258038,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997800672473969,
      "loss": 6.024,
      "mean_token_accuracy": 0.12057831212878227,
      "num_tokens": 5293082.0,
      "step": 2405
    },
    {
      "entropy": 5.8743205070495605,
      "epoch": 0.2391822151647479,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997784982430247,
      "loss": 5.8092,
      "mean_token_accuracy": 0.1349871814250946,
      "num_tokens": 5304039.0,
      "step": 2410
    },
    {
      "entropy": 5.888135862350464,
      "epoch": 0.23967844382691544,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997769236646301,
      "loss": 5.8874,
      "mean_token_accuracy": 0.12667458057403563,
      "num_tokens": 5315016.0,
      "step": 2415
    },
    {
      "entropy": 6.046527481079101,
      "epoch": 0.24017467248908297,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997753435122522,
      "loss": 6.0304,
      "mean_token_accuracy": 0.1209819495677948,
      "num_tokens": 5326090.0,
      "step": 2420
    },
    {
      "entropy": 5.8429787158966064,
      "epoch": 0.2406709011512505,
      "grad_norm": 1.0,
      "learning_rate": 0.00049977375778593,
      "loss": 5.8061,
      "mean_token_accuracy": 0.128286737203598,
      "num_tokens": 5336423.0,
      "step": 2425
    },
    {
      "entropy": 5.994162750244141,
      "epoch": 0.24116712981341804,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997721664857032,
      "loss": 5.9764,
      "mean_token_accuracy": 0.1251453071832657,
      "num_tokens": 5346982.0,
      "step": 2430
    },
    {
      "entropy": 5.9075380802154545,
      "epoch": 0.24166335847558554,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997705696116108,
      "loss": 5.8802,
      "mean_token_accuracy": 0.13289263024926184,
      "num_tokens": 5359210.0,
      "step": 2435
    },
    {
      "entropy": 5.939221620559692,
      "epoch": 0.24215958713775307,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997689671636926,
      "loss": 5.8914,
      "mean_token_accuracy": 0.11953317150473594,
      "num_tokens": 5369891.0,
      "step": 2440
    },
    {
      "entropy": 5.891269207000732,
      "epoch": 0.2426558157999206,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997673591419884,
      "loss": 5.8777,
      "mean_token_accuracy": 0.12496770992875099,
      "num_tokens": 5380850.0,
      "step": 2445
    },
    {
      "entropy": 5.939934349060058,
      "epoch": 0.24315204446208813,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004997657455465381,
      "loss": 5.9339,
      "mean_token_accuracy": 0.12732330337166786,
      "num_tokens": 5391903.0,
      "step": 2450
    },
    {
      "entropy": 5.993845796585083,
      "epoch": 0.24364827312425566,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997641263773815,
      "loss": 5.9966,
      "mean_token_accuracy": 0.12384477332234382,
      "num_tokens": 5402517.0,
      "step": 2455
    },
    {
      "entropy": 5.8536583423614506,
      "epoch": 0.2441445017864232,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499762501634559,
      "loss": 5.9017,
      "mean_token_accuracy": 0.1264289692044258,
      "num_tokens": 5414110.0,
      "step": 2460
    },
    {
      "entropy": 6.013503313064575,
      "epoch": 0.24464073044859072,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997608713181108,
      "loss": 5.9143,
      "mean_token_accuracy": 0.12651659920811653,
      "num_tokens": 5424466.0,
      "step": 2465
    },
    {
      "entropy": 5.9579092502594,
      "epoch": 0.24513695911075822,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997592354280772,
      "loss": 5.9171,
      "mean_token_accuracy": 0.13171297907829285,
      "num_tokens": 5434461.0,
      "step": 2470
    },
    {
      "entropy": 5.863575887680054,
      "epoch": 0.24563318777292575,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499757593964499,
      "loss": 5.8918,
      "mean_token_accuracy": 0.1223128654062748,
      "num_tokens": 5445921.0,
      "step": 2475
    },
    {
      "entropy": 5.945754766464233,
      "epoch": 0.24612941643509328,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997559469274166,
      "loss": 5.8846,
      "mean_token_accuracy": 0.12594258412718773,
      "num_tokens": 5456283.0,
      "step": 2480
    },
    {
      "entropy": 5.98380069732666,
      "epoch": 0.24662564509726081,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997542943168712,
      "loss": 5.9281,
      "mean_token_accuracy": 0.1247189499437809,
      "num_tokens": 5467616.0,
      "step": 2485
    },
    {
      "entropy": 5.927638578414917,
      "epoch": 0.24712187375942835,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997526361329035,
      "loss": 5.8764,
      "mean_token_accuracy": 0.13195810765028,
      "num_tokens": 5478050.0,
      "step": 2490
    },
    {
      "entropy": 5.97781229019165,
      "epoch": 0.24761810242159588,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997509723755547,
      "loss": 5.8846,
      "mean_token_accuracy": 0.12755061611533164,
      "num_tokens": 5487185.0,
      "step": 2495
    },
    {
      "entropy": 5.910194063186646,
      "epoch": 0.2481143310837634,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997493030448662,
      "loss": 5.9018,
      "mean_token_accuracy": 0.12408641576766968,
      "num_tokens": 5498802.0,
      "step": 2500
    },
    {
      "entropy": 5.906022214889527,
      "epoch": 0.24861055974593094,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997476281408793,
      "loss": 5.9155,
      "mean_token_accuracy": 0.12289765030145645,
      "num_tokens": 5510285.0,
      "step": 2505
    },
    {
      "entropy": 5.956116676330566,
      "epoch": 0.24910678840809844,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997459476636354,
      "loss": 5.8077,
      "mean_token_accuracy": 0.13291431069374085,
      "num_tokens": 5520028.0,
      "step": 2510
    },
    {
      "entropy": 5.836493730545044,
      "epoch": 0.24960301707026597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997442616131763,
      "loss": 5.9187,
      "mean_token_accuracy": 0.12413758859038353,
      "num_tokens": 5529271.0,
      "step": 2515
    },
    {
      "entropy": 5.923549509048462,
      "epoch": 0.25009924573243353,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997425699895438,
      "loss": 5.8478,
      "mean_token_accuracy": 0.12018645405769349,
      "num_tokens": 5539396.0,
      "step": 2520
    },
    {
      "entropy": 5.942753076553345,
      "epoch": 0.25059547439460106,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004997408727927798,
      "loss": 5.9803,
      "mean_token_accuracy": 0.12607812359929085,
      "num_tokens": 5551674.0,
      "step": 2525
    },
    {
      "entropy": 5.946086692810058,
      "epoch": 0.25109170305676853,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997391700229264,
      "loss": 5.9052,
      "mean_token_accuracy": 0.1281418062746525,
      "num_tokens": 5562749.0,
      "step": 2530
    },
    {
      "entropy": 5.937380933761597,
      "epoch": 0.25158793171893606,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499737461680026,
      "loss": 5.9167,
      "mean_token_accuracy": 0.12441470995545387,
      "num_tokens": 5573478.0,
      "step": 2535
    },
    {
      "entropy": 5.853802585601807,
      "epoch": 0.2520841603811036,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997357477641207,
      "loss": 5.9051,
      "mean_token_accuracy": 0.12962415516376496,
      "num_tokens": 5584546.0,
      "step": 2540
    },
    {
      "entropy": 5.911726713180542,
      "epoch": 0.2525803890432711,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997340282752532,
      "loss": 5.7971,
      "mean_token_accuracy": 0.13067309483885764,
      "num_tokens": 5594004.0,
      "step": 2545
    },
    {
      "entropy": 5.941144037246704,
      "epoch": 0.25307661770543866,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499732303213466,
      "loss": 5.8453,
      "mean_token_accuracy": 0.1209590882062912,
      "num_tokens": 5606095.0,
      "step": 2550
    },
    {
      "entropy": 5.907631063461304,
      "epoch": 0.2535728463676062,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997305725788019,
      "loss": 5.8957,
      "mean_token_accuracy": 0.11941977143287659,
      "num_tokens": 5615596.0,
      "step": 2555
    },
    {
      "entropy": 5.896251630783081,
      "epoch": 0.2540690750297737,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499728836371304,
      "loss": 5.827,
      "mean_token_accuracy": 0.12794101238250732,
      "num_tokens": 5626753.0,
      "step": 2560
    },
    {
      "entropy": 5.832917833328247,
      "epoch": 0.25456530369194125,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997270945910151,
      "loss": 5.8598,
      "mean_token_accuracy": 0.12301537916064262,
      "num_tokens": 5636135.0,
      "step": 2565
    },
    {
      "entropy": 5.945082187652588,
      "epoch": 0.2550615323541088,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997253472379785,
      "loss": 5.9181,
      "mean_token_accuracy": 0.1255082905292511,
      "num_tokens": 5647074.0,
      "step": 2570
    },
    {
      "entropy": 5.991348171234131,
      "epoch": 0.2555577610162763,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004997235943122375,
      "loss": 5.9454,
      "mean_token_accuracy": 0.12038756832480431,
      "num_tokens": 5658600.0,
      "step": 2575
    },
    {
      "entropy": 5.845171403884888,
      "epoch": 0.25605398967844384,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997218358138356,
      "loss": 5.8206,
      "mean_token_accuracy": 0.12867578938603402,
      "num_tokens": 5668453.0,
      "step": 2580
    },
    {
      "entropy": 5.950637292861939,
      "epoch": 0.25655021834061137,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997200717428165,
      "loss": 5.9086,
      "mean_token_accuracy": 0.1230459950864315,
      "num_tokens": 5678855.0,
      "step": 2585
    },
    {
      "entropy": 5.937513780593872,
      "epoch": 0.2570464470027789,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997183020992237,
      "loss": 5.9609,
      "mean_token_accuracy": 0.12079612538218498,
      "num_tokens": 5690063.0,
      "step": 2590
    },
    {
      "entropy": 5.882065010070801,
      "epoch": 0.25754267566494643,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997165268831014,
      "loss": 5.8518,
      "mean_token_accuracy": 0.12909654155373573,
      "num_tokens": 5701487.0,
      "step": 2595
    },
    {
      "entropy": 5.876619815826416,
      "epoch": 0.25803890432711396,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997147460944933,
      "loss": 5.8497,
      "mean_token_accuracy": 0.12846047356724738,
      "num_tokens": 5713034.0,
      "step": 2600
    },
    {
      "entropy": 5.968547630310058,
      "epoch": 0.25853513298928144,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997129597334438,
      "loss": 5.8875,
      "mean_token_accuracy": 0.1237348847091198,
      "num_tokens": 5725469.0,
      "step": 2605
    },
    {
      "entropy": 5.886534070968628,
      "epoch": 0.25903136165144897,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997111677999971,
      "loss": 5.8853,
      "mean_token_accuracy": 0.12469863295555114,
      "num_tokens": 5737099.0,
      "step": 2610
    },
    {
      "entropy": 5.934344339370727,
      "epoch": 0.2595275903136165,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997093702941976,
      "loss": 5.8837,
      "mean_token_accuracy": 0.12119583413004875,
      "num_tokens": 5748756.0,
      "step": 2615
    },
    {
      "entropy": 5.952655696868897,
      "epoch": 0.260023818975784,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00049970756721609,
      "loss": 5.8767,
      "mean_token_accuracy": 0.12939024344086647,
      "num_tokens": 5759598.0,
      "step": 2620
    },
    {
      "entropy": 5.8704029560089115,
      "epoch": 0.26052004763795156,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499705758565719,
      "loss": 5.9176,
      "mean_token_accuracy": 0.12523015439510346,
      "num_tokens": 5770397.0,
      "step": 2625
    },
    {
      "entropy": 5.990549516677857,
      "epoch": 0.2610162763001191,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997039443431292,
      "loss": 6.0049,
      "mean_token_accuracy": 0.1170310489833355,
      "num_tokens": 5780935.0,
      "step": 2630
    },
    {
      "entropy": 5.979841566085815,
      "epoch": 0.2615125049622866,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997021245483659,
      "loss": 5.8621,
      "mean_token_accuracy": 0.12379876151680946,
      "num_tokens": 5791557.0,
      "step": 2635
    },
    {
      "entropy": 5.852493476867676,
      "epoch": 0.26200873362445415,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499700299181474,
      "loss": 5.8745,
      "mean_token_accuracy": 0.12755739986896514,
      "num_tokens": 5802233.0,
      "step": 2640
    },
    {
      "entropy": 5.857036972045899,
      "epoch": 0.2625049622866217,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996984682424989,
      "loss": 5.8121,
      "mean_token_accuracy": 0.13497790470719337,
      "num_tokens": 5813424.0,
      "step": 2645
    },
    {
      "entropy": 5.838797950744629,
      "epoch": 0.2630011909487892,
      "grad_norm": 0.7890625,
      "learning_rate": 0.000499696631731486,
      "loss": 5.7849,
      "mean_token_accuracy": 0.12850171998143195,
      "num_tokens": 5825652.0,
      "step": 2650
    },
    {
      "entropy": 5.9219934940338135,
      "epoch": 0.26349741961095674,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996947896484809,
      "loss": 5.882,
      "mean_token_accuracy": 0.12964971140027046,
      "num_tokens": 5836603.0,
      "step": 2655
    },
    {
      "entropy": 5.882232713699341,
      "epoch": 0.26399364827312427,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499692941993529,
      "loss": 5.9252,
      "mean_token_accuracy": 0.12897176370024682,
      "num_tokens": 5849365.0,
      "step": 2660
    },
    {
      "entropy": 5.976453876495361,
      "epoch": 0.2644898769352918,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996910887666765,
      "loss": 5.8506,
      "mean_token_accuracy": 0.130551964789629,
      "num_tokens": 5860305.0,
      "step": 2665
    },
    {
      "entropy": 5.838088703155518,
      "epoch": 0.26498610559745933,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499689229967969,
      "loss": 5.7878,
      "mean_token_accuracy": 0.1278667666018009,
      "num_tokens": 5872272.0,
      "step": 2670
    },
    {
      "entropy": 5.881339263916016,
      "epoch": 0.26548233425962686,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996873655974529,
      "loss": 5.8187,
      "mean_token_accuracy": 0.12291400730609894,
      "num_tokens": 5884107.0,
      "step": 2675
    },
    {
      "entropy": 5.816775846481323,
      "epoch": 0.26597856292179434,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996854956551743,
      "loss": 5.8165,
      "mean_token_accuracy": 0.13490741476416587,
      "num_tokens": 5894060.0,
      "step": 2680
    },
    {
      "entropy": 5.898659563064575,
      "epoch": 0.26647479158396187,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996836201411794,
      "loss": 5.8378,
      "mean_token_accuracy": 0.1263569936156273,
      "num_tokens": 5904804.0,
      "step": 2685
    },
    {
      "entropy": 5.890391683578491,
      "epoch": 0.2669710202461294,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499681739055515,
      "loss": 5.8976,
      "mean_token_accuracy": 0.12535513564944267,
      "num_tokens": 5916496.0,
      "step": 2690
    },
    {
      "entropy": 5.904550838470459,
      "epoch": 0.26746724890829693,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996798523982276,
      "loss": 5.8488,
      "mean_token_accuracy": 0.12368987798690796,
      "num_tokens": 5926920.0,
      "step": 2695
    },
    {
      "entropy": 6.00138578414917,
      "epoch": 0.26796347757046446,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499677960169364,
      "loss": 5.9644,
      "mean_token_accuracy": 0.12470580637454987,
      "num_tokens": 5940098.0,
      "step": 2700
    },
    {
      "entropy": 5.859459924697876,
      "epoch": 0.268459706232632,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996760623689711,
      "loss": 5.8033,
      "mean_token_accuracy": 0.12954767867922784,
      "num_tokens": 5949687.0,
      "step": 2705
    },
    {
      "entropy": 5.925769805908203,
      "epoch": 0.2689559348947995,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499674158997096,
      "loss": 5.9077,
      "mean_token_accuracy": 0.12583340257406234,
      "num_tokens": 5960777.0,
      "step": 2710
    },
    {
      "entropy": 5.945502376556396,
      "epoch": 0.26945216355696705,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996722500537859,
      "loss": 5.8254,
      "mean_token_accuracy": 0.12575791031122208,
      "num_tokens": 5972298.0,
      "step": 2715
    },
    {
      "entropy": 5.855357360839844,
      "epoch": 0.2699483922191346,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996703355390881,
      "loss": 5.9376,
      "mean_token_accuracy": 0.12119691222906112,
      "num_tokens": 5985613.0,
      "step": 2720
    },
    {
      "entropy": 5.991959714889527,
      "epoch": 0.2704446208813021,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049966841545305,
      "loss": 5.9844,
      "mean_token_accuracy": 0.1248553104698658,
      "num_tokens": 5997530.0,
      "step": 2725
    },
    {
      "entropy": 5.943222093582153,
      "epoch": 0.27094084954346964,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996664897957195,
      "loss": 5.8614,
      "mean_token_accuracy": 0.12187350690364837,
      "num_tokens": 6008123.0,
      "step": 2730
    },
    {
      "entropy": 5.869727277755738,
      "epoch": 0.2714370782056372,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996645585671441,
      "loss": 5.8702,
      "mean_token_accuracy": 0.1277486801147461,
      "num_tokens": 6020578.0,
      "step": 2735
    },
    {
      "entropy": 5.862465906143188,
      "epoch": 0.2719333068678047,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996626217673717,
      "loss": 5.7927,
      "mean_token_accuracy": 0.13384850472211837,
      "num_tokens": 6031625.0,
      "step": 2740
    },
    {
      "entropy": 5.851309490203858,
      "epoch": 0.27242953552997223,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996606793964504,
      "loss": 5.8372,
      "mean_token_accuracy": 0.13454477563500405,
      "num_tokens": 6041320.0,
      "step": 2745
    },
    {
      "entropy": 5.875042676925659,
      "epoch": 0.27292576419213976,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996587314544284,
      "loss": 5.8325,
      "mean_token_accuracy": 0.1278717115521431,
      "num_tokens": 6052627.0,
      "step": 2750
    },
    {
      "entropy": 5.901455354690552,
      "epoch": 0.27342199285430724,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996567779413539,
      "loss": 5.8164,
      "mean_token_accuracy": 0.13307981044054032,
      "num_tokens": 6063468.0,
      "step": 2755
    },
    {
      "entropy": 5.850373125076294,
      "epoch": 0.27391822151647477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996548188572755,
      "loss": 5.7943,
      "mean_token_accuracy": 0.134738889336586,
      "num_tokens": 6074208.0,
      "step": 2760
    },
    {
      "entropy": 5.893016290664673,
      "epoch": 0.2744144501786423,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996528542022416,
      "loss": 5.8448,
      "mean_token_accuracy": 0.1275886908173561,
      "num_tokens": 6085068.0,
      "step": 2765
    },
    {
      "entropy": 5.887360763549805,
      "epoch": 0.27491067884080983,
      "grad_norm": 0.875,
      "learning_rate": 0.000499650883976301,
      "loss": 5.8585,
      "mean_token_accuracy": 0.12763815000653267,
      "num_tokens": 6096445.0,
      "step": 2770
    },
    {
      "entropy": 5.812702941894531,
      "epoch": 0.27540690750297736,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996489081795026,
      "loss": 5.7673,
      "mean_token_accuracy": 0.12923398315906526,
      "num_tokens": 6106527.0,
      "step": 2775
    },
    {
      "entropy": 5.8115157127380375,
      "epoch": 0.2759031361651449,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996469268118953,
      "loss": 5.7841,
      "mean_token_accuracy": 0.13443052023649216,
      "num_tokens": 6115923.0,
      "step": 2780
    },
    {
      "entropy": 5.800508975982666,
      "epoch": 0.2763993648273124,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996449398735283,
      "loss": 5.6704,
      "mean_token_accuracy": 0.13257666900753975,
      "num_tokens": 6125861.0,
      "step": 2785
    },
    {
      "entropy": 5.881771564483643,
      "epoch": 0.27689559348947995,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996429473644509,
      "loss": 5.8686,
      "mean_token_accuracy": 0.1287565641105175,
      "num_tokens": 6137189.0,
      "step": 2790
    },
    {
      "entropy": 5.870003747940063,
      "epoch": 0.2773918221516475,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004996409492847124,
      "loss": 5.7946,
      "mean_token_accuracy": 0.129666505753994,
      "num_tokens": 6147920.0,
      "step": 2795
    },
    {
      "entropy": 5.8141978740692135,
      "epoch": 0.277888050813815,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996389456343624,
      "loss": 5.7874,
      "mean_token_accuracy": 0.13501562774181367,
      "num_tokens": 6159371.0,
      "step": 2800
    },
    {
      "entropy": 5.945824146270752,
      "epoch": 0.27838427947598254,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004996369364134507,
      "loss": 5.9822,
      "mean_token_accuracy": 0.12169614806771278,
      "num_tokens": 6172679.0,
      "step": 2805
    },
    {
      "entropy": 5.854594326019287,
      "epoch": 0.2788805081381501,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499634921622027,
      "loss": 5.8441,
      "mean_token_accuracy": 0.12827418893575668,
      "num_tokens": 6184244.0,
      "step": 2810
    },
    {
      "entropy": 5.838413953781128,
      "epoch": 0.2793767368003176,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996329012601413,
      "loss": 5.7281,
      "mean_token_accuracy": 0.12748083770275115,
      "num_tokens": 6194826.0,
      "step": 2815
    },
    {
      "entropy": 5.906148958206177,
      "epoch": 0.27987296546248513,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996308753278437,
      "loss": 5.8572,
      "mean_token_accuracy": 0.12726689577102662,
      "num_tokens": 6206808.0,
      "step": 2820
    },
    {
      "entropy": 5.886915063858032,
      "epoch": 0.28036919412465267,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996288438251843,
      "loss": 5.8789,
      "mean_token_accuracy": 0.12301005497574806,
      "num_tokens": 6217303.0,
      "step": 2825
    },
    {
      "entropy": 5.944884872436523,
      "epoch": 0.28086542278682014,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996268067522137,
      "loss": 5.8766,
      "mean_token_accuracy": 0.12609019502997398,
      "num_tokens": 6227965.0,
      "step": 2830
    },
    {
      "entropy": 5.942210054397583,
      "epoch": 0.28136165144898767,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996247641089823,
      "loss": 5.9284,
      "mean_token_accuracy": 0.12285050153732299,
      "num_tokens": 6239011.0,
      "step": 2835
    },
    {
      "entropy": 5.796023035049439,
      "epoch": 0.2818578801111552,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996227158955409,
      "loss": 5.7826,
      "mean_token_accuracy": 0.13204773068428038,
      "num_tokens": 6249084.0,
      "step": 2840
    },
    {
      "entropy": 5.8865186214447025,
      "epoch": 0.28235410877332273,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049962066211194,
      "loss": 5.8446,
      "mean_token_accuracy": 0.12603942677378654,
      "num_tokens": 6260706.0,
      "step": 2845
    },
    {
      "entropy": 5.841519212722778,
      "epoch": 0.28285033743549026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996186027582308,
      "loss": 5.7821,
      "mean_token_accuracy": 0.13092539235949516,
      "num_tokens": 6272848.0,
      "step": 2850
    },
    {
      "entropy": 5.836033535003662,
      "epoch": 0.2833465660976578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996165378344642,
      "loss": 5.7843,
      "mean_token_accuracy": 0.13042048811912538,
      "num_tokens": 6283820.0,
      "step": 2855
    },
    {
      "entropy": 5.854663515090943,
      "epoch": 0.2838427947598253,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996144673406914,
      "loss": 5.9457,
      "mean_token_accuracy": 0.12040838748216628,
      "num_tokens": 6294193.0,
      "step": 2860
    },
    {
      "entropy": 5.911272716522217,
      "epoch": 0.28433902342199285,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996123912769639,
      "loss": 5.7933,
      "mean_token_accuracy": 0.13677143752574922,
      "num_tokens": 6304407.0,
      "step": 2865
    },
    {
      "entropy": 5.796115159988403,
      "epoch": 0.2848352520841604,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996103096433332,
      "loss": 5.867,
      "mean_token_accuracy": 0.12861735597252846,
      "num_tokens": 6316311.0,
      "step": 2870
    },
    {
      "entropy": 5.917466163635254,
      "epoch": 0.2853314807463279,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996082224398506,
      "loss": 5.8353,
      "mean_token_accuracy": 0.13029164150357248,
      "num_tokens": 6327248.0,
      "step": 2875
    },
    {
      "entropy": 5.865206480026245,
      "epoch": 0.28582770940849545,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996061296665682,
      "loss": 5.8516,
      "mean_token_accuracy": 0.13090819716453553,
      "num_tokens": 6338494.0,
      "step": 2880
    },
    {
      "entropy": 5.866081285476684,
      "epoch": 0.286323938070663,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996040313235377,
      "loss": 5.7413,
      "mean_token_accuracy": 0.1322721965610981,
      "num_tokens": 6349038.0,
      "step": 2885
    },
    {
      "entropy": 5.829685258865356,
      "epoch": 0.2868201667328305,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996019274108112,
      "loss": 5.847,
      "mean_token_accuracy": 0.12434035912156105,
      "num_tokens": 6360180.0,
      "step": 2890
    },
    {
      "entropy": 5.818691205978394,
      "epoch": 0.28731639539499804,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499599817928441,
      "loss": 5.7882,
      "mean_token_accuracy": 0.1332053415477276,
      "num_tokens": 6370931.0,
      "step": 2895
    },
    {
      "entropy": 5.768586444854736,
      "epoch": 0.28781262405716557,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995977028764792,
      "loss": 5.7551,
      "mean_token_accuracy": 0.13535102233290672,
      "num_tokens": 6381810.0,
      "step": 2900
    },
    {
      "entropy": 5.846061563491821,
      "epoch": 0.28830885271933304,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995955822549783,
      "loss": 5.7766,
      "mean_token_accuracy": 0.13201335817575455,
      "num_tokens": 6392268.0,
      "step": 2905
    },
    {
      "entropy": 5.875686454772949,
      "epoch": 0.2888050813815006,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995934560639909,
      "loss": 5.8363,
      "mean_token_accuracy": 0.12547914758324624,
      "num_tokens": 6403650.0,
      "step": 2910
    },
    {
      "entropy": 5.90367488861084,
      "epoch": 0.2893013100436681,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995913243035699,
      "loss": 5.804,
      "mean_token_accuracy": 0.12914050742983818,
      "num_tokens": 6414672.0,
      "step": 2915
    },
    {
      "entropy": 5.841298294067383,
      "epoch": 0.28979753870583563,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995891869737678,
      "loss": 5.8772,
      "mean_token_accuracy": 0.13006574884057046,
      "num_tokens": 6424916.0,
      "step": 2920
    },
    {
      "entropy": 5.897180557250977,
      "epoch": 0.29029376736800316,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995870440746379,
      "loss": 5.8448,
      "mean_token_accuracy": 0.13426648080348969,
      "num_tokens": 6435950.0,
      "step": 2925
    },
    {
      "entropy": 5.772582006454468,
      "epoch": 0.2907899960301707,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995848956062333,
      "loss": 5.7987,
      "mean_token_accuracy": 0.1304109625518322,
      "num_tokens": 6446336.0,
      "step": 2930
    },
    {
      "entropy": 5.867930555343628,
      "epoch": 0.2912862246923382,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995827415686071,
      "loss": 5.8551,
      "mean_token_accuracy": 0.12857231721282006,
      "num_tokens": 6457655.0,
      "step": 2935
    },
    {
      "entropy": 5.871946573257446,
      "epoch": 0.29178245335450576,
      "grad_norm": 1.0,
      "learning_rate": 0.000499580581961813,
      "loss": 5.7452,
      "mean_token_accuracy": 0.13132286220788955,
      "num_tokens": 6469627.0,
      "step": 2940
    },
    {
      "entropy": 5.811709070205689,
      "epoch": 0.2922786820166733,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995784167859043,
      "loss": 5.854,
      "mean_token_accuracy": 0.12665076926350594,
      "num_tokens": 6480351.0,
      "step": 2945
    },
    {
      "entropy": 5.942422008514404,
      "epoch": 0.2927749106788408,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995762460409349,
      "loss": 5.8867,
      "mean_token_accuracy": 0.13129101172089577,
      "num_tokens": 6491651.0,
      "step": 2950
    },
    {
      "entropy": 5.803939962387085,
      "epoch": 0.29327113934100835,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995740697269584,
      "loss": 5.7101,
      "mean_token_accuracy": 0.13372641056776047,
      "num_tokens": 6502531.0,
      "step": 2955
    },
    {
      "entropy": 5.742104434967041,
      "epoch": 0.2937673680031759,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995718878440289,
      "loss": 5.7394,
      "mean_token_accuracy": 0.13551529124379158,
      "num_tokens": 6511938.0,
      "step": 2960
    },
    {
      "entropy": 5.916677808761596,
      "epoch": 0.2942635966653434,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004995697003922006,
      "loss": 5.881,
      "mean_token_accuracy": 0.12271229401230813,
      "num_tokens": 6522775.0,
      "step": 2965
    },
    {
      "entropy": 5.853166818618774,
      "epoch": 0.29475982532751094,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995675073715274,
      "loss": 5.7017,
      "mean_token_accuracy": 0.12792166620492934,
      "num_tokens": 6534660.0,
      "step": 2970
    },
    {
      "entropy": 5.804327535629272,
      "epoch": 0.29525605398967847,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004995653087820641,
      "loss": 5.8612,
      "mean_token_accuracy": 0.12866377979516982,
      "num_tokens": 6545673.0,
      "step": 2975
    },
    {
      "entropy": 5.902544212341309,
      "epoch": 0.29575228265184594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499563104623865,
      "loss": 5.7985,
      "mean_token_accuracy": 0.13151154071092605,
      "num_tokens": 6556912.0,
      "step": 2980
    },
    {
      "entropy": 5.905428886413574,
      "epoch": 0.2962485113140135,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995608948969849,
      "loss": 5.9056,
      "mean_token_accuracy": 0.12208881825208664,
      "num_tokens": 6567532.0,
      "step": 2985
    },
    {
      "entropy": 5.880066728591919,
      "epoch": 0.296744739976181,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995586796014784,
      "loss": 5.8723,
      "mean_token_accuracy": 0.12799570932984353,
      "num_tokens": 6577830.0,
      "step": 2990
    },
    {
      "entropy": 5.8459772109985355,
      "epoch": 0.29724096863834853,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995564587374005,
      "loss": 5.7618,
      "mean_token_accuracy": 0.1300479270517826,
      "num_tokens": 6587254.0,
      "step": 2995
    },
    {
      "entropy": 5.901417541503906,
      "epoch": 0.29773719730051607,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995542323048063,
      "loss": 5.9346,
      "mean_token_accuracy": 0.12555868849158286,
      "num_tokens": 6598294.0,
      "step": 3000
    },
    {
      "epoch": 0.29773719730051607,
      "eval_entropy": 5.737077051513012,
      "eval_loss": 5.8997344970703125,
      "eval_mean_token_accuracy": 0.1301897406935635,
      "eval_num_tokens": 6598294.0,
      "eval_runtime": 26.6497,
      "eval_samples_per_second": 1263.389,
      "eval_steps_per_second": 157.938,
      "step": 3000
    },
    {
      "entropy": 5.908546400070191,
      "epoch": 0.2982334259626836,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499552000303751,
      "loss": 5.8254,
      "mean_token_accuracy": 0.13424069806933403,
      "num_tokens": 6608772.0,
      "step": 3005
    },
    {
      "entropy": 5.863250255584717,
      "epoch": 0.2987296546248511,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049954976273429,
      "loss": 5.8225,
      "mean_token_accuracy": 0.13384647592902182,
      "num_tokens": 6619973.0,
      "step": 3010
    },
    {
      "entropy": 5.848355102539062,
      "epoch": 0.29922588328701866,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995475195964787,
      "loss": 5.856,
      "mean_token_accuracy": 0.1277636930346489,
      "num_tokens": 6631172.0,
      "step": 3015
    },
    {
      "entropy": 5.830797290802002,
      "epoch": 0.2997221119491862,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995452708903727,
      "loss": 5.8381,
      "mean_token_accuracy": 0.1252981200814247,
      "num_tokens": 6641343.0,
      "step": 3020
    },
    {
      "entropy": 5.853375816345215,
      "epoch": 0.3002183406113537,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499543016616028,
      "loss": 5.7881,
      "mean_token_accuracy": 0.12906054332852362,
      "num_tokens": 6652892.0,
      "step": 3025
    },
    {
      "entropy": 5.813239049911499,
      "epoch": 0.30071456927352125,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995407567735002,
      "loss": 5.8103,
      "mean_token_accuracy": 0.1291997916996479,
      "num_tokens": 6664177.0,
      "step": 3030
    },
    {
      "entropy": 5.8448244571685795,
      "epoch": 0.3012107979356888,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995384913628456,
      "loss": 5.7806,
      "mean_token_accuracy": 0.12814074456691743,
      "num_tokens": 6674402.0,
      "step": 3035
    },
    {
      "entropy": 5.886237621307373,
      "epoch": 0.3017070265978563,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995362203841202,
      "loss": 5.9118,
      "mean_token_accuracy": 0.12575332671403885,
      "num_tokens": 6684886.0,
      "step": 3040
    },
    {
      "entropy": 5.884566354751587,
      "epoch": 0.30220325526002384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995339438373804,
      "loss": 5.7383,
      "mean_token_accuracy": 0.1311859056353569,
      "num_tokens": 6694774.0,
      "step": 3045
    },
    {
      "entropy": 5.778323936462402,
      "epoch": 0.30269948392219137,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995316617226825,
      "loss": 5.8032,
      "mean_token_accuracy": 0.13333578556776046,
      "num_tokens": 6705346.0,
      "step": 3050
    },
    {
      "entropy": 5.849145555496216,
      "epoch": 0.30319571258435885,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995293740400833,
      "loss": 5.7601,
      "mean_token_accuracy": 0.1354569710791111,
      "num_tokens": 6715579.0,
      "step": 3055
    },
    {
      "entropy": 5.777801465988159,
      "epoch": 0.3036919412465264,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995270807896395,
      "loss": 5.7179,
      "mean_token_accuracy": 0.1327461652457714,
      "num_tokens": 6726382.0,
      "step": 3060
    },
    {
      "entropy": 5.833434057235718,
      "epoch": 0.3041881699086939,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995247819714079,
      "loss": 5.7975,
      "mean_token_accuracy": 0.12714954614639282,
      "num_tokens": 6737245.0,
      "step": 3065
    },
    {
      "entropy": 5.905564641952514,
      "epoch": 0.30468439857086144,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995224775854454,
      "loss": 6.0002,
      "mean_token_accuracy": 0.12385166138410568,
      "num_tokens": 6749113.0,
      "step": 3070
    },
    {
      "entropy": 5.978738498687744,
      "epoch": 0.30518062723302897,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995201676318093,
      "loss": 5.8518,
      "mean_token_accuracy": 0.13237554132938384,
      "num_tokens": 6761987.0,
      "step": 3075
    },
    {
      "entropy": 5.816679763793945,
      "epoch": 0.3056768558951965,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995178521105568,
      "loss": 5.7908,
      "mean_token_accuracy": 0.1237344652414322,
      "num_tokens": 6771279.0,
      "step": 3080
    },
    {
      "entropy": 5.852257347106933,
      "epoch": 0.30617308455736403,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995155310217454,
      "loss": 5.7286,
      "mean_token_accuracy": 0.13410176187753678,
      "num_tokens": 6781038.0,
      "step": 3085
    },
    {
      "entropy": 5.75889139175415,
      "epoch": 0.30666931321953156,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995132043654327,
      "loss": 5.8619,
      "mean_token_accuracy": 0.12660413309931756,
      "num_tokens": 6792686.0,
      "step": 3090
    },
    {
      "entropy": 5.84913330078125,
      "epoch": 0.3071655418816991,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995108721416763,
      "loss": 5.8304,
      "mean_token_accuracy": 0.1372841201722622,
      "num_tokens": 6804414.0,
      "step": 3095
    },
    {
      "entropy": 5.846239280700684,
      "epoch": 0.3076617705438666,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995085343505339,
      "loss": 5.7936,
      "mean_token_accuracy": 0.13219277635216714,
      "num_tokens": 6816180.0,
      "step": 3100
    },
    {
      "entropy": 5.857756471633911,
      "epoch": 0.30815799920603415,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995061909920638,
      "loss": 5.7985,
      "mean_token_accuracy": 0.12769369333982467,
      "num_tokens": 6827114.0,
      "step": 3105
    },
    {
      "entropy": 5.881041431427002,
      "epoch": 0.3086542278682017,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995038420663238,
      "loss": 5.8539,
      "mean_token_accuracy": 0.13202394172549248,
      "num_tokens": 6836953.0,
      "step": 3110
    },
    {
      "entropy": 5.872609329223633,
      "epoch": 0.3091504565303692,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995014875733724,
      "loss": 5.9082,
      "mean_token_accuracy": 0.12458430081605912,
      "num_tokens": 6847963.0,
      "step": 3115
    },
    {
      "entropy": 5.93118109703064,
      "epoch": 0.30964668519253674,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994991275132678,
      "loss": 5.9155,
      "mean_token_accuracy": 0.12210882604122161,
      "num_tokens": 6858264.0,
      "step": 3120
    },
    {
      "entropy": 5.832281446456909,
      "epoch": 0.31014291385470427,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994967618860685,
      "loss": 5.7974,
      "mean_token_accuracy": 0.13236498534679414,
      "num_tokens": 6868861.0,
      "step": 3125
    },
    {
      "entropy": 5.914390993118286,
      "epoch": 0.31063914251687175,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994943906918333,
      "loss": 5.8243,
      "mean_token_accuracy": 0.13056366965174676,
      "num_tokens": 6879797.0,
      "step": 3130
    },
    {
      "entropy": 5.8452612400054935,
      "epoch": 0.3111353711790393,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994920139306211,
      "loss": 5.7794,
      "mean_token_accuracy": 0.13094358667731285,
      "num_tokens": 6891323.0,
      "step": 3135
    },
    {
      "entropy": 5.814275884628296,
      "epoch": 0.3116315998412068,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994896316024906,
      "loss": 5.8629,
      "mean_token_accuracy": 0.12925291806459427,
      "num_tokens": 6903345.0,
      "step": 3140
    },
    {
      "entropy": 5.829950189590454,
      "epoch": 0.31212782850337434,
      "grad_norm": 1.0,
      "learning_rate": 0.000499487243707501,
      "loss": 5.787,
      "mean_token_accuracy": 0.12979979440569878,
      "num_tokens": 6914342.0,
      "step": 3145
    },
    {
      "entropy": 5.8646656513214115,
      "epoch": 0.31262405716554187,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004994848502457114,
      "loss": 5.8328,
      "mean_token_accuracy": 0.12842394262552262,
      "num_tokens": 6926691.0,
      "step": 3150
    },
    {
      "entropy": 5.829729843139648,
      "epoch": 0.3131202858277094,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994824512171812,
      "loss": 5.7117,
      "mean_token_accuracy": 0.13195387348532678,
      "num_tokens": 6937841.0,
      "step": 3155
    },
    {
      "entropy": 5.766626453399658,
      "epoch": 0.31361651448987693,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994800466219701,
      "loss": 5.8114,
      "mean_token_accuracy": 0.12932611480355263,
      "num_tokens": 6948601.0,
      "step": 3160
    },
    {
      "entropy": 5.855544185638427,
      "epoch": 0.31411274315204446,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994776364601374,
      "loss": 5.7358,
      "mean_token_accuracy": 0.13258148804306985,
      "num_tokens": 6959775.0,
      "step": 3165
    },
    {
      "entropy": 5.78338475227356,
      "epoch": 0.314608971814212,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994752207317431,
      "loss": 5.7707,
      "mean_token_accuracy": 0.13972898945212364,
      "num_tokens": 6969202.0,
      "step": 3170
    },
    {
      "entropy": 5.805487060546875,
      "epoch": 0.3151052004763795,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499472799436847,
      "loss": 5.7517,
      "mean_token_accuracy": 0.13298066705465317,
      "num_tokens": 6978633.0,
      "step": 3175
    },
    {
      "entropy": 5.894915008544922,
      "epoch": 0.31560142913854705,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994703725755092,
      "loss": 5.8186,
      "mean_token_accuracy": 0.12706844583153726,
      "num_tokens": 6989638.0,
      "step": 3180
    },
    {
      "entropy": 5.812424945831299,
      "epoch": 0.3160976578007146,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994679401477899,
      "loss": 5.7826,
      "mean_token_accuracy": 0.12963953241705894,
      "num_tokens": 7001606.0,
      "step": 3185
    },
    {
      "entropy": 5.813755512237549,
      "epoch": 0.3165938864628821,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994655021537492,
      "loss": 5.7604,
      "mean_token_accuracy": 0.1284653961658478,
      "num_tokens": 7012915.0,
      "step": 3190
    },
    {
      "entropy": 5.80635085105896,
      "epoch": 0.31709011512504964,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994630585934478,
      "loss": 5.8141,
      "mean_token_accuracy": 0.12510425895452498,
      "num_tokens": 7023191.0,
      "step": 3195
    },
    {
      "entropy": 5.844677400588989,
      "epoch": 0.3175863437872172,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994606094669463,
      "loss": 5.708,
      "mean_token_accuracy": 0.13605407029390335,
      "num_tokens": 7032506.0,
      "step": 3200
    },
    {
      "entropy": 5.819374847412109,
      "epoch": 0.3180825724493847,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994581547743053,
      "loss": 5.8462,
      "mean_token_accuracy": 0.12448615580797195,
      "num_tokens": 7046098.0,
      "step": 3205
    },
    {
      "entropy": 5.952435874938965,
      "epoch": 0.3185788011115522,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994556945155857,
      "loss": 5.8499,
      "mean_token_accuracy": 0.13355428129434585,
      "num_tokens": 7057434.0,
      "step": 3210
    },
    {
      "entropy": 5.792888545989991,
      "epoch": 0.3190750297737197,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994532286908486,
      "loss": 5.7786,
      "mean_token_accuracy": 0.12661978080868722,
      "num_tokens": 7067674.0,
      "step": 3215
    },
    {
      "entropy": 5.934251117706299,
      "epoch": 0.31957125843588724,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499450757300155,
      "loss": 5.8491,
      "mean_token_accuracy": 0.12803201824426652,
      "num_tokens": 7077898.0,
      "step": 3220
    },
    {
      "entropy": 5.873095369338989,
      "epoch": 0.32006748709805477,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994482803435664,
      "loss": 5.8114,
      "mean_token_accuracy": 0.12614069283008575,
      "num_tokens": 7088566.0,
      "step": 3225
    },
    {
      "entropy": 5.777619647979736,
      "epoch": 0.3205637157602223,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499445797821144,
      "loss": 5.8038,
      "mean_token_accuracy": 0.13038980588316917,
      "num_tokens": 7099533.0,
      "step": 3230
    },
    {
      "entropy": 5.901455020904541,
      "epoch": 0.32105994442238983,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994433097329494,
      "loss": 5.8054,
      "mean_token_accuracy": 0.12909889370203018,
      "num_tokens": 7112254.0,
      "step": 3235
    },
    {
      "entropy": 5.893978786468506,
      "epoch": 0.32155617308455736,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994408160790444,
      "loss": 5.8276,
      "mean_token_accuracy": 0.12560412287712097,
      "num_tokens": 7124704.0,
      "step": 3240
    },
    {
      "entropy": 5.970049476623535,
      "epoch": 0.3220524017467249,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994383168594907,
      "loss": 5.9963,
      "mean_token_accuracy": 0.11985118389129638,
      "num_tokens": 7137741.0,
      "step": 3245
    },
    {
      "entropy": 5.848871469497681,
      "epoch": 0.3225486304088924,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994358120743506,
      "loss": 5.7898,
      "mean_token_accuracy": 0.12539106085896493,
      "num_tokens": 7147888.0,
      "step": 3250
    },
    {
      "entropy": 5.752130508422852,
      "epoch": 0.32304485907105995,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994333017236858,
      "loss": 5.7357,
      "mean_token_accuracy": 0.13781259804964066,
      "num_tokens": 7158950.0,
      "step": 3255
    },
    {
      "entropy": 5.861024618148804,
      "epoch": 0.3235410877332275,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994307858075587,
      "loss": 5.8815,
      "mean_token_accuracy": 0.12370717674493789,
      "num_tokens": 7170156.0,
      "step": 3260
    },
    {
      "entropy": 5.8257866382598875,
      "epoch": 0.324037316395395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994282643260319,
      "loss": 5.7285,
      "mean_token_accuracy": 0.1357402391731739,
      "num_tokens": 7180576.0,
      "step": 3265
    },
    {
      "entropy": 5.768695497512818,
      "epoch": 0.32453354505756254,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994257372791676,
      "loss": 5.7863,
      "mean_token_accuracy": 0.13116331100463868,
      "num_tokens": 7192380.0,
      "step": 3270
    },
    {
      "entropy": 5.890905237197876,
      "epoch": 0.3250297737197301,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994232046670286,
      "loss": 5.8419,
      "mean_token_accuracy": 0.1313177950680256,
      "num_tokens": 7203898.0,
      "step": 3275
    },
    {
      "entropy": 5.896629285812378,
      "epoch": 0.3255260023818976,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994206664896778,
      "loss": 5.8155,
      "mean_token_accuracy": 0.1287803702056408,
      "num_tokens": 7215241.0,
      "step": 3280
    },
    {
      "entropy": 5.85573182106018,
      "epoch": 0.3260222310440651,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994181227471779,
      "loss": 5.7653,
      "mean_token_accuracy": 0.1314986191689968,
      "num_tokens": 7226039.0,
      "step": 3285
    },
    {
      "entropy": 5.7266041278839115,
      "epoch": 0.3265184597062326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994155734395923,
      "loss": 5.79,
      "mean_token_accuracy": 0.13011762425303458,
      "num_tokens": 7237148.0,
      "step": 3290
    },
    {
      "entropy": 5.825259494781494,
      "epoch": 0.32701468836840014,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499413018566984,
      "loss": 5.7883,
      "mean_token_accuracy": 0.1279461018741131,
      "num_tokens": 7249284.0,
      "step": 3295
    },
    {
      "entropy": 5.842443895339966,
      "epoch": 0.32751091703056767,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994104581294163,
      "loss": 5.7961,
      "mean_token_accuracy": 0.12790419161319733,
      "num_tokens": 7259967.0,
      "step": 3300
    },
    {
      "entropy": 5.762500810623169,
      "epoch": 0.3280071456927352,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994078921269529,
      "loss": 5.7547,
      "mean_token_accuracy": 0.13474935069680213,
      "num_tokens": 7271145.0,
      "step": 3305
    },
    {
      "entropy": 5.918145751953125,
      "epoch": 0.32850337435490273,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994053205596573,
      "loss": 5.7698,
      "mean_token_accuracy": 0.1367741271853447,
      "num_tokens": 7281872.0,
      "step": 3310
    },
    {
      "entropy": 5.750212907791138,
      "epoch": 0.32899960301707026,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994027434275934,
      "loss": 5.788,
      "mean_token_accuracy": 0.12981380969285966,
      "num_tokens": 7293330.0,
      "step": 3315
    },
    {
      "entropy": 5.967539882659912,
      "epoch": 0.3294958316792378,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994001607308249,
      "loss": 5.8512,
      "mean_token_accuracy": 0.12412105202674865,
      "num_tokens": 7304187.0,
      "step": 3320
    },
    {
      "entropy": 5.885866308212281,
      "epoch": 0.3299920603414053,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499397572469416,
      "loss": 5.7521,
      "mean_token_accuracy": 0.13047175034880637,
      "num_tokens": 7314756.0,
      "step": 3325
    },
    {
      "entropy": 5.751908969879151,
      "epoch": 0.33048828900357285,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993949786434308,
      "loss": 5.7817,
      "mean_token_accuracy": 0.1319439634680748,
      "num_tokens": 7325711.0,
      "step": 3330
    },
    {
      "entropy": 5.782564353942871,
      "epoch": 0.3309845176657404,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993923792529337,
      "loss": 5.7167,
      "mean_token_accuracy": 0.13527547493577002,
      "num_tokens": 7335639.0,
      "step": 3335
    },
    {
      "entropy": 5.883091735839844,
      "epoch": 0.3314807463279079,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993897742979892,
      "loss": 5.8349,
      "mean_token_accuracy": 0.1300416111946106,
      "num_tokens": 7346621.0,
      "step": 3340
    },
    {
      "entropy": 5.872514009475708,
      "epoch": 0.33197697499007545,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993871637786616,
      "loss": 5.7901,
      "mean_token_accuracy": 0.12466496452689171,
      "num_tokens": 7358950.0,
      "step": 3345
    },
    {
      "entropy": 5.732246828079224,
      "epoch": 0.332473203652243,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499384547695016,
      "loss": 5.6516,
      "mean_token_accuracy": 0.13917058929800988,
      "num_tokens": 7369629.0,
      "step": 3350
    },
    {
      "entropy": 5.782614231109619,
      "epoch": 0.3329694323144105,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499381926047117,
      "loss": 5.7491,
      "mean_token_accuracy": 0.12965478897094726,
      "num_tokens": 7379853.0,
      "step": 3355
    },
    {
      "entropy": 5.828644323348999,
      "epoch": 0.333465660976578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049937929883503,
      "loss": 5.7902,
      "mean_token_accuracy": 0.12588309049606322,
      "num_tokens": 7391323.0,
      "step": 3360
    },
    {
      "entropy": 5.824267625808716,
      "epoch": 0.3339618896387455,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004993766660588196,
      "loss": 5.7624,
      "mean_token_accuracy": 0.1319116823375225,
      "num_tokens": 7402201.0,
      "step": 3365
    },
    {
      "entropy": 5.922031736373901,
      "epoch": 0.33445811830091304,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004993740277185515,
      "loss": 5.9897,
      "mean_token_accuracy": 0.12059160396456718,
      "num_tokens": 7413783.0,
      "step": 3370
    },
    {
      "entropy": 5.8004532814025875,
      "epoch": 0.3349543469630806,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499371383814291,
      "loss": 5.7574,
      "mean_token_accuracy": 0.13145501986145974,
      "num_tokens": 7425909.0,
      "step": 3375
    },
    {
      "entropy": 5.759319877624511,
      "epoch": 0.3354505756252481,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993687343461036,
      "loss": 5.6652,
      "mean_token_accuracy": 0.13739446327090263,
      "num_tokens": 7436346.0,
      "step": 3380
    },
    {
      "entropy": 5.760404396057129,
      "epoch": 0.33594680428741563,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993660793140551,
      "loss": 5.7156,
      "mean_token_accuracy": 0.13563819900155066,
      "num_tokens": 7446596.0,
      "step": 3385
    },
    {
      "entropy": 5.784324502944946,
      "epoch": 0.33644303294958317,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993634187182113,
      "loss": 5.806,
      "mean_token_accuracy": 0.1257621392607689,
      "num_tokens": 7457697.0,
      "step": 3390
    },
    {
      "entropy": 5.917649269104004,
      "epoch": 0.3369392616117507,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993607525586381,
      "loss": 5.9285,
      "mean_token_accuracy": 0.1271599940955639,
      "num_tokens": 7468541.0,
      "step": 3395
    },
    {
      "entropy": 5.8661158084869385,
      "epoch": 0.3374354902739182,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993580808354018,
      "loss": 5.8033,
      "mean_token_accuracy": 0.13219190388917923,
      "num_tokens": 7479781.0,
      "step": 3400
    },
    {
      "entropy": 5.800511312484741,
      "epoch": 0.33793171893608576,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993554035485685,
      "loss": 5.7042,
      "mean_token_accuracy": 0.12884424477815629,
      "num_tokens": 7490722.0,
      "step": 3405
    },
    {
      "entropy": 5.8494329929351805,
      "epoch": 0.3384279475982533,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993527206982047,
      "loss": 5.8264,
      "mean_token_accuracy": 0.12879022359848022,
      "num_tokens": 7502293.0,
      "step": 3410
    },
    {
      "entropy": 5.8656847953796385,
      "epoch": 0.3389241762604208,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004993500322843767,
      "loss": 5.7507,
      "mean_token_accuracy": 0.13226759135723115,
      "num_tokens": 7513846.0,
      "step": 3415
    },
    {
      "entropy": 5.711007070541382,
      "epoch": 0.33942040492258835,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993473383071514,
      "loss": 5.7865,
      "mean_token_accuracy": 0.13022084534168243,
      "num_tokens": 7524062.0,
      "step": 3420
    },
    {
      "entropy": 5.862089157104492,
      "epoch": 0.3399166335847559,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993446387665956,
      "loss": 5.8329,
      "mean_token_accuracy": 0.13513429313898087,
      "num_tokens": 7535533.0,
      "step": 3425
    },
    {
      "entropy": 5.865432071685791,
      "epoch": 0.3404128622469234,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993419336627761,
      "loss": 5.7147,
      "mean_token_accuracy": 0.13362642973661423,
      "num_tokens": 7545951.0,
      "step": 3430
    },
    {
      "entropy": 5.711605072021484,
      "epoch": 0.3409090909090909,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993392229957602,
      "loss": 5.671,
      "mean_token_accuracy": 0.1315312497317791,
      "num_tokens": 7557320.0,
      "step": 3435
    },
    {
      "entropy": 5.690997362136841,
      "epoch": 0.3414053195712584,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993365067656148,
      "loss": 5.7777,
      "mean_token_accuracy": 0.1280876100063324,
      "num_tokens": 7568078.0,
      "step": 3440
    },
    {
      "entropy": 5.824246025085449,
      "epoch": 0.34190154823342594,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993337849724076,
      "loss": 5.7856,
      "mean_token_accuracy": 0.1305572994053364,
      "num_tokens": 7578940.0,
      "step": 3445
    },
    {
      "entropy": 5.8472373485565186,
      "epoch": 0.3423977768955935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993310576162059,
      "loss": 5.7803,
      "mean_token_accuracy": 0.13081290274858476,
      "num_tokens": 7589864.0,
      "step": 3450
    },
    {
      "entropy": 5.871004867553711,
      "epoch": 0.342894005557761,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993283246970772,
      "loss": 5.725,
      "mean_token_accuracy": 0.12843016013503075,
      "num_tokens": 7599795.0,
      "step": 3455
    },
    {
      "entropy": 5.8094720363616945,
      "epoch": 0.34339023421992854,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993255862150896,
      "loss": 5.7121,
      "mean_token_accuracy": 0.12965178340673447,
      "num_tokens": 7610056.0,
      "step": 3460
    },
    {
      "entropy": 5.721919441223145,
      "epoch": 0.34388646288209607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993228421703108,
      "loss": 5.7053,
      "mean_token_accuracy": 0.1275747649371624,
      "num_tokens": 7620218.0,
      "step": 3465
    },
    {
      "entropy": 5.75278491973877,
      "epoch": 0.3443826915442636,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993200925628088,
      "loss": 5.8105,
      "mean_token_accuracy": 0.1322753168642521,
      "num_tokens": 7632781.0,
      "step": 3470
    },
    {
      "entropy": 5.820105791091919,
      "epoch": 0.3448789202064311,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499317337392652,
      "loss": 5.7817,
      "mean_token_accuracy": 0.12995077893137932,
      "num_tokens": 7643845.0,
      "step": 3475
    },
    {
      "entropy": 5.8669881343841555,
      "epoch": 0.34537514886859866,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993145766599085,
      "loss": 5.8008,
      "mean_token_accuracy": 0.13089126273989676,
      "num_tokens": 7654583.0,
      "step": 3480
    },
    {
      "entropy": 5.7667046070098875,
      "epoch": 0.3458713775307662,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993118103646469,
      "loss": 5.6611,
      "mean_token_accuracy": 0.1325427420437336,
      "num_tokens": 7665714.0,
      "step": 3485
    },
    {
      "entropy": 5.769314384460449,
      "epoch": 0.3463676061929337,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993090385069357,
      "loss": 5.7067,
      "mean_token_accuracy": 0.1339074306190014,
      "num_tokens": 7676011.0,
      "step": 3490
    },
    {
      "entropy": 5.743018531799317,
      "epoch": 0.34686383485510125,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993062610868437,
      "loss": 5.763,
      "mean_token_accuracy": 0.1294798120856285,
      "num_tokens": 7686927.0,
      "step": 3495
    },
    {
      "entropy": 5.764451122283935,
      "epoch": 0.3473600635172688,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993034781044398,
      "loss": 5.7245,
      "mean_token_accuracy": 0.1329188510775566,
      "num_tokens": 7698419.0,
      "step": 3500
    },
    {
      "entropy": 5.829875087738037,
      "epoch": 0.3478562921794363,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499300689559793,
      "loss": 5.6837,
      "mean_token_accuracy": 0.13597684502601623,
      "num_tokens": 7709695.0,
      "step": 3505
    },
    {
      "entropy": 5.726664352416992,
      "epoch": 0.3483525208416038,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004992978954529724,
      "loss": 5.6869,
      "mean_token_accuracy": 0.136630592495203,
      "num_tokens": 7721309.0,
      "step": 3510
    },
    {
      "entropy": 5.803768014907837,
      "epoch": 0.3488487495037713,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992950957840473,
      "loss": 5.7554,
      "mean_token_accuracy": 0.132266054302454,
      "num_tokens": 7732074.0,
      "step": 3515
    },
    {
      "entropy": 5.816020488739014,
      "epoch": 0.34934497816593885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499292290553087,
      "loss": 5.8194,
      "mean_token_accuracy": 0.13066235557198524,
      "num_tokens": 7742869.0,
      "step": 3520
    },
    {
      "entropy": 5.7433083057403564,
      "epoch": 0.3498412068281064,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992894797601614,
      "loss": 5.7274,
      "mean_token_accuracy": 0.13449874967336656,
      "num_tokens": 7754895.0,
      "step": 3525
    },
    {
      "entropy": 5.830230379104615,
      "epoch": 0.3503374354902739,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049928666340534,
      "loss": 5.7012,
      "mean_token_accuracy": 0.12928271517157555,
      "num_tokens": 7765262.0,
      "step": 3530
    },
    {
      "entropy": 5.747873067855835,
      "epoch": 0.35083366415244144,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992838414886925,
      "loss": 5.7814,
      "mean_token_accuracy": 0.12928831875324248,
      "num_tokens": 7777348.0,
      "step": 3535
    },
    {
      "entropy": 5.870401191711426,
      "epoch": 0.35132989281460897,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992810140102891,
      "loss": 5.7629,
      "mean_token_accuracy": 0.13162485882639885,
      "num_tokens": 7788161.0,
      "step": 3540
    },
    {
      "entropy": 5.795786046981812,
      "epoch": 0.3518261214767765,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992781809701999,
      "loss": 5.7618,
      "mean_token_accuracy": 0.12922549918293952,
      "num_tokens": 7798629.0,
      "step": 3545
    },
    {
      "entropy": 5.7486223697662355,
      "epoch": 0.35232235013894403,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992753423684951,
      "loss": 5.7428,
      "mean_token_accuracy": 0.13399974927306174,
      "num_tokens": 7810148.0,
      "step": 3550
    },
    {
      "entropy": 5.806520700454712,
      "epoch": 0.35281857880111156,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992724982052449,
      "loss": 5.732,
      "mean_token_accuracy": 0.13214099556207656,
      "num_tokens": 7821667.0,
      "step": 3555
    },
    {
      "entropy": 5.731984186172485,
      "epoch": 0.3533148074632791,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049926964848052,
      "loss": 5.7207,
      "mean_token_accuracy": 0.13429917991161347,
      "num_tokens": 7832184.0,
      "step": 3560
    },
    {
      "entropy": 5.8297947406768795,
      "epoch": 0.3538110361254466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992667931943913,
      "loss": 5.7387,
      "mean_token_accuracy": 0.130974131077528,
      "num_tokens": 7842983.0,
      "step": 3565
    },
    {
      "entropy": 5.766924142837524,
      "epoch": 0.35430726478761415,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992639323469293,
      "loss": 5.6723,
      "mean_token_accuracy": 0.1407732456922531,
      "num_tokens": 7854783.0,
      "step": 3570
    },
    {
      "entropy": 5.724614095687866,
      "epoch": 0.3548034934497817,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499261065938205,
      "loss": 5.695,
      "mean_token_accuracy": 0.1421818196773529,
      "num_tokens": 7864684.0,
      "step": 3575
    },
    {
      "entropy": 5.8322038650512695,
      "epoch": 0.3552997221119492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992581939682895,
      "loss": 5.8576,
      "mean_token_accuracy": 0.12228314802050591,
      "num_tokens": 7876846.0,
      "step": 3580
    },
    {
      "entropy": 5.905322122573852,
      "epoch": 0.3557959507741167,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499255316437254,
      "loss": 5.7116,
      "mean_token_accuracy": 0.1306779131293297,
      "num_tokens": 7886826.0,
      "step": 3585
    },
    {
      "entropy": 5.699662113189698,
      "epoch": 0.3562921794362842,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992524333451699,
      "loss": 5.7923,
      "mean_token_accuracy": 0.1284032307565212,
      "num_tokens": 7898941.0,
      "step": 3590
    },
    {
      "entropy": 5.777550268173218,
      "epoch": 0.35678840809845175,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992495446921087,
      "loss": 5.6991,
      "mean_token_accuracy": 0.1283227153122425,
      "num_tokens": 7909916.0,
      "step": 3595
    },
    {
      "entropy": 5.718798351287842,
      "epoch": 0.3572846367606193,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992466504781419,
      "loss": 5.723,
      "mean_token_accuracy": 0.13592231124639512,
      "num_tokens": 7922137.0,
      "step": 3600
    },
    {
      "entropy": 5.725904083251953,
      "epoch": 0.3577808654227868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992437507033415,
      "loss": 5.6544,
      "mean_token_accuracy": 0.13926874473690987,
      "num_tokens": 7932940.0,
      "step": 3605
    },
    {
      "entropy": 5.8571995258331295,
      "epoch": 0.35827709408495434,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992408453677792,
      "loss": 5.8249,
      "mean_token_accuracy": 0.12425437346100807,
      "num_tokens": 7943735.0,
      "step": 3610
    },
    {
      "entropy": 5.761858224868774,
      "epoch": 0.35877332274712187,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992379344715273,
      "loss": 5.7084,
      "mean_token_accuracy": 0.13300363793969155,
      "num_tokens": 7955323.0,
      "step": 3615
    },
    {
      "entropy": 5.747606563568115,
      "epoch": 0.3592695514092894,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992350180146577,
      "loss": 5.656,
      "mean_token_accuracy": 0.13175439611077308,
      "num_tokens": 7966073.0,
      "step": 3620
    },
    {
      "entropy": 5.769119215011597,
      "epoch": 0.35976578007145693,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992320959972428,
      "loss": 5.8377,
      "mean_token_accuracy": 0.12840200886130332,
      "num_tokens": 7977027.0,
      "step": 3625
    },
    {
      "entropy": 5.802391815185547,
      "epoch": 0.36026200873362446,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992291684193551,
      "loss": 5.6963,
      "mean_token_accuracy": 0.1378728047013283,
      "num_tokens": 7987368.0,
      "step": 3630
    },
    {
      "entropy": 5.810581016540527,
      "epoch": 0.360758237395792,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992262352810672,
      "loss": 5.7589,
      "mean_token_accuracy": 0.13187605291604995,
      "num_tokens": 8000060.0,
      "step": 3635
    },
    {
      "entropy": 5.778606843948364,
      "epoch": 0.3612544660579595,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992232965824518,
      "loss": 5.8151,
      "mean_token_accuracy": 0.1257768027484417,
      "num_tokens": 8011388.0,
      "step": 3640
    },
    {
      "entropy": 5.7404289722442625,
      "epoch": 0.36175069472012705,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499220352323582,
      "loss": 5.6808,
      "mean_token_accuracy": 0.13884039595723152,
      "num_tokens": 8021682.0,
      "step": 3645
    },
    {
      "entropy": 5.798848342895508,
      "epoch": 0.3622469233822946,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992174025045304,
      "loss": 5.6941,
      "mean_token_accuracy": 0.13487540781497956,
      "num_tokens": 8032799.0,
      "step": 3650
    },
    {
      "entropy": 5.706443929672242,
      "epoch": 0.3627431520444621,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992144471253704,
      "loss": 5.5725,
      "mean_token_accuracy": 0.14214823096990586,
      "num_tokens": 8042742.0,
      "step": 3655
    },
    {
      "entropy": 5.763505363464356,
      "epoch": 0.3632393807066296,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992114861861753,
      "loss": 5.7613,
      "mean_token_accuracy": 0.13013699501752854,
      "num_tokens": 8054068.0,
      "step": 3660
    },
    {
      "entropy": 5.813911485671997,
      "epoch": 0.3637356093687971,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992085196870184,
      "loss": 5.7419,
      "mean_token_accuracy": 0.1301536478102207,
      "num_tokens": 8065286.0,
      "step": 3665
    },
    {
      "entropy": 5.763180732727051,
      "epoch": 0.36423183803096465,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992055476279734,
      "loss": 5.7993,
      "mean_token_accuracy": 0.12314197868108749,
      "num_tokens": 8077122.0,
      "step": 3670
    },
    {
      "entropy": 5.856541395187378,
      "epoch": 0.3647280666931322,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992025700091139,
      "loss": 5.7467,
      "mean_token_accuracy": 0.1376276656985283,
      "num_tokens": 8087036.0,
      "step": 3675
    },
    {
      "entropy": 5.808696556091308,
      "epoch": 0.3652242953552997,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991995868305138,
      "loss": 5.7756,
      "mean_token_accuracy": 0.128545381128788,
      "num_tokens": 8098517.0,
      "step": 3680
    },
    {
      "entropy": 5.738351106643677,
      "epoch": 0.36572052401746724,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991965980922471,
      "loss": 5.6491,
      "mean_token_accuracy": 0.13984912484884263,
      "num_tokens": 8107475.0,
      "step": 3685
    },
    {
      "entropy": 5.764809942245483,
      "epoch": 0.36621675267963477,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991936037943878,
      "loss": 5.7499,
      "mean_token_accuracy": 0.13070235177874565,
      "num_tokens": 8118529.0,
      "step": 3690
    },
    {
      "entropy": 5.828148555755615,
      "epoch": 0.3667129813418023,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991906039370104,
      "loss": 5.6623,
      "mean_token_accuracy": 0.13350174129009246,
      "num_tokens": 8128951.0,
      "step": 3695
    },
    {
      "entropy": 5.638001918792725,
      "epoch": 0.36720921000396983,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991875985201889,
      "loss": 5.6795,
      "mean_token_accuracy": 0.13438779413700103,
      "num_tokens": 8139385.0,
      "step": 3700
    },
    {
      "entropy": 5.794062376022339,
      "epoch": 0.36770543866613736,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499184587543998,
      "loss": 5.7161,
      "mean_token_accuracy": 0.13666265308856965,
      "num_tokens": 8149376.0,
      "step": 3705
    },
    {
      "entropy": 5.7886322975158695,
      "epoch": 0.3682016673283049,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004991815710085126,
      "loss": 5.7389,
      "mean_token_accuracy": 0.13325911313295363,
      "num_tokens": 8160687.0,
      "step": 3710
    },
    {
      "entropy": 5.797859144210816,
      "epoch": 0.3686978959904724,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991785489138072,
      "loss": 5.7453,
      "mean_token_accuracy": 0.12990956529974937,
      "num_tokens": 8171762.0,
      "step": 3715
    },
    {
      "entropy": 5.650119733810425,
      "epoch": 0.36919412465263995,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991755212599569,
      "loss": 5.5183,
      "mean_token_accuracy": 0.1405628427863121,
      "num_tokens": 8181643.0,
      "step": 3720
    },
    {
      "entropy": 5.618019437789917,
      "epoch": 0.3696903533148075,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991724880470366,
      "loss": 5.7692,
      "mean_token_accuracy": 0.13439096063375472,
      "num_tokens": 8192692.0,
      "step": 3725
    },
    {
      "entropy": 5.8731317043304445,
      "epoch": 0.370186581976975,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991694492751217,
      "loss": 5.7165,
      "mean_token_accuracy": 0.1306888848543167,
      "num_tokens": 8202941.0,
      "step": 3730
    },
    {
      "entropy": 5.741935539245605,
      "epoch": 0.3706828106391425,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991664049442876,
      "loss": 5.721,
      "mean_token_accuracy": 0.13041145205497742,
      "num_tokens": 8214206.0,
      "step": 3735
    },
    {
      "entropy": 5.719397354125976,
      "epoch": 0.37117903930131,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004991633550546097,
      "loss": 5.5478,
      "mean_token_accuracy": 0.14364924728870393,
      "num_tokens": 8223672.0,
      "step": 3740
    },
    {
      "entropy": 5.712414312362671,
      "epoch": 0.37167526796347755,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991602996061635,
      "loss": 5.6767,
      "mean_token_accuracy": 0.13784719556570052,
      "num_tokens": 8234189.0,
      "step": 3745
    },
    {
      "entropy": 5.789827442169189,
      "epoch": 0.3721714966256451,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991572385990249,
      "loss": 5.7125,
      "mean_token_accuracy": 0.13273593932390212,
      "num_tokens": 8245382.0,
      "step": 3750
    },
    {
      "entropy": 5.801346778869629,
      "epoch": 0.3726677252878126,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991541720332698,
      "loss": 5.6675,
      "mean_token_accuracy": 0.13517806306481361,
      "num_tokens": 8254530.0,
      "step": 3755
    },
    {
      "entropy": 5.70973916053772,
      "epoch": 0.37316395394998014,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991510999089743,
      "loss": 5.7233,
      "mean_token_accuracy": 0.13753556311130524,
      "num_tokens": 8264030.0,
      "step": 3760
    },
    {
      "entropy": 5.7114022731781,
      "epoch": 0.3736601826121477,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004991480222262145,
      "loss": 5.6906,
      "mean_token_accuracy": 0.13835713118314744,
      "num_tokens": 8273537.0,
      "step": 3765
    },
    {
      "entropy": 5.805447435379028,
      "epoch": 0.3741564112743152,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004991449389850668,
      "loss": 5.7144,
      "mean_token_accuracy": 0.131557959318161,
      "num_tokens": 8286123.0,
      "step": 3770
    },
    {
      "entropy": 5.782719993591309,
      "epoch": 0.37465263993648273,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991418501856076,
      "loss": 5.7359,
      "mean_token_accuracy": 0.1322738565504551,
      "num_tokens": 8298115.0,
      "step": 3775
    },
    {
      "entropy": 5.8190759181976315,
      "epoch": 0.37514886859865026,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991387558279134,
      "loss": 5.7701,
      "mean_token_accuracy": 0.13059676066040993,
      "num_tokens": 8309376.0,
      "step": 3780
    },
    {
      "entropy": 5.72836856842041,
      "epoch": 0.3756450972608178,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991356559120612,
      "loss": 5.7162,
      "mean_token_accuracy": 0.13034249395132064,
      "num_tokens": 8319272.0,
      "step": 3785
    },
    {
      "entropy": 5.763256311416626,
      "epoch": 0.3761413259229853,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991325504381275,
      "loss": 5.7034,
      "mean_token_accuracy": 0.1289884179830551,
      "num_tokens": 8330042.0,
      "step": 3790
    },
    {
      "entropy": 5.748146390914917,
      "epoch": 0.37663755458515286,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991294394061896,
      "loss": 5.6656,
      "mean_token_accuracy": 0.13286871761083602,
      "num_tokens": 8340394.0,
      "step": 3795
    },
    {
      "entropy": 5.679359817504883,
      "epoch": 0.3771337832473204,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991263228163246,
      "loss": 5.7017,
      "mean_token_accuracy": 0.1317191243171692,
      "num_tokens": 8351103.0,
      "step": 3800
    },
    {
      "entropy": 5.786353254318238,
      "epoch": 0.3776300119094879,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991232006686096,
      "loss": 5.7306,
      "mean_token_accuracy": 0.13431305363774299,
      "num_tokens": 8363215.0,
      "step": 3805
    },
    {
      "entropy": 5.844559383392334,
      "epoch": 0.3781262405716554,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991200729631223,
      "loss": 5.7776,
      "mean_token_accuracy": 0.12665583416819573,
      "num_tokens": 8374602.0,
      "step": 3810
    },
    {
      "entropy": 5.70382399559021,
      "epoch": 0.3786224692338229,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991169396999401,
      "loss": 5.7011,
      "mean_token_accuracy": 0.1290666401386261,
      "num_tokens": 8386470.0,
      "step": 3815
    },
    {
      "entropy": 5.795304489135742,
      "epoch": 0.37911869789599045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004991138008791408,
      "loss": 5.7216,
      "mean_token_accuracy": 0.1339471235871315,
      "num_tokens": 8396284.0,
      "step": 3820
    },
    {
      "entropy": 5.835215473175049,
      "epoch": 0.379614926558158,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991106565008021,
      "loss": 5.7438,
      "mean_token_accuracy": 0.12994968369603158,
      "num_tokens": 8407856.0,
      "step": 3825
    },
    {
      "entropy": 5.737123155593872,
      "epoch": 0.3801111552203255,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991075065650019,
      "loss": 5.6464,
      "mean_token_accuracy": 0.1383263237774372,
      "num_tokens": 8419560.0,
      "step": 3830
    },
    {
      "entropy": 5.79041862487793,
      "epoch": 0.38060738388249304,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991043510718186,
      "loss": 5.6539,
      "mean_token_accuracy": 0.13635289072990417,
      "num_tokens": 8430594.0,
      "step": 3835
    },
    {
      "entropy": 5.666716289520264,
      "epoch": 0.3811036125446606,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991011900213303,
      "loss": 5.6231,
      "mean_token_accuracy": 0.13809854835271834,
      "num_tokens": 8440549.0,
      "step": 3840
    },
    {
      "entropy": 5.663881063461304,
      "epoch": 0.3815998412068281,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990980234136152,
      "loss": 5.6503,
      "mean_token_accuracy": 0.1358008362352848,
      "num_tokens": 8451553.0,
      "step": 3845
    },
    {
      "entropy": 5.762479019165039,
      "epoch": 0.38209606986899564,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990948512487522,
      "loss": 5.6673,
      "mean_token_accuracy": 0.13858909234404565,
      "num_tokens": 8462899.0,
      "step": 3850
    },
    {
      "entropy": 5.748224115371704,
      "epoch": 0.38259229853116317,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990916735268198,
      "loss": 5.7106,
      "mean_token_accuracy": 0.13554939329624177,
      "num_tokens": 8474566.0,
      "step": 3855
    },
    {
      "entropy": 5.790412521362304,
      "epoch": 0.3830885271933307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990884902478966,
      "loss": 5.7276,
      "mean_token_accuracy": 0.13737209066748618,
      "num_tokens": 8484786.0,
      "step": 3860
    },
    {
      "entropy": 5.723254299163818,
      "epoch": 0.3835847558554982,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990853014120619,
      "loss": 5.6614,
      "mean_token_accuracy": 0.13724090456962584,
      "num_tokens": 8496416.0,
      "step": 3865
    },
    {
      "entropy": 5.744536972045898,
      "epoch": 0.38408098451766576,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990821070193944,
      "loss": 5.6919,
      "mean_token_accuracy": 0.13323881477117538,
      "num_tokens": 8507287.0,
      "step": 3870
    },
    {
      "entropy": 5.806073141098023,
      "epoch": 0.3845772131798333,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990789070699736,
      "loss": 5.7355,
      "mean_token_accuracy": 0.1301907204091549,
      "num_tokens": 8517922.0,
      "step": 3875
    },
    {
      "entropy": 5.694337511062622,
      "epoch": 0.3850734418420008,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990757015638787,
      "loss": 5.6752,
      "mean_token_accuracy": 0.1399615302681923,
      "num_tokens": 8529532.0,
      "step": 3880
    },
    {
      "entropy": 5.744428539276123,
      "epoch": 0.3855696705041683,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990724905011894,
      "loss": 5.6327,
      "mean_token_accuracy": 0.13914796561002732,
      "num_tokens": 8540207.0,
      "step": 3885
    },
    {
      "entropy": 5.743565320968628,
      "epoch": 0.3860658991663358,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499069273881985,
      "loss": 5.6818,
      "mean_token_accuracy": 0.13373377323150634,
      "num_tokens": 8551634.0,
      "step": 3890
    },
    {
      "entropy": 5.73831787109375,
      "epoch": 0.38656212782850335,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990660517063455,
      "loss": 5.7359,
      "mean_token_accuracy": 0.12965402603149415,
      "num_tokens": 8563256.0,
      "step": 3895
    },
    {
      "entropy": 5.75383768081665,
      "epoch": 0.3870583564906709,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990628239743508,
      "loss": 5.7733,
      "mean_token_accuracy": 0.13520312905311585,
      "num_tokens": 8573833.0,
      "step": 3900
    },
    {
      "entropy": 5.775561237335205,
      "epoch": 0.3875545851528384,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990595906860807,
      "loss": 5.6575,
      "mean_token_accuracy": 0.13614343851804733,
      "num_tokens": 8584283.0,
      "step": 3905
    },
    {
      "entropy": 5.725812149047852,
      "epoch": 0.38805081381500595,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990563518416157,
      "loss": 5.715,
      "mean_token_accuracy": 0.13555167317390443,
      "num_tokens": 8595344.0,
      "step": 3910
    },
    {
      "entropy": 5.760308265686035,
      "epoch": 0.3885470424771735,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990531074410358,
      "loss": 5.6478,
      "mean_token_accuracy": 0.13378319814801215,
      "num_tokens": 8605625.0,
      "step": 3915
    },
    {
      "entropy": 5.705371856689453,
      "epoch": 0.389043271139341,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990498574844218,
      "loss": 5.6268,
      "mean_token_accuracy": 0.13527522459626198,
      "num_tokens": 8616259.0,
      "step": 3920
    },
    {
      "entropy": 5.714511823654175,
      "epoch": 0.38953949980150854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990466019718542,
      "loss": 5.6418,
      "mean_token_accuracy": 0.13701364398002625,
      "num_tokens": 8627179.0,
      "step": 3925
    },
    {
      "entropy": 5.776921558380127,
      "epoch": 0.39003572846367607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990433409034134,
      "loss": 5.7392,
      "mean_token_accuracy": 0.13890910968184472,
      "num_tokens": 8637448.0,
      "step": 3930
    },
    {
      "entropy": 5.656614494323731,
      "epoch": 0.3905319571258436,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990400742791806,
      "loss": 5.6588,
      "mean_token_accuracy": 0.13583290949463844,
      "num_tokens": 8649168.0,
      "step": 3935
    },
    {
      "entropy": 5.842466402053833,
      "epoch": 0.39102818578801113,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004990368020992368,
      "loss": 5.77,
      "mean_token_accuracy": 0.13096642941236497,
      "num_tokens": 8660653.0,
      "step": 3940
    },
    {
      "entropy": 5.779029989242554,
      "epoch": 0.39152441445017866,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499033524363663,
      "loss": 5.7419,
      "mean_token_accuracy": 0.13649297654628753,
      "num_tokens": 8672000.0,
      "step": 3945
    },
    {
      "entropy": 5.689775228500366,
      "epoch": 0.3920206431123462,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990302410725405,
      "loss": 5.6659,
      "mean_token_accuracy": 0.14118805006146432,
      "num_tokens": 8682455.0,
      "step": 3950
    },
    {
      "entropy": 5.721955442428589,
      "epoch": 0.3925168717745137,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990269522259509,
      "loss": 5.6392,
      "mean_token_accuracy": 0.13590310513973236,
      "num_tokens": 8692975.0,
      "step": 3955
    },
    {
      "entropy": 5.674862480163574,
      "epoch": 0.3930131004366812,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990236578239755,
      "loss": 5.6253,
      "mean_token_accuracy": 0.1390941359102726,
      "num_tokens": 8703569.0,
      "step": 3960
    },
    {
      "entropy": 5.759145498275757,
      "epoch": 0.3935093290988487,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990203578666962,
      "loss": 5.7267,
      "mean_token_accuracy": 0.13206208124756813,
      "num_tokens": 8713734.0,
      "step": 3965
    },
    {
      "entropy": 5.743315029144287,
      "epoch": 0.39400555776101626,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990170523541948,
      "loss": 5.6232,
      "mean_token_accuracy": 0.13998670354485512,
      "num_tokens": 8723852.0,
      "step": 3970
    },
    {
      "entropy": 5.716074228286743,
      "epoch": 0.3945017864231838,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990137412865531,
      "loss": 5.7699,
      "mean_token_accuracy": 0.13368171229958534,
      "num_tokens": 8734300.0,
      "step": 3975
    },
    {
      "entropy": 5.844010591506958,
      "epoch": 0.3949980150853513,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990104246638533,
      "loss": 5.7851,
      "mean_token_accuracy": 0.12670395150780678,
      "num_tokens": 8745022.0,
      "step": 3980
    },
    {
      "entropy": 5.770476388931274,
      "epoch": 0.39549424374751885,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004990071024861778,
      "loss": 5.6707,
      "mean_token_accuracy": 0.13448970913887023,
      "num_tokens": 8754870.0,
      "step": 3985
    },
    {
      "entropy": 5.8409770965576175,
      "epoch": 0.3959904724096864,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990037747536087,
      "loss": 5.7111,
      "mean_token_accuracy": 0.13520667031407357,
      "num_tokens": 8765536.0,
      "step": 3990
    },
    {
      "entropy": 5.734442949295044,
      "epoch": 0.3964867010718539,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990004414662288,
      "loss": 5.6826,
      "mean_token_accuracy": 0.13069720268249513,
      "num_tokens": 8776898.0,
      "step": 3995
    },
    {
      "entropy": 5.730175352096557,
      "epoch": 0.39698292973402144,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989971026241205,
      "loss": 5.7419,
      "mean_token_accuracy": 0.13170261383056642,
      "num_tokens": 8788176.0,
      "step": 4000
    },
    {
      "entropy": 5.692081499099731,
      "epoch": 0.39747915839618897,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989937582273668,
      "loss": 5.6738,
      "mean_token_accuracy": 0.1344978965818882,
      "num_tokens": 8798939.0,
      "step": 4005
    },
    {
      "entropy": 5.766072082519531,
      "epoch": 0.3979753870583565,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989904082760506,
      "loss": 5.7043,
      "mean_token_accuracy": 0.13863254114985465,
      "num_tokens": 8809092.0,
      "step": 4010
    },
    {
      "entropy": 5.7443944931030275,
      "epoch": 0.39847161572052403,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498987052770255,
      "loss": 5.6311,
      "mean_token_accuracy": 0.14151344299316407,
      "num_tokens": 8819352.0,
      "step": 4015
    },
    {
      "entropy": 5.7538347244262695,
      "epoch": 0.39896784438269156,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498983691710063,
      "loss": 5.7579,
      "mean_token_accuracy": 0.13163795024156572,
      "num_tokens": 8829878.0,
      "step": 4020
    },
    {
      "entropy": 5.75619592666626,
      "epoch": 0.3994640730448591,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498980325095558,
      "loss": 5.6265,
      "mean_token_accuracy": 0.1341935031116009,
      "num_tokens": 8840364.0,
      "step": 4025
    },
    {
      "entropy": 5.720586156845092,
      "epoch": 0.3999603017070266,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989769529268238,
      "loss": 5.7439,
      "mean_token_accuracy": 0.12914118841290473,
      "num_tokens": 8851340.0,
      "step": 4030
    },
    {
      "entropy": 5.715707445144654,
      "epoch": 0.4004565303691941,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989735752039436,
      "loss": 5.6986,
      "mean_token_accuracy": 0.13207997828722,
      "num_tokens": 8862571.0,
      "step": 4035
    },
    {
      "entropy": 5.753764533996582,
      "epoch": 0.4009527590313616,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989701919270015,
      "loss": 5.7026,
      "mean_token_accuracy": 0.13718848526477814,
      "num_tokens": 8873927.0,
      "step": 4040
    },
    {
      "entropy": 5.802546691894531,
      "epoch": 0.40144898769352916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989668030960812,
      "loss": 5.8891,
      "mean_token_accuracy": 0.12460704743862153,
      "num_tokens": 8886542.0,
      "step": 4045
    },
    {
      "entropy": 5.795607662200927,
      "epoch": 0.4019452163556967,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989634087112667,
      "loss": 5.6369,
      "mean_token_accuracy": 0.13462657555937768,
      "num_tokens": 8896114.0,
      "step": 4050
    },
    {
      "entropy": 5.739918994903564,
      "epoch": 0.4024414450178642,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989600087726425,
      "loss": 5.7415,
      "mean_token_accuracy": 0.12707119584083557,
      "num_tokens": 8906303.0,
      "step": 4055
    },
    {
      "entropy": 5.770398187637329,
      "epoch": 0.40293767368003175,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004989566032802924,
      "loss": 5.7568,
      "mean_token_accuracy": 0.12863401621580123,
      "num_tokens": 8916961.0,
      "step": 4060
    },
    {
      "entropy": 5.899053716659546,
      "epoch": 0.4034339023421993,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989531922343013,
      "loss": 5.8522,
      "mean_token_accuracy": 0.1265175573527813,
      "num_tokens": 8927700.0,
      "step": 4065
    },
    {
      "entropy": 5.775335073471069,
      "epoch": 0.4039301310043668,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989497756347534,
      "loss": 5.6768,
      "mean_token_accuracy": 0.1359332799911499,
      "num_tokens": 8939655.0,
      "step": 4070
    },
    {
      "entropy": 5.701972103118896,
      "epoch": 0.40442635966653434,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989463534817339,
      "loss": 5.6153,
      "mean_token_accuracy": 0.13829075172543526,
      "num_tokens": 8951238.0,
      "step": 4075
    },
    {
      "entropy": 5.765563440322876,
      "epoch": 0.40492258832870187,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989429257753273,
      "loss": 5.69,
      "mean_token_accuracy": 0.14064479097723961,
      "num_tokens": 8962019.0,
      "step": 4080
    },
    {
      "entropy": 5.6876383304595945,
      "epoch": 0.4054188169908694,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989394925156187,
      "loss": 5.6758,
      "mean_token_accuracy": 0.13742703646421434,
      "num_tokens": 8973861.0,
      "step": 4085
    },
    {
      "entropy": 5.685515975952148,
      "epoch": 0.40591504565303693,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989360537026932,
      "loss": 5.6754,
      "mean_token_accuracy": 0.13646822348237037,
      "num_tokens": 8985981.0,
      "step": 4090
    },
    {
      "entropy": 5.736924839019776,
      "epoch": 0.40641127431520446,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498932609336636,
      "loss": 5.5817,
      "mean_token_accuracy": 0.14304670542478562,
      "num_tokens": 8995318.0,
      "step": 4095
    },
    {
      "entropy": 5.731611108779907,
      "epoch": 0.406907502977372,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989291594175327,
      "loss": 5.6133,
      "mean_token_accuracy": 0.13422033861279486,
      "num_tokens": 9004679.0,
      "step": 4100
    },
    {
      "entropy": 5.830239486694336,
      "epoch": 0.4074037316395395,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989257039454689,
      "loss": 5.738,
      "mean_token_accuracy": 0.133600851893425,
      "num_tokens": 9016422.0,
      "step": 4105
    },
    {
      "entropy": 5.641452980041504,
      "epoch": 0.40789996030170705,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049892224292053,
      "loss": 5.6599,
      "mean_token_accuracy": 0.13614288121461868,
      "num_tokens": 9027914.0,
      "step": 4110
    },
    {
      "entropy": 5.735976791381836,
      "epoch": 0.40839618896387453,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989187763428021,
      "loss": 5.6392,
      "mean_token_accuracy": 0.1383671998977661,
      "num_tokens": 9038701.0,
      "step": 4115
    },
    {
      "entropy": 5.70672926902771,
      "epoch": 0.40889241762604206,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498915304212371,
      "loss": 5.6956,
      "mean_token_accuracy": 0.13438485115766524,
      "num_tokens": 9050620.0,
      "step": 4120
    },
    {
      "entropy": 5.784137392044068,
      "epoch": 0.4093886462882096,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989118265293228,
      "loss": 5.7157,
      "mean_token_accuracy": 0.13783123940229416,
      "num_tokens": 9060900.0,
      "step": 4125
    },
    {
      "entropy": 5.692770051956177,
      "epoch": 0.4098848749503771,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989083432937439,
      "loss": 5.6358,
      "mean_token_accuracy": 0.13960246965289116,
      "num_tokens": 9072122.0,
      "step": 4130
    },
    {
      "entropy": 5.695958614349365,
      "epoch": 0.41038110361254465,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004989048545057204,
      "loss": 5.6621,
      "mean_token_accuracy": 0.13656162321567536,
      "num_tokens": 9082780.0,
      "step": 4135
    },
    {
      "entropy": 5.6983644485473635,
      "epoch": 0.4108773322747122,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989013601653391,
      "loss": 5.6915,
      "mean_token_accuracy": 0.13537346571683884,
      "num_tokens": 9093518.0,
      "step": 4140
    },
    {
      "entropy": 5.7246016502380375,
      "epoch": 0.4113735609368797,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988978602726866,
      "loss": 5.716,
      "mean_token_accuracy": 0.14137836471199988,
      "num_tokens": 9105035.0,
      "step": 4145
    },
    {
      "entropy": 5.734498691558838,
      "epoch": 0.41186978959904724,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988943548278494,
      "loss": 5.6336,
      "mean_token_accuracy": 0.13698274716734887,
      "num_tokens": 9117154.0,
      "step": 4150
    },
    {
      "entropy": 5.706672430038452,
      "epoch": 0.4123660182612148,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988908438309149,
      "loss": 5.7051,
      "mean_token_accuracy": 0.1334255114197731,
      "num_tokens": 9128818.0,
      "step": 4155
    },
    {
      "entropy": 5.768472909927368,
      "epoch": 0.4128622469233823,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988873272819697,
      "loss": 5.7179,
      "mean_token_accuracy": 0.1378393739461899,
      "num_tokens": 9140188.0,
      "step": 4160
    },
    {
      "entropy": 5.915740299224853,
      "epoch": 0.41335847558554983,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988838051811014,
      "loss": 5.8077,
      "mean_token_accuracy": 0.12831991240382196,
      "num_tokens": 9150565.0,
      "step": 4165
    },
    {
      "entropy": 5.691146659851074,
      "epoch": 0.41385470424771736,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498880277528397,
      "loss": 5.659,
      "mean_token_accuracy": 0.13782408088445663,
      "num_tokens": 9160130.0,
      "step": 4170
    },
    {
      "entropy": 5.652090454101563,
      "epoch": 0.4143509329098849,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988767443239443,
      "loss": 5.652,
      "mean_token_accuracy": 0.14071165323257445,
      "num_tokens": 9171740.0,
      "step": 4175
    },
    {
      "entropy": 5.747271871566772,
      "epoch": 0.4148471615720524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988732055678307,
      "loss": 5.609,
      "mean_token_accuracy": 0.15193828493356704,
      "num_tokens": 9182459.0,
      "step": 4180
    },
    {
      "entropy": 5.772048234939575,
      "epoch": 0.41534339023421996,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498869661260144,
      "loss": 5.6562,
      "mean_token_accuracy": 0.13707233741879463,
      "num_tokens": 9192284.0,
      "step": 4185
    },
    {
      "entropy": 5.678713798522949,
      "epoch": 0.41583961889638743,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498866111400972,
      "loss": 5.6578,
      "mean_token_accuracy": 0.13831765949726105,
      "num_tokens": 9202700.0,
      "step": 4190
    },
    {
      "entropy": 5.702182912826538,
      "epoch": 0.41633584755855496,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988625559904029,
      "loss": 5.6639,
      "mean_token_accuracy": 0.1302245832979679,
      "num_tokens": 9213013.0,
      "step": 4195
    },
    {
      "entropy": 5.792655801773071,
      "epoch": 0.4168320762207225,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004988589950285248,
      "loss": 5.7366,
      "mean_token_accuracy": 0.13581578209996223,
      "num_tokens": 9224568.0,
      "step": 4200
    },
    {
      "entropy": 5.735432529449463,
      "epoch": 0.41732830488289,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988554285154259,
      "loss": 5.6132,
      "mean_token_accuracy": 0.13576259464025497,
      "num_tokens": 9233791.0,
      "step": 4205
    },
    {
      "entropy": 5.714122867584228,
      "epoch": 0.41782453354505755,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988518564511948,
      "loss": 5.6566,
      "mean_token_accuracy": 0.13220425993204116,
      "num_tokens": 9245048.0,
      "step": 4210
    },
    {
      "entropy": 5.798601531982422,
      "epoch": 0.4183207622072251,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004988482788359202,
      "loss": 5.6919,
      "mean_token_accuracy": 0.14111457020044327,
      "num_tokens": 9256508.0,
      "step": 4215
    },
    {
      "entropy": 5.750736999511719,
      "epoch": 0.4188169908693926,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988446956696904,
      "loss": 5.7043,
      "mean_token_accuracy": 0.13517057597637178,
      "num_tokens": 9267871.0,
      "step": 4220
    },
    {
      "entropy": 5.788787364959717,
      "epoch": 0.41931321953156014,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004988411069525947,
      "loss": 5.7662,
      "mean_token_accuracy": 0.13160478845238685,
      "num_tokens": 9279287.0,
      "step": 4225
    },
    {
      "entropy": 5.814012050628662,
      "epoch": 0.4198094481937277,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988375126847216,
      "loss": 5.7846,
      "mean_token_accuracy": 0.12892357259988785,
      "num_tokens": 9291257.0,
      "step": 4230
    },
    {
      "entropy": 5.751958894729614,
      "epoch": 0.4203056768558952,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004988339128661607,
      "loss": 5.7387,
      "mean_token_accuracy": 0.13436317667365075,
      "num_tokens": 9302590.0,
      "step": 4235
    },
    {
      "entropy": 5.624151992797851,
      "epoch": 0.42080190551806274,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498830307497001,
      "loss": 5.5986,
      "mean_token_accuracy": 0.13606150150299073,
      "num_tokens": 9313174.0,
      "step": 4240
    },
    {
      "entropy": 5.751293659210205,
      "epoch": 0.42129813418023027,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498826696577332,
      "loss": 5.6813,
      "mean_token_accuracy": 0.13548489436507224,
      "num_tokens": 9325034.0,
      "step": 4245
    },
    {
      "entropy": 5.739410495758056,
      "epoch": 0.4217943628423978,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988230801072432,
      "loss": 5.6622,
      "mean_token_accuracy": 0.13756145015358925,
      "num_tokens": 9336834.0,
      "step": 4250
    },
    {
      "entropy": 5.642874622344971,
      "epoch": 0.4222905915045653,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988194580868244,
      "loss": 5.6777,
      "mean_token_accuracy": 0.13649028465151786,
      "num_tokens": 9347747.0,
      "step": 4255
    },
    {
      "entropy": 5.70621657371521,
      "epoch": 0.42278682016673286,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988158305161654,
      "loss": 5.69,
      "mean_token_accuracy": 0.13723429888486863,
      "num_tokens": 9359186.0,
      "step": 4260
    },
    {
      "entropy": 5.729804754257202,
      "epoch": 0.42328304882890033,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498812197395356,
      "loss": 5.6134,
      "mean_token_accuracy": 0.1391861580312252,
      "num_tokens": 9369485.0,
      "step": 4265
    },
    {
      "entropy": 5.728609323501587,
      "epoch": 0.42377927749106786,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004988085587244863,
      "loss": 5.6961,
      "mean_token_accuracy": 0.13410992473363875,
      "num_tokens": 9382658.0,
      "step": 4270
    },
    {
      "entropy": 5.681890535354614,
      "epoch": 0.4242755061532354,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004988049145036466,
      "loss": 5.674,
      "mean_token_accuracy": 0.13061357513070107,
      "num_tokens": 9394151.0,
      "step": 4275
    },
    {
      "entropy": 5.812398242950439,
      "epoch": 0.4247717348154029,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988012647329273,
      "loss": 5.7482,
      "mean_token_accuracy": 0.13043404445052148,
      "num_tokens": 9405956.0,
      "step": 4280
    },
    {
      "entropy": 5.757098770141601,
      "epoch": 0.42526796347757045,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498797609412419,
      "loss": 5.6348,
      "mean_token_accuracy": 0.1390029162168503,
      "num_tokens": 9416889.0,
      "step": 4285
    },
    {
      "entropy": 5.752142858505249,
      "epoch": 0.425764192139738,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987939485422122,
      "loss": 5.7588,
      "mean_token_accuracy": 0.12929404750466347,
      "num_tokens": 9427602.0,
      "step": 4290
    },
    {
      "entropy": 5.780097198486328,
      "epoch": 0.4262604208019055,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004987902821223976,
      "loss": 5.6649,
      "mean_token_accuracy": 0.13806519359350206,
      "num_tokens": 9438557.0,
      "step": 4295
    },
    {
      "entropy": 5.725652647018433,
      "epoch": 0.42675664946407305,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987866101530663,
      "loss": 5.7168,
      "mean_token_accuracy": 0.13426719680428506,
      "num_tokens": 9449598.0,
      "step": 4300
    },
    {
      "entropy": 5.707887935638428,
      "epoch": 0.4272528781262406,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987829326343092,
      "loss": 5.6417,
      "mean_token_accuracy": 0.1364934429526329,
      "num_tokens": 9459740.0,
      "step": 4305
    },
    {
      "entropy": 5.700517320632935,
      "epoch": 0.4277491067884081,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987792495662177,
      "loss": 5.659,
      "mean_token_accuracy": 0.13137739896774292,
      "num_tokens": 9469835.0,
      "step": 4310
    },
    {
      "entropy": 5.794715166091919,
      "epoch": 0.42824533545057564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498775560948883,
      "loss": 5.741,
      "mean_token_accuracy": 0.13609717413783073,
      "num_tokens": 9480735.0,
      "step": 4315
    },
    {
      "entropy": 5.718795871734619,
      "epoch": 0.42874156411274317,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987718667823965,
      "loss": 5.6311,
      "mean_token_accuracy": 0.14181154519319533,
      "num_tokens": 9491321.0,
      "step": 4320
    },
    {
      "entropy": 5.699181270599365,
      "epoch": 0.4292377927749107,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00049876816706685,
      "loss": 5.7662,
      "mean_token_accuracy": 0.12978544905781747,
      "num_tokens": 9503507.0,
      "step": 4325
    },
    {
      "entropy": 5.800131225585938,
      "epoch": 0.42973402143707823,
      "grad_norm": 1.8984375,
      "learning_rate": 0.000498764461802335,
      "loss": 5.7704,
      "mean_token_accuracy": 0.13711531162261964,
      "num_tokens": 9514404.0,
      "step": 4330
    },
    {
      "entropy": 5.757583141326904,
      "epoch": 0.43023025009924576,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004987607509889437,
      "loss": 5.631,
      "mean_token_accuracy": 0.1362014576792717,
      "num_tokens": 9525555.0,
      "step": 4335
    },
    {
      "entropy": 5.753689956665039,
      "epoch": 0.43072647876141323,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987570346267677,
      "loss": 5.6269,
      "mean_token_accuracy": 0.1425912782549858,
      "num_tokens": 9535084.0,
      "step": 4340
    },
    {
      "entropy": 5.658543395996094,
      "epoch": 0.43122270742358076,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987533127158996,
      "loss": 5.6504,
      "mean_token_accuracy": 0.1396719478070736,
      "num_tokens": 9546170.0,
      "step": 4345
    },
    {
      "entropy": 5.718197584152222,
      "epoch": 0.4317189360857483,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987495852564314,
      "loss": 5.6516,
      "mean_token_accuracy": 0.1370712161064148,
      "num_tokens": 9556793.0,
      "step": 4350
    },
    {
      "entropy": 5.743592357635498,
      "epoch": 0.4322151647479158,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987458522484556,
      "loss": 5.6103,
      "mean_token_accuracy": 0.1390673503279686,
      "num_tokens": 9568114.0,
      "step": 4355
    },
    {
      "entropy": 5.67595329284668,
      "epoch": 0.43271139341008336,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004987421136920648,
      "loss": 5.607,
      "mean_token_accuracy": 0.13263363316655158,
      "num_tokens": 9579119.0,
      "step": 4360
    },
    {
      "entropy": 5.631229734420776,
      "epoch": 0.4332076220722509,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004987383695873517,
      "loss": 5.6654,
      "mean_token_accuracy": 0.13817091807723045,
      "num_tokens": 9590259.0,
      "step": 4365
    },
    {
      "entropy": 5.799467992782593,
      "epoch": 0.4337038507344184,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987346199344092,
      "loss": 5.7005,
      "mean_token_accuracy": 0.12761994451284409,
      "num_tokens": 9600431.0,
      "step": 4370
    },
    {
      "entropy": 5.686568689346314,
      "epoch": 0.43420007939658595,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987308647333303,
      "loss": 5.675,
      "mean_token_accuracy": 0.1352570503950119,
      "num_tokens": 9610286.0,
      "step": 4375
    },
    {
      "entropy": 5.780293607711792,
      "epoch": 0.4346963080587535,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004987271039842079,
      "loss": 5.691,
      "mean_token_accuracy": 0.13502518460154533,
      "num_tokens": 9621148.0,
      "step": 4380
    },
    {
      "entropy": 5.705330896377563,
      "epoch": 0.435192536720921,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004987233376871355,
      "loss": 5.577,
      "mean_token_accuracy": 0.13657931089401246,
      "num_tokens": 9631300.0,
      "step": 4385
    },
    {
      "entropy": 5.670144701004029,
      "epoch": 0.43568876538308854,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987195658422063,
      "loss": 5.6862,
      "mean_token_accuracy": 0.13548445776104928,
      "num_tokens": 9641986.0,
      "step": 4390
    },
    {
      "entropy": 5.737906885147095,
      "epoch": 0.43618499404525607,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987157884495141,
      "loss": 5.6967,
      "mean_token_accuracy": 0.1371404081583023,
      "num_tokens": 9652962.0,
      "step": 4395
    },
    {
      "entropy": 5.726242446899414,
      "epoch": 0.4366812227074236,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987120055091523,
      "loss": 5.7136,
      "mean_token_accuracy": 0.13460650369524957,
      "num_tokens": 9664084.0,
      "step": 4400
    },
    {
      "entropy": 5.708842039108276,
      "epoch": 0.43717745136959113,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987082170212149,
      "loss": 5.6156,
      "mean_token_accuracy": 0.13606675416231157,
      "num_tokens": 9675821.0,
      "step": 4405
    },
    {
      "entropy": 5.75541615486145,
      "epoch": 0.43767368003175866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987044229857957,
      "loss": 5.7358,
      "mean_token_accuracy": 0.13673339784145355,
      "num_tokens": 9687964.0,
      "step": 4410
    },
    {
      "entropy": 5.705683469772339,
      "epoch": 0.43816990869392614,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987006234029889,
      "loss": 5.7125,
      "mean_token_accuracy": 0.1395598903298378,
      "num_tokens": 9699653.0,
      "step": 4415
    },
    {
      "entropy": 5.797198009490967,
      "epoch": 0.43866613735609367,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986968182728885,
      "loss": 5.7274,
      "mean_token_accuracy": 0.13438003659248351,
      "num_tokens": 9709872.0,
      "step": 4420
    },
    {
      "entropy": 5.767149305343628,
      "epoch": 0.4391623660182612,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004986930075955892,
      "loss": 5.6454,
      "mean_token_accuracy": 0.13367924019694327,
      "num_tokens": 9720589.0,
      "step": 4425
    },
    {
      "entropy": 5.612972831726074,
      "epoch": 0.4396585946804287,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004986891913711853,
      "loss": 5.5894,
      "mean_token_accuracy": 0.14131948202848435,
      "num_tokens": 9731862.0,
      "step": 4430
    },
    {
      "entropy": 5.685967874526978,
      "epoch": 0.44015482334259626,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986853695997715,
      "loss": 5.7326,
      "mean_token_accuracy": 0.1408631294965744,
      "num_tokens": 9743710.0,
      "step": 4435
    },
    {
      "entropy": 5.753010702133179,
      "epoch": 0.4406510520047638,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004986815422814425,
      "loss": 5.6725,
      "mean_token_accuracy": 0.12475050538778305,
      "num_tokens": 9753855.0,
      "step": 4440
    },
    {
      "entropy": 5.8120142936706545,
      "epoch": 0.4411472806669313,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986777094162932,
      "loss": 5.6878,
      "mean_token_accuracy": 0.14027968794107437,
      "num_tokens": 9763883.0,
      "step": 4445
    },
    {
      "entropy": 5.672481060028076,
      "epoch": 0.44164350932909885,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986738710044187,
      "loss": 5.6472,
      "mean_token_accuracy": 0.13763445019721984,
      "num_tokens": 9774890.0,
      "step": 4450
    },
    {
      "entropy": 5.662131595611572,
      "epoch": 0.4421397379912664,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986700270459142,
      "loss": 5.635,
      "mean_token_accuracy": 0.13798378705978392,
      "num_tokens": 9785249.0,
      "step": 4455
    },
    {
      "entropy": 5.721683168411255,
      "epoch": 0.4426359666534339,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986661775408749,
      "loss": 5.6417,
      "mean_token_accuracy": 0.13845691978931426,
      "num_tokens": 9795723.0,
      "step": 4460
    },
    {
      "entropy": 5.741473865509033,
      "epoch": 0.44313219531560144,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004986623224893965,
      "loss": 5.5845,
      "mean_token_accuracy": 0.1356796771287918,
      "num_tokens": 9807637.0,
      "step": 4465
    },
    {
      "entropy": 5.690484952926636,
      "epoch": 0.44362842397776897,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986584618915744,
      "loss": 5.6982,
      "mean_token_accuracy": 0.13647759109735488,
      "num_tokens": 9820298.0,
      "step": 4470
    },
    {
      "entropy": 5.721742868423462,
      "epoch": 0.4441246526399365,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986545957475045,
      "loss": 5.5906,
      "mean_token_accuracy": 0.14316966831684114,
      "num_tokens": 9831251.0,
      "step": 4475
    },
    {
      "entropy": 5.652020597457886,
      "epoch": 0.44462088130210403,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004986507240572825,
      "loss": 5.6525,
      "mean_token_accuracy": 0.13798210993409157,
      "num_tokens": 9842884.0,
      "step": 4480
    },
    {
      "entropy": 5.637630605697632,
      "epoch": 0.44511710996427156,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004986468468210042,
      "loss": 5.5504,
      "mean_token_accuracy": 0.13658547177910804,
      "num_tokens": 9854413.0,
      "step": 4485
    },
    {
      "entropy": 5.747673511505127,
      "epoch": 0.44561333862643904,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986429640387663,
      "loss": 5.6892,
      "mean_token_accuracy": 0.13593099191784858,
      "num_tokens": 9865771.0,
      "step": 4490
    },
    {
      "entropy": 5.705465602874756,
      "epoch": 0.44610956728860657,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004986390757106648,
      "loss": 5.6164,
      "mean_token_accuracy": 0.14181874468922614,
      "num_tokens": 9876717.0,
      "step": 4495
    },
    {
      "entropy": 5.68349666595459,
      "epoch": 0.4466057959507741,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986351818367959,
      "loss": 5.69,
      "mean_token_accuracy": 0.1398135282099247,
      "num_tokens": 9887616.0,
      "step": 4500
    },
    {
      "entropy": 5.7810155868530275,
      "epoch": 0.44710202461294163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986312824172566,
      "loss": 5.661,
      "mean_token_accuracy": 0.13780627772212029,
      "num_tokens": 9898979.0,
      "step": 4505
    },
    {
      "entropy": 5.6977684020996096,
      "epoch": 0.44759825327510916,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004986273774521433,
      "loss": 5.6925,
      "mean_token_accuracy": 0.13116807863116264,
      "num_tokens": 9912158.0,
      "step": 4510
    },
    {
      "entropy": 5.686586380004883,
      "epoch": 0.4480944819372767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986234669415529,
      "loss": 5.6842,
      "mean_token_accuracy": 0.13569653257727624,
      "num_tokens": 9923886.0,
      "step": 4515
    },
    {
      "entropy": 5.7598377704620365,
      "epoch": 0.4485907105994442,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986195508855824,
      "loss": 5.6603,
      "mean_token_accuracy": 0.13824379518628122,
      "num_tokens": 9934979.0,
      "step": 4520
    },
    {
      "entropy": 5.7572063446044925,
      "epoch": 0.44908693926161175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498615629284329,
      "loss": 5.7586,
      "mean_token_accuracy": 0.128438288718462,
      "num_tokens": 9944715.0,
      "step": 4525
    },
    {
      "entropy": 5.716220664978027,
      "epoch": 0.4495831679237793,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986117021378896,
      "loss": 5.6142,
      "mean_token_accuracy": 0.13333934247493745,
      "num_tokens": 9954736.0,
      "step": 4530
    },
    {
      "entropy": 5.768646383285523,
      "epoch": 0.4500793965859468,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498607769446362,
      "loss": 5.7149,
      "mean_token_accuracy": 0.13724260032176971,
      "num_tokens": 9965952.0,
      "step": 4535
    },
    {
      "entropy": 5.718695592880249,
      "epoch": 0.45057562524811434,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986038312098435,
      "loss": 5.6581,
      "mean_token_accuracy": 0.14347179606556892,
      "num_tokens": 9976428.0,
      "step": 4540
    },
    {
      "entropy": 5.752839088439941,
      "epoch": 0.4510718539102819,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004985998874284319,
      "loss": 5.5975,
      "mean_token_accuracy": 0.1452907532453537,
      "num_tokens": 9986100.0,
      "step": 4545
    },
    {
      "entropy": 5.5815338611602785,
      "epoch": 0.4515680825724494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985959381022248,
      "loss": 5.5761,
      "mean_token_accuracy": 0.1382984571158886,
      "num_tokens": 9996584.0,
      "step": 4550
    },
    {
      "entropy": 5.693500804901123,
      "epoch": 0.45206431123461693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004985919832313203,
      "loss": 5.6048,
      "mean_token_accuracy": 0.13820491433143617,
      "num_tokens": 10007792.0,
      "step": 4555
    },
    {
      "entropy": 5.809909439086914,
      "epoch": 0.45256053989678446,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985880228158164,
      "loss": 5.7928,
      "mean_token_accuracy": 0.12971990257501603,
      "num_tokens": 10019615.0,
      "step": 4560
    },
    {
      "entropy": 5.679464054107666,
      "epoch": 0.45305676855895194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985840568558112,
      "loss": 5.635,
      "mean_token_accuracy": 0.1400697186589241,
      "num_tokens": 10030223.0,
      "step": 4565
    },
    {
      "entropy": 5.715747213363647,
      "epoch": 0.45355299722111947,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985800853514033,
      "loss": 5.6386,
      "mean_token_accuracy": 0.13305401057004929,
      "num_tokens": 10041522.0,
      "step": 4570
    },
    {
      "entropy": 5.730625104904175,
      "epoch": 0.454049225883287,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985761083026911,
      "loss": 5.5584,
      "mean_token_accuracy": 0.1452631801366806,
      "num_tokens": 10052847.0,
      "step": 4575
    },
    {
      "entropy": 5.649207019805909,
      "epoch": 0.45454545454545453,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985721257097731,
      "loss": 5.5915,
      "mean_token_accuracy": 0.13804994523525238,
      "num_tokens": 10062259.0,
      "step": 4580
    },
    {
      "entropy": 5.785450220108032,
      "epoch": 0.45504168320762206,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985681375727482,
      "loss": 5.7556,
      "mean_token_accuracy": 0.13155426308512688,
      "num_tokens": 10073341.0,
      "step": 4585
    },
    {
      "entropy": 5.724898719787598,
      "epoch": 0.4555379118697896,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985641438917151,
      "loss": 5.6249,
      "mean_token_accuracy": 0.14122139066457748,
      "num_tokens": 10083992.0,
      "step": 4590
    },
    {
      "entropy": 5.7234375,
      "epoch": 0.4560341405319571,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985601446667731,
      "loss": 5.6081,
      "mean_token_accuracy": 0.13897195234894752,
      "num_tokens": 10094245.0,
      "step": 4595
    },
    {
      "entropy": 5.732874202728271,
      "epoch": 0.45653036919412465,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985561398980213,
      "loss": 5.5905,
      "mean_token_accuracy": 0.14286764189600945,
      "num_tokens": 10105976.0,
      "step": 4600
    },
    {
      "entropy": 5.620663595199585,
      "epoch": 0.4570265978562922,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985521295855588,
      "loss": 5.6699,
      "mean_token_accuracy": 0.13586745038628578,
      "num_tokens": 10117373.0,
      "step": 4605
    },
    {
      "entropy": 5.703548860549927,
      "epoch": 0.4575228265184597,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985481137294853,
      "loss": 5.612,
      "mean_token_accuracy": 0.13380024060606957,
      "num_tokens": 10128470.0,
      "step": 4610
    },
    {
      "entropy": 5.61048436164856,
      "epoch": 0.45801905518062724,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985440923299002,
      "loss": 5.5932,
      "mean_token_accuracy": 0.14190664738416672,
      "num_tokens": 10140217.0,
      "step": 4615
    },
    {
      "entropy": 5.656254386901855,
      "epoch": 0.4585152838427948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985400653869034,
      "loss": 5.6962,
      "mean_token_accuracy": 0.13559751883149146,
      "num_tokens": 10152308.0,
      "step": 4620
    },
    {
      "entropy": 5.842973804473877,
      "epoch": 0.4590115125049623,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985360329005945,
      "loss": 5.7106,
      "mean_token_accuracy": 0.13268141373991965,
      "num_tokens": 10164234.0,
      "step": 4625
    },
    {
      "entropy": 5.825473308563232,
      "epoch": 0.45950774116712984,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985319948710738,
      "loss": 5.7591,
      "mean_token_accuracy": 0.13398344442248344,
      "num_tokens": 10175287.0,
      "step": 4630
    },
    {
      "entropy": 5.714378023147583,
      "epoch": 0.46000396982929737,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985279512984413,
      "loss": 5.648,
      "mean_token_accuracy": 0.14291325584053993,
      "num_tokens": 10184658.0,
      "step": 4635
    },
    {
      "entropy": 5.6979296684265135,
      "epoch": 0.46050019849146484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985239021827972,
      "loss": 5.6358,
      "mean_token_accuracy": 0.13907306492328644,
      "num_tokens": 10194593.0,
      "step": 4640
    },
    {
      "entropy": 5.647254276275635,
      "epoch": 0.46099642715363237,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498519847524242,
      "loss": 5.5415,
      "mean_token_accuracy": 0.13873762339353563,
      "num_tokens": 10205200.0,
      "step": 4645
    },
    {
      "entropy": 5.642165088653565,
      "epoch": 0.4614926558157999,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985157873228762,
      "loss": 5.58,
      "mean_token_accuracy": 0.13976380601525307,
      "num_tokens": 10216848.0,
      "step": 4650
    },
    {
      "entropy": 5.7167640209198,
      "epoch": 0.46198888447796743,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004985117215788005,
      "loss": 5.6333,
      "mean_token_accuracy": 0.13801254853606224,
      "num_tokens": 10228188.0,
      "step": 4655
    },
    {
      "entropy": 5.695525503158569,
      "epoch": 0.46248511314013496,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985076502921158,
      "loss": 5.7119,
      "mean_token_accuracy": 0.13519898056983948,
      "num_tokens": 10238253.0,
      "step": 4660
    },
    {
      "entropy": 5.7912256717681885,
      "epoch": 0.4629813418023025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498503573462923,
      "loss": 5.6643,
      "mean_token_accuracy": 0.13310108035802842,
      "num_tokens": 10248169.0,
      "step": 4665
    },
    {
      "entropy": 5.646713399887085,
      "epoch": 0.46347757046447,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984994910913232,
      "loss": 5.5364,
      "mean_token_accuracy": 0.13920091986656188,
      "num_tokens": 10258962.0,
      "step": 4670
    },
    {
      "entropy": 5.681157016754151,
      "epoch": 0.46397379912663755,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004984954031774175,
      "loss": 5.6515,
      "mean_token_accuracy": 0.14098910093307496,
      "num_tokens": 10270303.0,
      "step": 4675
    },
    {
      "entropy": 5.650723314285278,
      "epoch": 0.4644700277888051,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984913097213075,
      "loss": 5.5564,
      "mean_token_accuracy": 0.1381896175444126,
      "num_tokens": 10281782.0,
      "step": 4680
    },
    {
      "entropy": 5.697708034515381,
      "epoch": 0.4649662564509726,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984872107230945,
      "loss": 5.5888,
      "mean_token_accuracy": 0.14147323966026307,
      "num_tokens": 10292937.0,
      "step": 4685
    },
    {
      "entropy": 5.6948168754577635,
      "epoch": 0.46546248511314015,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984831061828803,
      "loss": 5.6582,
      "mean_token_accuracy": 0.13611662909388542,
      "num_tokens": 10305443.0,
      "step": 4690
    },
    {
      "entropy": 5.675160598754883,
      "epoch": 0.4659587137753077,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984789961007667,
      "loss": 5.6069,
      "mean_token_accuracy": 0.1387884348630905,
      "num_tokens": 10316489.0,
      "step": 4695
    },
    {
      "entropy": 5.629416275024414,
      "epoch": 0.4664549424374752,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984748804768554,
      "loss": 5.6583,
      "mean_token_accuracy": 0.1354706920683384,
      "num_tokens": 10326690.0,
      "step": 4700
    },
    {
      "entropy": 5.758411979675293,
      "epoch": 0.46695117109964274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984707593112487,
      "loss": 5.6436,
      "mean_token_accuracy": 0.13553577437996864,
      "num_tokens": 10337059.0,
      "step": 4705
    },
    {
      "entropy": 5.687545251846314,
      "epoch": 0.46744739976181027,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004984666326040487,
      "loss": 5.5146,
      "mean_token_accuracy": 0.14385205060243605,
      "num_tokens": 10347035.0,
      "step": 4710
    },
    {
      "entropy": 5.659956359863282,
      "epoch": 0.46794362842397774,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984625003553578,
      "loss": 5.6523,
      "mean_token_accuracy": 0.13844538107514381,
      "num_tokens": 10359492.0,
      "step": 4715
    },
    {
      "entropy": 5.71567964553833,
      "epoch": 0.4684398570861453,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984583625652783,
      "loss": 5.6339,
      "mean_token_accuracy": 0.13988834768533706,
      "num_tokens": 10370128.0,
      "step": 4720
    },
    {
      "entropy": 5.68728232383728,
      "epoch": 0.4689360857483128,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498454219233913,
      "loss": 5.6302,
      "mean_token_accuracy": 0.13606520891189575,
      "num_tokens": 10380276.0,
      "step": 4725
    },
    {
      "entropy": 5.617489004135132,
      "epoch": 0.46943231441048033,
      "grad_norm": 1.0,
      "learning_rate": 0.0004984500703613645,
      "loss": 5.621,
      "mean_token_accuracy": 0.1403240218758583,
      "num_tokens": 10392489.0,
      "step": 4730
    },
    {
      "entropy": 5.8123973369598385,
      "epoch": 0.46992854307264786,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004984459159477358,
      "loss": 5.699,
      "mean_token_accuracy": 0.13760418444871902,
      "num_tokens": 10404259.0,
      "step": 4735
    },
    {
      "entropy": 5.617903470993042,
      "epoch": 0.4704247717348154,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984417559931298,
      "loss": 5.5535,
      "mean_token_accuracy": 0.14490504711866378,
      "num_tokens": 10415000.0,
      "step": 4740
    },
    {
      "entropy": 5.712674140930176,
      "epoch": 0.4709210003969829,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984375904976498,
      "loss": 5.6848,
      "mean_token_accuracy": 0.1319558687508106,
      "num_tokens": 10426078.0,
      "step": 4745
    },
    {
      "entropy": 5.728394031524658,
      "epoch": 0.47141722905915046,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984334194613989,
      "loss": 5.6254,
      "mean_token_accuracy": 0.13398135229945182,
      "num_tokens": 10436838.0,
      "step": 4750
    },
    {
      "entropy": 5.687630033493042,
      "epoch": 0.471913457721318,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984292428844809,
      "loss": 5.5996,
      "mean_token_accuracy": 0.13922261223196983,
      "num_tokens": 10448544.0,
      "step": 4755
    },
    {
      "entropy": 5.7294745445251465,
      "epoch": 0.4724096863834855,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004984250607669989,
      "loss": 5.6673,
      "mean_token_accuracy": 0.13299350515007974,
      "num_tokens": 10459730.0,
      "step": 4760
    },
    {
      "entropy": 5.798430013656616,
      "epoch": 0.47290591504565305,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004984208731090569,
      "loss": 5.7226,
      "mean_token_accuracy": 0.13423294350504875,
      "num_tokens": 10470257.0,
      "step": 4765
    },
    {
      "entropy": 5.7634024143219,
      "epoch": 0.4734021437078206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984166799107587,
      "loss": 5.6724,
      "mean_token_accuracy": 0.13788133040070533,
      "num_tokens": 10480835.0,
      "step": 4770
    },
    {
      "entropy": 5.634775018692016,
      "epoch": 0.4738983723699881,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004984124811722081,
      "loss": 5.707,
      "mean_token_accuracy": 0.1415470227599144,
      "num_tokens": 10491720.0,
      "step": 4775
    },
    {
      "entropy": 5.792123174667358,
      "epoch": 0.47439460103215564,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984082768935096,
      "loss": 5.7232,
      "mean_token_accuracy": 0.13873425871133804,
      "num_tokens": 10502313.0,
      "step": 4780
    },
    {
      "entropy": 5.727884674072266,
      "epoch": 0.47489082969432317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984040670747672,
      "loss": 5.602,
      "mean_token_accuracy": 0.14081369638442992,
      "num_tokens": 10513333.0,
      "step": 4785
    },
    {
      "entropy": 5.630955696105957,
      "epoch": 0.47538705835649064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983998517160852,
      "loss": 5.5825,
      "mean_token_accuracy": 0.14495161324739456,
      "num_tokens": 10522766.0,
      "step": 4790
    },
    {
      "entropy": 5.686036777496338,
      "epoch": 0.4758832870186582,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004983956308175684,
      "loss": 5.6204,
      "mean_token_accuracy": 0.138974317163229,
      "num_tokens": 10534362.0,
      "step": 4795
    },
    {
      "entropy": 5.655626821517944,
      "epoch": 0.4763795156808257,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983914043793212,
      "loss": 5.6309,
      "mean_token_accuracy": 0.1388360910117626,
      "num_tokens": 10545072.0,
      "step": 4800
    },
    {
      "entropy": 5.630531120300293,
      "epoch": 0.47687574434299324,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983871724014487,
      "loss": 5.5675,
      "mean_token_accuracy": 0.1337438069283962,
      "num_tokens": 10555953.0,
      "step": 4805
    },
    {
      "entropy": 5.7464807510375975,
      "epoch": 0.47737197300516077,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004983829348840556,
      "loss": 5.6218,
      "mean_token_accuracy": 0.14246708303689956,
      "num_tokens": 10567757.0,
      "step": 4810
    },
    {
      "entropy": 5.569893312454224,
      "epoch": 0.4778682016673283,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498378691827247,
      "loss": 5.516,
      "mean_token_accuracy": 0.14553011506795882,
      "num_tokens": 10579390.0,
      "step": 4815
    },
    {
      "entropy": 5.7074120998382565,
      "epoch": 0.4783644303294958,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983744432311282,
      "loss": 5.7099,
      "mean_token_accuracy": 0.13435145616531372,
      "num_tokens": 10589624.0,
      "step": 4820
    },
    {
      "entropy": 5.779653644561767,
      "epoch": 0.47886065899166336,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983701890958046,
      "loss": 5.686,
      "mean_token_accuracy": 0.13085080310702324,
      "num_tokens": 10601927.0,
      "step": 4825
    },
    {
      "entropy": 5.600057601928711,
      "epoch": 0.4793568876538309,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004983659294213815,
      "loss": 5.5707,
      "mean_token_accuracy": 0.1411229468882084,
      "num_tokens": 10614073.0,
      "step": 4830
    },
    {
      "entropy": 5.6393006324768065,
      "epoch": 0.4798531163159984,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983616642079649,
      "loss": 5.6096,
      "mean_token_accuracy": 0.1381899058818817,
      "num_tokens": 10625535.0,
      "step": 4835
    },
    {
      "entropy": 5.789355182647705,
      "epoch": 0.48034934497816595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983573934556601,
      "loss": 5.5666,
      "mean_token_accuracy": 0.1391058512032032,
      "num_tokens": 10637160.0,
      "step": 4840
    },
    {
      "entropy": 5.716371202468872,
      "epoch": 0.4808455736403335,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983531171645734,
      "loss": 5.6816,
      "mean_token_accuracy": 0.1327929750084877,
      "num_tokens": 10649061.0,
      "step": 4845
    },
    {
      "entropy": 5.577041244506836,
      "epoch": 0.481341802302501,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983488353348106,
      "loss": 5.549,
      "mean_token_accuracy": 0.13940927758812904,
      "num_tokens": 10661130.0,
      "step": 4850
    },
    {
      "entropy": 5.676693344116211,
      "epoch": 0.48183803096466854,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983445479664779,
      "loss": 5.603,
      "mean_token_accuracy": 0.1405676446855068,
      "num_tokens": 10673017.0,
      "step": 4855
    },
    {
      "entropy": 5.712483978271484,
      "epoch": 0.48233425962683607,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983402550596817,
      "loss": 5.6683,
      "mean_token_accuracy": 0.13832890316843988,
      "num_tokens": 10683268.0,
      "step": 4860
    },
    {
      "entropy": 5.790274238586425,
      "epoch": 0.48283048828900355,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983359566145285,
      "loss": 5.6758,
      "mean_token_accuracy": 0.13833464533090592,
      "num_tokens": 10694220.0,
      "step": 4865
    },
    {
      "entropy": 5.6954034805297855,
      "epoch": 0.4833267169511711,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983316526311248,
      "loss": 5.5939,
      "mean_token_accuracy": 0.1338975928723812,
      "num_tokens": 10704829.0,
      "step": 4870
    },
    {
      "entropy": 5.71824460029602,
      "epoch": 0.4838229456133386,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004983273431095773,
      "loss": 5.7732,
      "mean_token_accuracy": 0.127245507389307,
      "num_tokens": 10717160.0,
      "step": 4875
    },
    {
      "entropy": 5.743475103378296,
      "epoch": 0.48431917427550614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983230280499929,
      "loss": 5.5908,
      "mean_token_accuracy": 0.14171051159501075,
      "num_tokens": 10728638.0,
      "step": 4880
    },
    {
      "entropy": 5.668865203857422,
      "epoch": 0.48481540293767367,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004983187074524787,
      "loss": 5.6492,
      "mean_token_accuracy": 0.13725975155830383,
      "num_tokens": 10740122.0,
      "step": 4885
    },
    {
      "entropy": 5.668728685379028,
      "epoch": 0.4853116315998412,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004983143813171418,
      "loss": 5.5982,
      "mean_token_accuracy": 0.1439150556921959,
      "num_tokens": 10751103.0,
      "step": 4890
    },
    {
      "entropy": 5.711651372909546,
      "epoch": 0.48580786026200873,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004983100496440895,
      "loss": 5.5642,
      "mean_token_accuracy": 0.14204839766025543,
      "num_tokens": 10761937.0,
      "step": 4895
    },
    {
      "entropy": 5.758747482299805,
      "epoch": 0.48630408892417626,
      "grad_norm": 0.96875,
      "learning_rate": 0.000498305712433429,
      "loss": 5.7783,
      "mean_token_accuracy": 0.13372551202774047,
      "num_tokens": 10773527.0,
      "step": 4900
    },
    {
      "entropy": 5.714013719558716,
      "epoch": 0.4868003175863438,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498301369685268,
      "loss": 5.6167,
      "mean_token_accuracy": 0.1352584406733513,
      "num_tokens": 10783873.0,
      "step": 4905
    },
    {
      "entropy": 5.625506639480591,
      "epoch": 0.4872965462485113,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004982970213997142,
      "loss": 5.4722,
      "mean_token_accuracy": 0.14466914534568787,
      "num_tokens": 10796124.0,
      "step": 4910
    },
    {
      "entropy": 5.614594602584839,
      "epoch": 0.48779277491067885,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982926675768756,
      "loss": 5.5741,
      "mean_token_accuracy": 0.14142126515507697,
      "num_tokens": 10806468.0,
      "step": 4915
    },
    {
      "entropy": 5.673980665206909,
      "epoch": 0.4882890035728464,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982883082168599,
      "loss": 5.6164,
      "mean_token_accuracy": 0.1384105958044529,
      "num_tokens": 10817126.0,
      "step": 4920
    },
    {
      "entropy": 5.736818456649781,
      "epoch": 0.4887852322350139,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982839433197752,
      "loss": 5.6523,
      "mean_token_accuracy": 0.13175676092505456,
      "num_tokens": 10828985.0,
      "step": 4925
    },
    {
      "entropy": 5.645940256118775,
      "epoch": 0.48928146089718144,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049827957288573,
      "loss": 5.5669,
      "mean_token_accuracy": 0.13916460573673248,
      "num_tokens": 10840528.0,
      "step": 4930
    },
    {
      "entropy": 5.6961616516113285,
      "epoch": 0.48977768955934897,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982751969148324,
      "loss": 5.5796,
      "mean_token_accuracy": 0.13913155645132064,
      "num_tokens": 10851312.0,
      "step": 4935
    },
    {
      "entropy": 5.611951398849487,
      "epoch": 0.49027391822151645,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982708154071912,
      "loss": 5.5415,
      "mean_token_accuracy": 0.14414673000574113,
      "num_tokens": 10861882.0,
      "step": 4940
    },
    {
      "entropy": 5.652145767211914,
      "epoch": 0.490770146883684,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004982664283629147,
      "loss": 5.6924,
      "mean_token_accuracy": 0.13200924098491668,
      "num_tokens": 10873187.0,
      "step": 4945
    },
    {
      "entropy": 5.704933023452758,
      "epoch": 0.4912663755458515,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982620357821118,
      "loss": 5.5722,
      "mean_token_accuracy": 0.14038059413433074,
      "num_tokens": 10884254.0,
      "step": 4950
    },
    {
      "entropy": 5.631897354125977,
      "epoch": 0.49176260420801904,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004982576376648918,
      "loss": 5.5608,
      "mean_token_accuracy": 0.14028386399149895,
      "num_tokens": 10895770.0,
      "step": 4955
    },
    {
      "entropy": 5.675318670272827,
      "epoch": 0.49225883287018657,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982532340113633,
      "loss": 5.5861,
      "mean_token_accuracy": 0.1447230599820614,
      "num_tokens": 10906436.0,
      "step": 4960
    },
    {
      "entropy": 5.669725561141968,
      "epoch": 0.4927550615323541,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982488248216357,
      "loss": 5.547,
      "mean_token_accuracy": 0.14086605086922646,
      "num_tokens": 10917655.0,
      "step": 4965
    },
    {
      "entropy": 5.621703243255615,
      "epoch": 0.49325129019452163,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982444100958182,
      "loss": 5.5673,
      "mean_token_accuracy": 0.1466407895088196,
      "num_tokens": 10928359.0,
      "step": 4970
    },
    {
      "entropy": 5.687765741348267,
      "epoch": 0.49374751885668916,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982399898340205,
      "loss": 5.7322,
      "mean_token_accuracy": 0.13373062685132026,
      "num_tokens": 10940494.0,
      "step": 4975
    },
    {
      "entropy": 5.796570491790772,
      "epoch": 0.4942437475188567,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498235564036352,
      "loss": 5.7147,
      "mean_token_accuracy": 0.13257310763001443,
      "num_tokens": 10951298.0,
      "step": 4980
    },
    {
      "entropy": 5.7094841480255125,
      "epoch": 0.4947399761810242,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982311327029226,
      "loss": 5.6223,
      "mean_token_accuracy": 0.14185833930969238,
      "num_tokens": 10961680.0,
      "step": 4985
    },
    {
      "entropy": 5.760863256454468,
      "epoch": 0.49523620484319175,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004982266958338421,
      "loss": 5.7317,
      "mean_token_accuracy": 0.1363532744348049,
      "num_tokens": 10974968.0,
      "step": 4990
    },
    {
      "entropy": 5.669859409332275,
      "epoch": 0.4957324335053593,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982222534292205,
      "loss": 5.6744,
      "mean_token_accuracy": 0.1317122034728527,
      "num_tokens": 10986383.0,
      "step": 4995
    },
    {
      "entropy": 5.73351149559021,
      "epoch": 0.4962286621675268,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004982178054891681,
      "loss": 5.7072,
      "mean_token_accuracy": 0.13707910105586052,
      "num_tokens": 10998384.0,
      "step": 5000
    },
    {
      "entropy": 5.655369520187378,
      "epoch": 0.49672489082969434,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982133520137951,
      "loss": 5.4711,
      "mean_token_accuracy": 0.14474554359912872,
      "num_tokens": 11009420.0,
      "step": 5005
    },
    {
      "entropy": 5.682604455947876,
      "epoch": 0.4972211194918619,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982088930032119,
      "loss": 5.6522,
      "mean_token_accuracy": 0.13571663498878478,
      "num_tokens": 11019601.0,
      "step": 5010
    },
    {
      "entropy": 5.638244915008545,
      "epoch": 0.4977173481540294,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004982044284575293,
      "loss": 5.5649,
      "mean_token_accuracy": 0.14048463478684425,
      "num_tokens": 11031138.0,
      "step": 5015
    },
    {
      "entropy": 5.659575700759888,
      "epoch": 0.4982135768161969,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981999583768576,
      "loss": 5.5857,
      "mean_token_accuracy": 0.1401713401079178,
      "num_tokens": 11041643.0,
      "step": 5020
    },
    {
      "entropy": 5.65928955078125,
      "epoch": 0.4987098054783644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498195482761308,
      "loss": 5.5992,
      "mean_token_accuracy": 0.1333266206085682,
      "num_tokens": 11052116.0,
      "step": 5025
    },
    {
      "entropy": 5.857709455490112,
      "epoch": 0.49920603414053194,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981910016109914,
      "loss": 5.7914,
      "mean_token_accuracy": 0.13015363439917566,
      "num_tokens": 11064796.0,
      "step": 5030
    },
    {
      "entropy": 5.7740106105804445,
      "epoch": 0.49970226280269947,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004981865149260189,
      "loss": 5.6277,
      "mean_token_accuracy": 0.1381896987557411,
      "num_tokens": 11075029.0,
      "step": 5035
    },
    {
      "entropy": 5.702889013290405,
      "epoch": 0.5001984914648671,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981820227065016,
      "loss": 5.6478,
      "mean_token_accuracy": 0.13852666169404984,
      "num_tokens": 11086702.0,
      "step": 5040
    },
    {
      "entropy": 5.650403547286987,
      "epoch": 0.5006947201270345,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981775249525513,
      "loss": 5.6376,
      "mean_token_accuracy": 0.13566814363002777,
      "num_tokens": 11097417.0,
      "step": 5045
    },
    {
      "entropy": 5.620085144042969,
      "epoch": 0.5011909487892021,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981730216642791,
      "loss": 5.5452,
      "mean_token_accuracy": 0.1435072675347328,
      "num_tokens": 11108458.0,
      "step": 5050
    },
    {
      "entropy": 5.6926917552948,
      "epoch": 0.5016871774513696,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981685128417969,
      "loss": 5.7132,
      "mean_token_accuracy": 0.13271809369325638,
      "num_tokens": 11121466.0,
      "step": 5055
    },
    {
      "entropy": 5.681808233261108,
      "epoch": 0.5021834061135371,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004981639984852165,
      "loss": 5.6499,
      "mean_token_accuracy": 0.13097651302814484,
      "num_tokens": 11132489.0,
      "step": 5060
    },
    {
      "entropy": 5.709500074386597,
      "epoch": 0.5026796347757047,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981594785946498,
      "loss": 5.6078,
      "mean_token_accuracy": 0.13881729319691657,
      "num_tokens": 11143087.0,
      "step": 5065
    },
    {
      "entropy": 5.733527803421021,
      "epoch": 0.5031758634378721,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981549531702089,
      "loss": 5.6467,
      "mean_token_accuracy": 0.132865297049284,
      "num_tokens": 11154229.0,
      "step": 5070
    },
    {
      "entropy": 5.670539712905883,
      "epoch": 0.5036720921000397,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498150422212006,
      "loss": 5.6091,
      "mean_token_accuracy": 0.1388283282518387,
      "num_tokens": 11165698.0,
      "step": 5075
    },
    {
      "entropy": 5.66808500289917,
      "epoch": 0.5041683207622072,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004981458857201534,
      "loss": 5.5995,
      "mean_token_accuracy": 0.1360068902373314,
      "num_tokens": 11177025.0,
      "step": 5080
    },
    {
      "entropy": 5.711760187149048,
      "epoch": 0.5046645494243748,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004981413436947639,
      "loss": 5.6881,
      "mean_token_accuracy": 0.13852130770683288,
      "num_tokens": 11188766.0,
      "step": 5085
    },
    {
      "entropy": 5.712253618240356,
      "epoch": 0.5051607780865423,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981367961359496,
      "loss": 5.6996,
      "mean_token_accuracy": 0.12998327538371085,
      "num_tokens": 11201626.0,
      "step": 5090
    },
    {
      "entropy": 5.721598958969116,
      "epoch": 0.5056570067487098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981322430438239,
      "loss": 5.6179,
      "mean_token_accuracy": 0.13517760410904883,
      "num_tokens": 11213975.0,
      "step": 5095
    },
    {
      "entropy": 5.614361000061035,
      "epoch": 0.5061532354108773,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981276844184991,
      "loss": 5.5788,
      "mean_token_accuracy": 0.14420121163129807,
      "num_tokens": 11223636.0,
      "step": 5100
    },
    {
      "entropy": 5.614958047866821,
      "epoch": 0.5066494640730449,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981231202600886,
      "loss": 5.5454,
      "mean_token_accuracy": 0.14166136980056762,
      "num_tokens": 11233549.0,
      "step": 5105
    },
    {
      "entropy": 5.635168075561523,
      "epoch": 0.5071456927352124,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981185505687056,
      "loss": 5.5515,
      "mean_token_accuracy": 0.1432757556438446,
      "num_tokens": 11242915.0,
      "step": 5110
    },
    {
      "entropy": 5.779569911956787,
      "epoch": 0.50764192139738,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981139753444632,
      "loss": 5.6672,
      "mean_token_accuracy": 0.13216919153928758,
      "num_tokens": 11254370.0,
      "step": 5115
    },
    {
      "entropy": 5.714865493774414,
      "epoch": 0.5081381500595474,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981093945874752,
      "loss": 5.6075,
      "mean_token_accuracy": 0.14307128861546517,
      "num_tokens": 11264226.0,
      "step": 5120
    },
    {
      "entropy": 5.591129541397095,
      "epoch": 0.508634378721715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981048082978547,
      "loss": 5.5089,
      "mean_token_accuracy": 0.14743735790252685,
      "num_tokens": 11275304.0,
      "step": 5125
    },
    {
      "entropy": 5.604630851745606,
      "epoch": 0.5091306073838825,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981002164757158,
      "loss": 5.582,
      "mean_token_accuracy": 0.13893767669796944,
      "num_tokens": 11286179.0,
      "step": 5130
    },
    {
      "entropy": 5.739297389984131,
      "epoch": 0.50962683604605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980956191211723,
      "loss": 5.6016,
      "mean_token_accuracy": 0.138123244792223,
      "num_tokens": 11297244.0,
      "step": 5135
    },
    {
      "entropy": 5.725628232955932,
      "epoch": 0.5101230647082176,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980910162343382,
      "loss": 5.604,
      "mean_token_accuracy": 0.1446719452738762,
      "num_tokens": 11307251.0,
      "step": 5140
    },
    {
      "entropy": 5.683520317077637,
      "epoch": 0.510619293370385,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004980864078153275,
      "loss": 5.5841,
      "mean_token_accuracy": 0.13879474326968194,
      "num_tokens": 11318441.0,
      "step": 5145
    },
    {
      "entropy": 5.578943824768066,
      "epoch": 0.5111155220325526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980817938642548,
      "loss": 5.5705,
      "mean_token_accuracy": 0.14160208776593208,
      "num_tokens": 11328743.0,
      "step": 5150
    },
    {
      "entropy": 5.591680717468262,
      "epoch": 0.5116117506947201,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498077174381234,
      "loss": 5.5533,
      "mean_token_accuracy": 0.14184579849243165,
      "num_tokens": 11340745.0,
      "step": 5155
    },
    {
      "entropy": 5.689578866958618,
      "epoch": 0.5121079793568877,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004980725493663802,
      "loss": 5.6165,
      "mean_token_accuracy": 0.14150336384773254,
      "num_tokens": 11352060.0,
      "step": 5160
    },
    {
      "entropy": 5.689644479751587,
      "epoch": 0.5126042080190552,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980679188198078,
      "loss": 5.6365,
      "mean_token_accuracy": 0.13577819392085075,
      "num_tokens": 11363097.0,
      "step": 5165
    },
    {
      "entropy": 5.711014699935913,
      "epoch": 0.5131004366812227,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980632827416317,
      "loss": 5.6585,
      "mean_token_accuracy": 0.14745849817991258,
      "num_tokens": 11373410.0,
      "step": 5170
    },
    {
      "entropy": 5.753261661529541,
      "epoch": 0.5135966653433902,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004980586411319669,
      "loss": 5.7296,
      "mean_token_accuracy": 0.13415995016694068,
      "num_tokens": 11385615.0,
      "step": 5175
    },
    {
      "entropy": 5.596803760528564,
      "epoch": 0.5140928940055578,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004980539939909283,
      "loss": 5.5276,
      "mean_token_accuracy": 0.13587600886821746,
      "num_tokens": 11396492.0,
      "step": 5180
    },
    {
      "entropy": 5.7470043182373045,
      "epoch": 0.5145891226677253,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004980493413186314,
      "loss": 5.6269,
      "mean_token_accuracy": 0.1332610622048378,
      "num_tokens": 11406377.0,
      "step": 5185
    },
    {
      "entropy": 5.6779357433319095,
      "epoch": 0.5150853513298929,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980446831151914,
      "loss": 5.6101,
      "mean_token_accuracy": 0.14095872342586518,
      "num_tokens": 11416862.0,
      "step": 5190
    },
    {
      "entropy": 5.6290970802307125,
      "epoch": 0.5155815799920603,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980400193807239,
      "loss": 5.5687,
      "mean_token_accuracy": 0.14431979805231093,
      "num_tokens": 11428246.0,
      "step": 5195
    },
    {
      "entropy": 5.705343961715698,
      "epoch": 0.5160778086542279,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980353501153446,
      "loss": 5.6012,
      "mean_token_accuracy": 0.13967292085289956,
      "num_tokens": 11438147.0,
      "step": 5200
    },
    {
      "entropy": 5.709432315826416,
      "epoch": 0.5165740373163954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980306753191691,
      "loss": 5.7081,
      "mean_token_accuracy": 0.1305124878883362,
      "num_tokens": 11449154.0,
      "step": 5205
    },
    {
      "entropy": 5.751299238204956,
      "epoch": 0.5170702659785629,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980259949923135,
      "loss": 5.6124,
      "mean_token_accuracy": 0.13552996069192885,
      "num_tokens": 11459262.0,
      "step": 5210
    },
    {
      "entropy": 5.671953725814819,
      "epoch": 0.5175664946407305,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004980213091348938,
      "loss": 5.5976,
      "mean_token_accuracy": 0.13675811737775803,
      "num_tokens": 11471991.0,
      "step": 5215
    },
    {
      "entropy": 5.656467247009277,
      "epoch": 0.5180627233028979,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004980166177470262,
      "loss": 5.5399,
      "mean_token_accuracy": 0.13962042704224586,
      "num_tokens": 11483176.0,
      "step": 5220
    },
    {
      "entropy": 5.702274036407471,
      "epoch": 0.5185589519650655,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980119208288271,
      "loss": 5.6268,
      "mean_token_accuracy": 0.13966499343514444,
      "num_tokens": 11494015.0,
      "step": 5225
    },
    {
      "entropy": 5.585013294219971,
      "epoch": 0.519055180627233,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004980072183804128,
      "loss": 5.5451,
      "mean_token_accuracy": 0.13971148952841758,
      "num_tokens": 11506365.0,
      "step": 5230
    },
    {
      "entropy": 5.696497058868408,
      "epoch": 0.5195514092894006,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980025104019,
      "loss": 5.642,
      "mean_token_accuracy": 0.14106786623597145,
      "num_tokens": 11517199.0,
      "step": 5235
    },
    {
      "entropy": 5.651773166656494,
      "epoch": 0.520047637951568,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004979977968934056,
      "loss": 5.4995,
      "mean_token_accuracy": 0.14559734538197516,
      "num_tokens": 11527591.0,
      "step": 5240
    },
    {
      "entropy": 5.688210153579712,
      "epoch": 0.5205438666137356,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004979930778550463,
      "loss": 5.665,
      "mean_token_accuracy": 0.1405502274632454,
      "num_tokens": 11538657.0,
      "step": 5245
    },
    {
      "entropy": 5.663222169876098,
      "epoch": 0.5210400952759031,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979883532869392,
      "loss": 5.4846,
      "mean_token_accuracy": 0.15589443147182463,
      "num_tokens": 11549513.0,
      "step": 5250
    },
    {
      "entropy": 5.668476867675781,
      "epoch": 0.5215363239380707,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979836231892013,
      "loss": 5.5658,
      "mean_token_accuracy": 0.1353978134691715,
      "num_tokens": 11560376.0,
      "step": 5255
    },
    {
      "entropy": 5.587079095840454,
      "epoch": 0.5220325526002382,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979788875619501,
      "loss": 5.557,
      "mean_token_accuracy": 0.13797224387526513,
      "num_tokens": 11571218.0,
      "step": 5260
    },
    {
      "entropy": 5.676311302185058,
      "epoch": 0.5225287812624058,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497974146405303,
      "loss": 5.6194,
      "mean_token_accuracy": 0.1385549485683441,
      "num_tokens": 11581981.0,
      "step": 5265
    },
    {
      "entropy": 5.653447818756104,
      "epoch": 0.5230250099245732,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979693997193776,
      "loss": 5.5821,
      "mean_token_accuracy": 0.13908568024635315,
      "num_tokens": 11592561.0,
      "step": 5270
    },
    {
      "entropy": 5.684787130355835,
      "epoch": 0.5235212385867408,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004979646475042914,
      "loss": 5.5489,
      "mean_token_accuracy": 0.14116989001631736,
      "num_tokens": 11601933.0,
      "step": 5275
    },
    {
      "entropy": 5.618759107589722,
      "epoch": 0.5240174672489083,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979598897601625,
      "loss": 5.4843,
      "mean_token_accuracy": 0.14859692752361298,
      "num_tokens": 11613938.0,
      "step": 5280
    },
    {
      "entropy": 5.7467328071594235,
      "epoch": 0.5245136959110758,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004979551264871088,
      "loss": 5.6766,
      "mean_token_accuracy": 0.13326044529676437,
      "num_tokens": 11624890.0,
      "step": 5285
    },
    {
      "entropy": 5.669732666015625,
      "epoch": 0.5250099245732434,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979503576852482,
      "loss": 5.6504,
      "mean_token_accuracy": 0.1438838467001915,
      "num_tokens": 11635080.0,
      "step": 5290
    },
    {
      "entropy": 5.576575374603271,
      "epoch": 0.5255061532354108,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979455833546994,
      "loss": 5.5554,
      "mean_token_accuracy": 0.14168847277760505,
      "num_tokens": 11645180.0,
      "step": 5295
    },
    {
      "entropy": 5.7158714771270756,
      "epoch": 0.5260023818975784,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979408034955804,
      "loss": 5.5907,
      "mean_token_accuracy": 0.13906485587358475,
      "num_tokens": 11657599.0,
      "step": 5300
    },
    {
      "entropy": 5.743302679061889,
      "epoch": 0.5264986105597459,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004979360181080099,
      "loss": 5.6472,
      "mean_token_accuracy": 0.1282064363360405,
      "num_tokens": 11669177.0,
      "step": 5305
    },
    {
      "entropy": 5.601555204391479,
      "epoch": 0.5269948392219135,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979312271921065,
      "loss": 5.5045,
      "mean_token_accuracy": 0.13916629999876023,
      "num_tokens": 11679756.0,
      "step": 5310
    },
    {
      "entropy": 5.641785907745361,
      "epoch": 0.527491067884081,
      "grad_norm": 1.0,
      "learning_rate": 0.000497926430747989,
      "loss": 5.573,
      "mean_token_accuracy": 0.14440273195505143,
      "num_tokens": 11689805.0,
      "step": 5315
    },
    {
      "entropy": 5.68066782951355,
      "epoch": 0.5279872965462485,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979216287757764,
      "loss": 5.5939,
      "mean_token_accuracy": 0.13892873302102088,
      "num_tokens": 11701150.0,
      "step": 5320
    },
    {
      "entropy": 5.569952583312988,
      "epoch": 0.528483525208416,
      "grad_norm": 0.953125,
      "learning_rate": 0.000497916821275588,
      "loss": 5.5064,
      "mean_token_accuracy": 0.14227857440710068,
      "num_tokens": 11712256.0,
      "step": 5325
    },
    {
      "entropy": 5.684609746932983,
      "epoch": 0.5289797538705836,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979120082475424,
      "loss": 5.5706,
      "mean_token_accuracy": 0.14501353949308396,
      "num_tokens": 11722547.0,
      "step": 5330
    },
    {
      "entropy": 5.679641437530518,
      "epoch": 0.5294759825327511,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979071896917595,
      "loss": 5.5902,
      "mean_token_accuracy": 0.14043826460838318,
      "num_tokens": 11732309.0,
      "step": 5335
    },
    {
      "entropy": 5.695653581619263,
      "epoch": 0.5299722111949187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979023656083586,
      "loss": 5.6359,
      "mean_token_accuracy": 0.13951552361249925,
      "num_tokens": 11743982.0,
      "step": 5340
    },
    {
      "entropy": 5.662636709213257,
      "epoch": 0.5304684398570861,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004978975359974594,
      "loss": 5.5874,
      "mean_token_accuracy": 0.14191087186336518,
      "num_tokens": 11755506.0,
      "step": 5345
    },
    {
      "entropy": 5.689707183837891,
      "epoch": 0.5309646685192537,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978927008591816,
      "loss": 5.5879,
      "mean_token_accuracy": 0.1396624781191349,
      "num_tokens": 11767024.0,
      "step": 5350
    },
    {
      "entropy": 5.733471298217774,
      "epoch": 0.5314608971814212,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004978878601936451,
      "loss": 5.5983,
      "mean_token_accuracy": 0.13833921551704406,
      "num_tokens": 11777625.0,
      "step": 5355
    },
    {
      "entropy": 5.586740970611572,
      "epoch": 0.5319571258435887,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978830140009698,
      "loss": 5.5512,
      "mean_token_accuracy": 0.1478976860642433,
      "num_tokens": 11788610.0,
      "step": 5360
    },
    {
      "entropy": 5.6329813480377195,
      "epoch": 0.5324533545057563,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978781622812762,
      "loss": 5.665,
      "mean_token_accuracy": 0.1432223603129387,
      "num_tokens": 11800828.0,
      "step": 5365
    },
    {
      "entropy": 5.692148733139038,
      "epoch": 0.5329495831679237,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978733050346844,
      "loss": 5.6581,
      "mean_token_accuracy": 0.1416023150086403,
      "num_tokens": 11811729.0,
      "step": 5370
    },
    {
      "entropy": 5.688501024246216,
      "epoch": 0.5334458118300913,
      "grad_norm": 1.125,
      "learning_rate": 0.0004978684422613149,
      "loss": 5.6875,
      "mean_token_accuracy": 0.13832212015986442,
      "num_tokens": 11822081.0,
      "step": 5375
    },
    {
      "entropy": 5.696527004241943,
      "epoch": 0.5339420404922588,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978635739612882,
      "loss": 5.5622,
      "mean_token_accuracy": 0.14342110902070998,
      "num_tokens": 11832984.0,
      "step": 5380
    },
    {
      "entropy": 5.6571948528289795,
      "epoch": 0.5344382691544264,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978587001347252,
      "loss": 5.5092,
      "mean_token_accuracy": 0.1376336097717285,
      "num_tokens": 11842918.0,
      "step": 5385
    },
    {
      "entropy": 5.60179328918457,
      "epoch": 0.5349344978165939,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004978538207817465,
      "loss": 5.5555,
      "mean_token_accuracy": 0.14260684698820114,
      "num_tokens": 11854364.0,
      "step": 5390
    },
    {
      "entropy": 5.65335168838501,
      "epoch": 0.5354307264787614,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004978489359024734,
      "loss": 5.6382,
      "mean_token_accuracy": 0.1374676614999771,
      "num_tokens": 11866400.0,
      "step": 5395
    },
    {
      "entropy": 5.74446587562561,
      "epoch": 0.5359269551409289,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004978440454970268,
      "loss": 5.6181,
      "mean_token_accuracy": 0.13733726665377616,
      "num_tokens": 11877094.0,
      "step": 5400
    },
    {
      "entropy": 5.667227172851563,
      "epoch": 0.5364231838030965,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978391495655281,
      "loss": 5.5841,
      "mean_token_accuracy": 0.1390235625207424,
      "num_tokens": 11887289.0,
      "step": 5405
    },
    {
      "entropy": 5.692342138290405,
      "epoch": 0.536919412465264,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004978342481080987,
      "loss": 5.555,
      "mean_token_accuracy": 0.1351594567298889,
      "num_tokens": 11897700.0,
      "step": 5410
    },
    {
      "entropy": 5.5396300792694095,
      "epoch": 0.5374156411274316,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978293411248601,
      "loss": 5.4873,
      "mean_token_accuracy": 0.14424988850951195,
      "num_tokens": 11908623.0,
      "step": 5415
    },
    {
      "entropy": 5.636839294433594,
      "epoch": 0.537911869789599,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004978244286159339,
      "loss": 5.5237,
      "mean_token_accuracy": 0.14568247199058532,
      "num_tokens": 11917883.0,
      "step": 5420
    },
    {
      "entropy": 5.5925007343292235,
      "epoch": 0.5384080984517666,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004978195105814421,
      "loss": 5.548,
      "mean_token_accuracy": 0.14374667555093765,
      "num_tokens": 11929145.0,
      "step": 5425
    },
    {
      "entropy": 5.692426109313965,
      "epoch": 0.5389043271139341,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004978145870215065,
      "loss": 5.5511,
      "mean_token_accuracy": 0.13841745555400847,
      "num_tokens": 11940073.0,
      "step": 5430
    },
    {
      "entropy": 5.681694412231446,
      "epoch": 0.5394005557761016,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978096579362493,
      "loss": 5.6635,
      "mean_token_accuracy": 0.13862710148096086,
      "num_tokens": 11953140.0,
      "step": 5435
    },
    {
      "entropy": 5.658891391754151,
      "epoch": 0.5398967844382692,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004978047233257927,
      "loss": 5.587,
      "mean_token_accuracy": 0.14590081721544265,
      "num_tokens": 11963853.0,
      "step": 5440
    },
    {
      "entropy": 5.626586198806763,
      "epoch": 0.5403930131004366,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004977997831902591,
      "loss": 5.5616,
      "mean_token_accuracy": 0.1489775650203228,
      "num_tokens": 11975469.0,
      "step": 5445
    },
    {
      "entropy": 5.679513597488404,
      "epoch": 0.5408892417626042,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004977948375297709,
      "loss": 5.5389,
      "mean_token_accuracy": 0.14958526194095612,
      "num_tokens": 11985227.0,
      "step": 5450
    },
    {
      "entropy": 5.535920810699463,
      "epoch": 0.5413854704247717,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004977898863444509,
      "loss": 5.5404,
      "mean_token_accuracy": 0.14351814836263657,
      "num_tokens": 11995459.0,
      "step": 5455
    },
    {
      "entropy": 5.629694604873658,
      "epoch": 0.5418816990869393,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004977849296344217,
      "loss": 5.5895,
      "mean_token_accuracy": 0.14124198332428933,
      "num_tokens": 12007705.0,
      "step": 5460
    },
    {
      "entropy": 5.60025725364685,
      "epoch": 0.5423779277491068,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004977799673998063,
      "loss": 5.4983,
      "mean_token_accuracy": 0.1446823038160801,
      "num_tokens": 12019729.0,
      "step": 5465
    },
    {
      "entropy": 5.651974487304687,
      "epoch": 0.5428741564112743,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977749996407278,
      "loss": 5.5737,
      "mean_token_accuracy": 0.13683584704995155,
      "num_tokens": 12031009.0,
      "step": 5470
    },
    {
      "entropy": 5.686952304840088,
      "epoch": 0.5433703850734418,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004977700263573094,
      "loss": 5.534,
      "mean_token_accuracy": 0.14341633021831512,
      "num_tokens": 12041396.0,
      "step": 5475
    },
    {
      "entropy": 5.643152189254761,
      "epoch": 0.5438666137356094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977650475496743,
      "loss": 5.6958,
      "mean_token_accuracy": 0.1346547156572342,
      "num_tokens": 12053112.0,
      "step": 5480
    },
    {
      "entropy": 5.705150461196899,
      "epoch": 0.5443628423977769,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004977600632179459,
      "loss": 5.5545,
      "mean_token_accuracy": 0.13762412071228028,
      "num_tokens": 12064823.0,
      "step": 5485
    },
    {
      "entropy": 5.654553079605103,
      "epoch": 0.5448590710599445,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977550733622482,
      "loss": 5.6758,
      "mean_token_accuracy": 0.13756117820739747,
      "num_tokens": 12076656.0,
      "step": 5490
    },
    {
      "entropy": 5.667914390563965,
      "epoch": 0.5453552997221119,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977500779827045,
      "loss": 5.4727,
      "mean_token_accuracy": 0.14255305528640747,
      "num_tokens": 12086324.0,
      "step": 5495
    },
    {
      "entropy": 5.517617225646973,
      "epoch": 0.5458515283842795,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977450770794389,
      "loss": 5.511,
      "mean_token_accuracy": 0.1395778737962246,
      "num_tokens": 12098076.0,
      "step": 5500
    },
    {
      "entropy": 5.608063220977783,
      "epoch": 0.546347757046447,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004977400706525755,
      "loss": 5.5807,
      "mean_token_accuracy": 0.1360511936247349,
      "num_tokens": 12111282.0,
      "step": 5505
    },
    {
      "entropy": 5.660810661315918,
      "epoch": 0.5468439857086145,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977350587022381,
      "loss": 5.5235,
      "mean_token_accuracy": 0.14769414961338043,
      "num_tokens": 12121932.0,
      "step": 5510
    },
    {
      "entropy": 5.603894424438477,
      "epoch": 0.5473402143707821,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977300412285513,
      "loss": 5.5111,
      "mean_token_accuracy": 0.14588091373443604,
      "num_tokens": 12133040.0,
      "step": 5515
    },
    {
      "entropy": 5.646938419342041,
      "epoch": 0.5478364430329495,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004977250182316394,
      "loss": 5.5897,
      "mean_token_accuracy": 0.14238848984241487,
      "num_tokens": 12142577.0,
      "step": 5520
    },
    {
      "entropy": 5.619527673721313,
      "epoch": 0.5483326716951171,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497719989711627,
      "loss": 5.5633,
      "mean_token_accuracy": 0.1430242106318474,
      "num_tokens": 12153941.0,
      "step": 5525
    },
    {
      "entropy": 5.63915753364563,
      "epoch": 0.5488289003572846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977149556686387,
      "loss": 5.621,
      "mean_token_accuracy": 0.13971951976418495,
      "num_tokens": 12165347.0,
      "step": 5530
    },
    {
      "entropy": 5.595114469528198,
      "epoch": 0.5493251290194522,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977099161027996,
      "loss": 5.4943,
      "mean_token_accuracy": 0.14236187785863877,
      "num_tokens": 12175705.0,
      "step": 5535
    },
    {
      "entropy": 5.639246988296509,
      "epoch": 0.5498213576816197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977048710142342,
      "loss": 5.5436,
      "mean_token_accuracy": 0.14288176596164703,
      "num_tokens": 12186781.0,
      "step": 5540
    },
    {
      "entropy": 5.679701423645019,
      "epoch": 0.5503175863437872,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976998204030681,
      "loss": 5.5818,
      "mean_token_accuracy": 0.13985614106059074,
      "num_tokens": 12197145.0,
      "step": 5545
    },
    {
      "entropy": 5.598316383361817,
      "epoch": 0.5508138150059547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004976947642694262,
      "loss": 5.5746,
      "mean_token_accuracy": 0.143681750446558,
      "num_tokens": 12207637.0,
      "step": 5550
    },
    {
      "entropy": 5.597862482070923,
      "epoch": 0.5513100436681223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004976897026134341,
      "loss": 5.5142,
      "mean_token_accuracy": 0.1443020895123482,
      "num_tokens": 12218156.0,
      "step": 5555
    },
    {
      "entropy": 5.6665167808532715,
      "epoch": 0.5518062723302898,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004976846354352172,
      "loss": 5.5992,
      "mean_token_accuracy": 0.1401934579014778,
      "num_tokens": 12229987.0,
      "step": 5560
    },
    {
      "entropy": 5.672375059127807,
      "epoch": 0.5523025009924574,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976795627349012,
      "loss": 5.569,
      "mean_token_accuracy": 0.13795140534639358,
      "num_tokens": 12240508.0,
      "step": 5565
    },
    {
      "entropy": 5.598812580108643,
      "epoch": 0.5527987296546248,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004976744845126118,
      "loss": 5.5086,
      "mean_token_accuracy": 0.14796727448701857,
      "num_tokens": 12250317.0,
      "step": 5570
    },
    {
      "entropy": 5.604941272735596,
      "epoch": 0.5532949583167924,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976694007684751,
      "loss": 5.5522,
      "mean_token_accuracy": 0.14205480813980104,
      "num_tokens": 12262104.0,
      "step": 5575
    },
    {
      "entropy": 5.717854738235474,
      "epoch": 0.5537911869789599,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497664311502617,
      "loss": 5.5855,
      "mean_token_accuracy": 0.1443232163786888,
      "num_tokens": 12273259.0,
      "step": 5580
    },
    {
      "entropy": 5.621715688705445,
      "epoch": 0.5542874156411274,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976592167151638,
      "loss": 5.6283,
      "mean_token_accuracy": 0.1344245620071888,
      "num_tokens": 12285504.0,
      "step": 5585
    },
    {
      "entropy": 5.691664218902588,
      "epoch": 0.554783644303295,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004976541164062418,
      "loss": 5.5621,
      "mean_token_accuracy": 0.1360063761472702,
      "num_tokens": 12296512.0,
      "step": 5590
    },
    {
      "entropy": 5.55618257522583,
      "epoch": 0.5552798729654624,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004976490105759774,
      "loss": 5.422,
      "mean_token_accuracy": 0.14680247455835344,
      "num_tokens": 12307948.0,
      "step": 5595
    },
    {
      "entropy": 5.599129724502563,
      "epoch": 0.55577610162763,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004976438992244976,
      "loss": 5.5904,
      "mean_token_accuracy": 0.1344291388988495,
      "num_tokens": 12320020.0,
      "step": 5600
    },
    {
      "entropy": 5.646237850189209,
      "epoch": 0.5562723302897975,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976387823519286,
      "loss": 5.4605,
      "mean_token_accuracy": 0.14551302939653396,
      "num_tokens": 12331091.0,
      "step": 5605
    },
    {
      "entropy": 5.672171211242675,
      "epoch": 0.5567685589519651,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004976336599583975,
      "loss": 5.5695,
      "mean_token_accuracy": 0.13842309042811393,
      "num_tokens": 12342621.0,
      "step": 5610
    },
    {
      "entropy": 5.710894536972046,
      "epoch": 0.5572647876141326,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004976285320440315,
      "loss": 5.8307,
      "mean_token_accuracy": 0.13015428707003593,
      "num_tokens": 12353330.0,
      "step": 5615
    },
    {
      "entropy": 5.714410448074341,
      "epoch": 0.5577610162763001,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976233986089577,
      "loss": 5.5702,
      "mean_token_accuracy": 0.1422278732061386,
      "num_tokens": 12364418.0,
      "step": 5620
    },
    {
      "entropy": 5.62553653717041,
      "epoch": 0.5582572449384676,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004976182596533031,
      "loss": 5.5475,
      "mean_token_accuracy": 0.14112602323293685,
      "num_tokens": 12377022.0,
      "step": 5625
    },
    {
      "entropy": 5.702981090545654,
      "epoch": 0.5587534736006352,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004976131151771955,
      "loss": 5.6238,
      "mean_token_accuracy": 0.13408835902810096,
      "num_tokens": 12387799.0,
      "step": 5630
    },
    {
      "entropy": 5.595159196853638,
      "epoch": 0.5592497022628027,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004976079651807623,
      "loss": 5.5573,
      "mean_token_accuracy": 0.13506573513150216,
      "num_tokens": 12397387.0,
      "step": 5635
    },
    {
      "entropy": 5.600511312484741,
      "epoch": 0.5597459309249703,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976028096641313,
      "loss": 5.4878,
      "mean_token_accuracy": 0.14869325160980223,
      "num_tokens": 12407994.0,
      "step": 5640
    },
    {
      "entropy": 5.74589524269104,
      "epoch": 0.5602421595871377,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004975976486274302,
      "loss": 5.6628,
      "mean_token_accuracy": 0.13694743812084198,
      "num_tokens": 12419124.0,
      "step": 5645
    },
    {
      "entropy": 5.674187421798706,
      "epoch": 0.5607383882493053,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004975924820707872,
      "loss": 5.5547,
      "mean_token_accuracy": 0.14174269288778304,
      "num_tokens": 12428699.0,
      "step": 5650
    },
    {
      "entropy": 5.653844833374023,
      "epoch": 0.5612346169114728,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004975873099943302,
      "loss": 5.6174,
      "mean_token_accuracy": 0.13702422231435776,
      "num_tokens": 12439025.0,
      "step": 5655
    },
    {
      "entropy": 5.675407028198242,
      "epoch": 0.5617308455736403,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975821323981875,
      "loss": 5.5836,
      "mean_token_accuracy": 0.14168438836932182,
      "num_tokens": 12449439.0,
      "step": 5660
    },
    {
      "entropy": 5.67562084197998,
      "epoch": 0.5622270742358079,
      "grad_norm": 1.0,
      "learning_rate": 0.0004975769492824877,
      "loss": 5.5282,
      "mean_token_accuracy": 0.14081865772604943,
      "num_tokens": 12460066.0,
      "step": 5665
    },
    {
      "entropy": 5.611133432388305,
      "epoch": 0.5627233028979753,
      "grad_norm": 1.0,
      "learning_rate": 0.0004975717606473589,
      "loss": 5.5553,
      "mean_token_accuracy": 0.14057415798306466,
      "num_tokens": 12471093.0,
      "step": 5670
    },
    {
      "entropy": 5.6840332508087155,
      "epoch": 0.5632195315601429,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004975665664929302,
      "loss": 5.5663,
      "mean_token_accuracy": 0.13668758645653725,
      "num_tokens": 12481181.0,
      "step": 5675
    },
    {
      "entropy": 5.649426221847534,
      "epoch": 0.5637157602223104,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004975613668193303,
      "loss": 5.5932,
      "mean_token_accuracy": 0.1397934749722481,
      "num_tokens": 12493068.0,
      "step": 5680
    },
    {
      "entropy": 5.638387966156006,
      "epoch": 0.564211988884478,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497556161626688,
      "loss": 5.5406,
      "mean_token_accuracy": 0.14199285358190536,
      "num_tokens": 12504025.0,
      "step": 5685
    },
    {
      "entropy": 5.657924365997315,
      "epoch": 0.5647082175466455,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975509509151325,
      "loss": 5.554,
      "mean_token_accuracy": 0.1422862321138382,
      "num_tokens": 12514567.0,
      "step": 5690
    },
    {
      "entropy": 5.7168656349182125,
      "epoch": 0.565204446208813,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004975457346847929,
      "loss": 5.6092,
      "mean_token_accuracy": 0.13971372693777084,
      "num_tokens": 12525088.0,
      "step": 5695
    },
    {
      "entropy": 5.613651514053345,
      "epoch": 0.5657006748709805,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975405129357986,
      "loss": 5.5527,
      "mean_token_accuracy": 0.14063025638461113,
      "num_tokens": 12536002.0,
      "step": 5700
    },
    {
      "entropy": 5.65606837272644,
      "epoch": 0.5661969035331481,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004975352856682792,
      "loss": 5.5765,
      "mean_token_accuracy": 0.14051345139741897,
      "num_tokens": 12547468.0,
      "step": 5705
    },
    {
      "entropy": 5.7039741516113285,
      "epoch": 0.5666931321953156,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004975300528823642,
      "loss": 5.5173,
      "mean_token_accuracy": 0.14239700585603715,
      "num_tokens": 12557670.0,
      "step": 5710
    },
    {
      "entropy": 5.606233406066894,
      "epoch": 0.5671893608574832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004975248145781834,
      "loss": 5.5891,
      "mean_token_accuracy": 0.1419870972633362,
      "num_tokens": 12568425.0,
      "step": 5715
    },
    {
      "entropy": 5.56580696105957,
      "epoch": 0.5676855895196506,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004975195707558667,
      "loss": 5.5525,
      "mean_token_accuracy": 0.14227935597300528,
      "num_tokens": 12579138.0,
      "step": 5720
    },
    {
      "entropy": 5.643411302566529,
      "epoch": 0.5681818181818182,
      "grad_norm": 1.0,
      "learning_rate": 0.0004975143214155443,
      "loss": 5.565,
      "mean_token_accuracy": 0.14773009121418,
      "num_tokens": 12589743.0,
      "step": 5725
    },
    {
      "entropy": 5.697768497467041,
      "epoch": 0.5686780468439857,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004975090665573459,
      "loss": 5.5554,
      "mean_token_accuracy": 0.1414801999926567,
      "num_tokens": 12600648.0,
      "step": 5730
    },
    {
      "entropy": 5.628736925125122,
      "epoch": 0.5691742755061532,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975038061814023,
      "loss": 5.5563,
      "mean_token_accuracy": 0.14277652502059937,
      "num_tokens": 12611369.0,
      "step": 5735
    },
    {
      "entropy": 5.631605100631714,
      "epoch": 0.5696705041683208,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974985402878437,
      "loss": 5.5256,
      "mean_token_accuracy": 0.1400889627635479,
      "num_tokens": 12621992.0,
      "step": 5740
    },
    {
      "entropy": 5.6745518207550045,
      "epoch": 0.5701667328304882,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004974932688768007,
      "loss": 5.6084,
      "mean_token_accuracy": 0.13641005828976632,
      "num_tokens": 12632463.0,
      "step": 5745
    },
    {
      "entropy": 5.578773403167725,
      "epoch": 0.5706629614926558,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004974879919484042,
      "loss": 5.4683,
      "mean_token_accuracy": 0.14433844685554503,
      "num_tokens": 12642282.0,
      "step": 5750
    },
    {
      "entropy": 5.600848770141601,
      "epoch": 0.5711591901548233,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004974827095027848,
      "loss": 5.5055,
      "mean_token_accuracy": 0.14577008932828903,
      "num_tokens": 12653495.0,
      "step": 5755
    },
    {
      "entropy": 5.607797050476075,
      "epoch": 0.5716554188169909,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004974774215400736,
      "loss": 5.4378,
      "mean_token_accuracy": 0.14466027393937111,
      "num_tokens": 12664322.0,
      "step": 5760
    },
    {
      "entropy": 5.465302228927612,
      "epoch": 0.5721516474791584,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004974721280604019,
      "loss": 5.5297,
      "mean_token_accuracy": 0.14000416174530983,
      "num_tokens": 12673981.0,
      "step": 5765
    },
    {
      "entropy": 5.594118261337281,
      "epoch": 0.572647876141326,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004974668290639007,
      "loss": 5.5568,
      "mean_token_accuracy": 0.14178637713193892,
      "num_tokens": 12684778.0,
      "step": 5770
    },
    {
      "entropy": 5.7291247844696045,
      "epoch": 0.5731441048034934,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974615245507014,
      "loss": 5.5139,
      "mean_token_accuracy": 0.14449861347675325,
      "num_tokens": 12695287.0,
      "step": 5775
    },
    {
      "entropy": 5.5744297981262205,
      "epoch": 0.573640333465661,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004974562145209359,
      "loss": 5.5849,
      "mean_token_accuracy": 0.1379687674343586,
      "num_tokens": 12705773.0,
      "step": 5780
    },
    {
      "entropy": 5.574149894714355,
      "epoch": 0.5741365621278285,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004974508989747354,
      "loss": 5.5093,
      "mean_token_accuracy": 0.1443384602665901,
      "num_tokens": 12716783.0,
      "step": 5785
    },
    {
      "entropy": 5.629298830032349,
      "epoch": 0.5746327907899961,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004974455779122321,
      "loss": 5.5771,
      "mean_token_accuracy": 0.14027536660432816,
      "num_tokens": 12728146.0,
      "step": 5790
    },
    {
      "entropy": 5.6705690860748295,
      "epoch": 0.5751290194521635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974402513335577,
      "loss": 5.579,
      "mean_token_accuracy": 0.1366913639008999,
      "num_tokens": 12739683.0,
      "step": 5795
    },
    {
      "entropy": 5.693634462356568,
      "epoch": 0.5756252481143311,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004974349192388444,
      "loss": 5.5689,
      "mean_token_accuracy": 0.13982002437114716,
      "num_tokens": 12750257.0,
      "step": 5800
    },
    {
      "entropy": 5.732922649383545,
      "epoch": 0.5761214767764986,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974295816282245,
      "loss": 5.6395,
      "mean_token_accuracy": 0.13885747790336608,
      "num_tokens": 12761055.0,
      "step": 5805
    },
    {
      "entropy": 5.630460023880005,
      "epoch": 0.5766177054386661,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004974242385018303,
      "loss": 5.5709,
      "mean_token_accuracy": 0.13992611467838287,
      "num_tokens": 12771014.0,
      "step": 5810
    },
    {
      "entropy": 5.592522716522216,
      "epoch": 0.5771139341008337,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497418889859794,
      "loss": 5.4865,
      "mean_token_accuracy": 0.14401526898145675,
      "num_tokens": 12782623.0,
      "step": 5815
    },
    {
      "entropy": 5.690291833877564,
      "epoch": 0.5776101627630011,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004974135357022488,
      "loss": 5.6016,
      "mean_token_accuracy": 0.1452617257833481,
      "num_tokens": 12793533.0,
      "step": 5820
    },
    {
      "entropy": 5.548589134216309,
      "epoch": 0.5781063914251687,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974081760293269,
      "loss": 5.4656,
      "mean_token_accuracy": 0.14004177302122117,
      "num_tokens": 12804389.0,
      "step": 5825
    },
    {
      "entropy": 5.63265175819397,
      "epoch": 0.5786026200873362,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004974028108411617,
      "loss": 5.5731,
      "mean_token_accuracy": 0.14461976215243338,
      "num_tokens": 12815672.0,
      "step": 5830
    },
    {
      "entropy": 5.738226652145386,
      "epoch": 0.5790988487495038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973974401378859,
      "loss": 5.579,
      "mean_token_accuracy": 0.14301157742738724,
      "num_tokens": 12826113.0,
      "step": 5835
    },
    {
      "entropy": 5.558589267730713,
      "epoch": 0.5795950774116713,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973920639196329,
      "loss": 5.5078,
      "mean_token_accuracy": 0.14471369311213494,
      "num_tokens": 12835941.0,
      "step": 5840
    },
    {
      "entropy": 5.6044713973999025,
      "epoch": 0.5800913060738389,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973866821865358,
      "loss": 5.5725,
      "mean_token_accuracy": 0.13366698920726777,
      "num_tokens": 12846186.0,
      "step": 5845
    },
    {
      "entropy": 5.69937424659729,
      "epoch": 0.5805875347360063,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973812949387283,
      "loss": 5.5559,
      "mean_token_accuracy": 0.14240086674690247,
      "num_tokens": 12856380.0,
      "step": 5850
    },
    {
      "entropy": 5.54889440536499,
      "epoch": 0.5810837633981739,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004973759021763439,
      "loss": 5.5947,
      "mean_token_accuracy": 0.13854385763406754,
      "num_tokens": 12865999.0,
      "step": 5855
    },
    {
      "entropy": 5.616566944122314,
      "epoch": 0.5815799920603414,
      "grad_norm": 1.0,
      "learning_rate": 0.0004973705038995162,
      "loss": 5.4763,
      "mean_token_accuracy": 0.14360741078853606,
      "num_tokens": 12876388.0,
      "step": 5860
    },
    {
      "entropy": 5.634979963302612,
      "epoch": 0.582076220722509,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973651001083792,
      "loss": 5.4929,
      "mean_token_accuracy": 0.14220962598919867,
      "num_tokens": 12885862.0,
      "step": 5865
    },
    {
      "entropy": 5.571282958984375,
      "epoch": 0.5825724493846764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973596908030669,
      "loss": 5.561,
      "mean_token_accuracy": 0.13786889910697936,
      "num_tokens": 12897489.0,
      "step": 5870
    },
    {
      "entropy": 5.679914665222168,
      "epoch": 0.583068678046844,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004973542759837135,
      "loss": 5.5908,
      "mean_token_accuracy": 0.13352037519216536,
      "num_tokens": 12907395.0,
      "step": 5875
    },
    {
      "entropy": 5.6124382495880125,
      "epoch": 0.5835649067090115,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497348855650453,
      "loss": 5.532,
      "mean_token_accuracy": 0.14500339403748513,
      "num_tokens": 12919439.0,
      "step": 5880
    },
    {
      "entropy": 5.6644408226013185,
      "epoch": 0.584061135371179,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004973434298034201,
      "loss": 5.6109,
      "mean_token_accuracy": 0.138861233741045,
      "num_tokens": 12931892.0,
      "step": 5885
    },
    {
      "entropy": 5.736296224594116,
      "epoch": 0.5845573640333466,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004973379984427492,
      "loss": 5.6699,
      "mean_token_accuracy": 0.1363135479390621,
      "num_tokens": 12943189.0,
      "step": 5890
    },
    {
      "entropy": 5.682193660736084,
      "epoch": 0.585053592695514,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497332561568575,
      "loss": 5.6078,
      "mean_token_accuracy": 0.13762683048844337,
      "num_tokens": 12954662.0,
      "step": 5895
    },
    {
      "entropy": 5.6221672058105465,
      "epoch": 0.5855498213576816,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973271191810325,
      "loss": 5.5561,
      "mean_token_accuracy": 0.14156940579414368,
      "num_tokens": 12964924.0,
      "step": 5900
    },
    {
      "entropy": 5.638089656829834,
      "epoch": 0.5860460500198491,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004973216712802564,
      "loss": 5.6244,
      "mean_token_accuracy": 0.13616342917084695,
      "num_tokens": 12976483.0,
      "step": 5905
    },
    {
      "entropy": 5.591444826126098,
      "epoch": 0.5865422786820167,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497316217866382,
      "loss": 5.537,
      "mean_token_accuracy": 0.1453517943620682,
      "num_tokens": 12988054.0,
      "step": 5910
    },
    {
      "entropy": 5.677403211593628,
      "epoch": 0.5870385073441842,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973107589395444,
      "loss": 5.5777,
      "mean_token_accuracy": 0.14270464181900025,
      "num_tokens": 13000596.0,
      "step": 5915
    },
    {
      "entropy": 5.562046909332276,
      "epoch": 0.5875347360063518,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973052944998791,
      "loss": 5.4187,
      "mean_token_accuracy": 0.14594281539320947,
      "num_tokens": 13012360.0,
      "step": 5920
    },
    {
      "entropy": 5.601492261886596,
      "epoch": 0.5880309646685192,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004972998245475214,
      "loss": 5.576,
      "mean_token_accuracy": 0.13968132436275482,
      "num_tokens": 13023806.0,
      "step": 5925
    },
    {
      "entropy": 5.61679368019104,
      "epoch": 0.5885271933306868,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000497294349082607,
      "loss": 5.5009,
      "mean_token_accuracy": 0.14438683837652205,
      "num_tokens": 13034049.0,
      "step": 5930
    },
    {
      "entropy": 5.655051136016846,
      "epoch": 0.5890234219928543,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004972888681052719,
      "loss": 5.5951,
      "mean_token_accuracy": 0.1382374234497547,
      "num_tokens": 13045006.0,
      "step": 5935
    },
    {
      "entropy": 5.7110090255737305,
      "epoch": 0.5895196506550219,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972833816156518,
      "loss": 5.512,
      "mean_token_accuracy": 0.14079046621918678,
      "num_tokens": 13055337.0,
      "step": 5940
    },
    {
      "entropy": 5.549333000183106,
      "epoch": 0.5900158793171894,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004972778896138828,
      "loss": 5.506,
      "mean_token_accuracy": 0.1479167029261589,
      "num_tokens": 13065267.0,
      "step": 5945
    },
    {
      "entropy": 5.594224071502685,
      "epoch": 0.5905121079793569,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004972723921001011,
      "loss": 5.5886,
      "mean_token_accuracy": 0.13957096710801126,
      "num_tokens": 13075173.0,
      "step": 5950
    },
    {
      "entropy": 5.692317104339599,
      "epoch": 0.5910083366415244,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004972668890744431,
      "loss": 5.5502,
      "mean_token_accuracy": 0.13957106247544288,
      "num_tokens": 13086848.0,
      "step": 5955
    },
    {
      "entropy": 5.640107107162476,
      "epoch": 0.5915045653036919,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972613805370452,
      "loss": 5.5422,
      "mean_token_accuracy": 0.1432167962193489,
      "num_tokens": 13096865.0,
      "step": 5960
    },
    {
      "entropy": 5.574533891677857,
      "epoch": 0.5920007939658595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972558664880439,
      "loss": 5.5043,
      "mean_token_accuracy": 0.14752691388130187,
      "num_tokens": 13107907.0,
      "step": 5965
    },
    {
      "entropy": 5.533886098861695,
      "epoch": 0.592497022628027,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000497250346927576,
      "loss": 5.4469,
      "mean_token_accuracy": 0.14774303063750266,
      "num_tokens": 13118171.0,
      "step": 5970
    },
    {
      "entropy": 5.584347295761108,
      "epoch": 0.5929932512901945,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004972448218557785,
      "loss": 5.4152,
      "mean_token_accuracy": 0.1491492673754692,
      "num_tokens": 13127042.0,
      "step": 5975
    },
    {
      "entropy": 5.696653079986572,
      "epoch": 0.593489479952362,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972392912727882,
      "loss": 5.6837,
      "mean_token_accuracy": 0.14060692191123964,
      "num_tokens": 13137026.0,
      "step": 5980
    },
    {
      "entropy": 5.656273984909058,
      "epoch": 0.5939857086145296,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972337551787425,
      "loss": 5.5281,
      "mean_token_accuracy": 0.13935455083847045,
      "num_tokens": 13146947.0,
      "step": 5985
    },
    {
      "entropy": 5.662847995758057,
      "epoch": 0.5944819372766971,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972282135737785,
      "loss": 5.5781,
      "mean_token_accuracy": 0.14084555357694625,
      "num_tokens": 13158714.0,
      "step": 5990
    },
    {
      "entropy": 5.671388864517212,
      "epoch": 0.5949781659388647,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972226664580337,
      "loss": 5.5832,
      "mean_token_accuracy": 0.1438387528061867,
      "num_tokens": 13170156.0,
      "step": 5995
    },
    {
      "entropy": 5.593986845016479,
      "epoch": 0.5954743946010321,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004972171138316456,
      "loss": 5.4974,
      "mean_token_accuracy": 0.14140935316681863,
      "num_tokens": 13179919.0,
      "step": 6000
    },
    {
      "epoch": 0.5954743946010321,
      "eval_entropy": 5.511412963133069,
      "eval_loss": 5.631018161773682,
      "eval_mean_token_accuracy": 0.14324472136627356,
      "eval_num_tokens": 13179919.0,
      "eval_runtime": 26.7243,
      "eval_samples_per_second": 1259.863,
      "eval_steps_per_second": 157.497,
      "step": 6000
    },
    {
      "entropy": 5.617749547958374,
      "epoch": 0.5959706232631997,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972115556947518,
      "loss": 5.577,
      "mean_token_accuracy": 0.14357811138033866,
      "num_tokens": 13190777.0,
      "step": 6005
    },
    {
      "entropy": 5.689182901382447,
      "epoch": 0.5964668519253672,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972059920474904,
      "loss": 5.5613,
      "mean_token_accuracy": 0.1421850137412548,
      "num_tokens": 13202179.0,
      "step": 6010
    },
    {
      "entropy": 5.6273339748382565,
      "epoch": 0.5969630805875348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004972004228899991,
      "loss": 5.5154,
      "mean_token_accuracy": 0.1430058442056179,
      "num_tokens": 13212764.0,
      "step": 6015
    },
    {
      "entropy": 5.596985912322998,
      "epoch": 0.5974593092497023,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004971948482224162,
      "loss": 5.5546,
      "mean_token_accuracy": 0.14096368551254274,
      "num_tokens": 13223343.0,
      "step": 6020
    },
    {
      "entropy": 5.663975954055786,
      "epoch": 0.5979555379118698,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004971892680448798,
      "loss": 5.529,
      "mean_token_accuracy": 0.1436027392745018,
      "num_tokens": 13233839.0,
      "step": 6025
    },
    {
      "entropy": 5.6838490009307865,
      "epoch": 0.5984517665740373,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004971836823575283,
      "loss": 5.634,
      "mean_token_accuracy": 0.14012941718101501,
      "num_tokens": 13243854.0,
      "step": 6030
    },
    {
      "entropy": 5.656049728393555,
      "epoch": 0.5989479952362048,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004971780911605003,
      "loss": 5.5507,
      "mean_token_accuracy": 0.13965711668133735,
      "num_tokens": 13256073.0,
      "step": 6035
    },
    {
      "entropy": 5.647913217544556,
      "epoch": 0.5994442238983724,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971724944539343,
      "loss": 5.5874,
      "mean_token_accuracy": 0.13906916305422784,
      "num_tokens": 13267568.0,
      "step": 6040
    },
    {
      "entropy": 5.6795509338378904,
      "epoch": 0.5999404525605398,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004971668922379692,
      "loss": 5.561,
      "mean_token_accuracy": 0.13741008937358856,
      "num_tokens": 13278829.0,
      "step": 6045
    },
    {
      "entropy": 5.624333763122559,
      "epoch": 0.6004366812227074,
      "grad_norm": 1.078125,
      "learning_rate": 0.000497161284512744,
      "loss": 5.5367,
      "mean_token_accuracy": 0.14608361721038818,
      "num_tokens": 13289156.0,
      "step": 6050
    },
    {
      "entropy": 5.511394929885864,
      "epoch": 0.6009329098848749,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004971556712783976,
      "loss": 5.4094,
      "mean_token_accuracy": 0.15416535288095473,
      "num_tokens": 13299817.0,
      "step": 6055
    },
    {
      "entropy": 5.544901275634766,
      "epoch": 0.6014291385470425,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004971500525350692,
      "loss": 5.4605,
      "mean_token_accuracy": 0.14854633510112764,
      "num_tokens": 13310290.0,
      "step": 6060
    },
    {
      "entropy": 5.572713613510132,
      "epoch": 0.60192536720921,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004971444282828983,
      "loss": 5.5869,
      "mean_token_accuracy": 0.1431807592511177,
      "num_tokens": 13321747.0,
      "step": 6065
    },
    {
      "entropy": 5.713278722763062,
      "epoch": 0.6024215958713776,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004971387985220242,
      "loss": 5.5261,
      "mean_token_accuracy": 0.1459825560450554,
      "num_tokens": 13333249.0,
      "step": 6070
    },
    {
      "entropy": 5.526351451873779,
      "epoch": 0.602917824533545,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004971331632525866,
      "loss": 5.5026,
      "mean_token_accuracy": 0.14363699331879615,
      "num_tokens": 13345336.0,
      "step": 6075
    },
    {
      "entropy": 5.720795011520385,
      "epoch": 0.6034140531957126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497127522474725,
      "loss": 5.6493,
      "mean_token_accuracy": 0.1350759483873844,
      "num_tokens": 13355797.0,
      "step": 6080
    },
    {
      "entropy": 5.670672416687012,
      "epoch": 0.6039102818578801,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971218761885797,
      "loss": 5.5991,
      "mean_token_accuracy": 0.14188750311732293,
      "num_tokens": 13367058.0,
      "step": 6085
    },
    {
      "entropy": 5.602615547180176,
      "epoch": 0.6044065105200477,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004971162243942904,
      "loss": 5.5403,
      "mean_token_accuracy": 0.14086537510156633,
      "num_tokens": 13378579.0,
      "step": 6090
    },
    {
      "entropy": 5.6347862720489506,
      "epoch": 0.6049027391822152,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004971105670919975,
      "loss": 5.5244,
      "mean_token_accuracy": 0.14428060352802277,
      "num_tokens": 13389240.0,
      "step": 6095
    },
    {
      "entropy": 5.701527214050293,
      "epoch": 0.6053989678443827,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497104904281841,
      "loss": 5.6604,
      "mean_token_accuracy": 0.13817641511559486,
      "num_tokens": 13401895.0,
      "step": 6100
    },
    {
      "entropy": 5.722287559509278,
      "epoch": 0.6058951965065502,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004970992359639616,
      "loss": 5.6511,
      "mean_token_accuracy": 0.13274826034903525,
      "num_tokens": 13413334.0,
      "step": 6105
    },
    {
      "entropy": 5.644446325302124,
      "epoch": 0.6063914251687177,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004970935621384996,
      "loss": 5.5471,
      "mean_token_accuracy": 0.14178750440478324,
      "num_tokens": 13423371.0,
      "step": 6110
    },
    {
      "entropy": 5.668351078033448,
      "epoch": 0.6068876538308853,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004970878828055959,
      "loss": 5.5351,
      "mean_token_accuracy": 0.14577889665961266,
      "num_tokens": 13434782.0,
      "step": 6115
    },
    {
      "entropy": 5.61508002281189,
      "epoch": 0.6073838824930528,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004970821979653912,
      "loss": 5.5119,
      "mean_token_accuracy": 0.14235954582691193,
      "num_tokens": 13444493.0,
      "step": 6120
    },
    {
      "entropy": 5.611604452133179,
      "epoch": 0.6078801111552203,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004970765076180265,
      "loss": 5.5039,
      "mean_token_accuracy": 0.14342557340860368,
      "num_tokens": 13455518.0,
      "step": 6125
    },
    {
      "entropy": 5.679499340057373,
      "epoch": 0.6083763398173878,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004970708117636431,
      "loss": 5.606,
      "mean_token_accuracy": 0.14624184966087342,
      "num_tokens": 13467631.0,
      "step": 6130
    },
    {
      "entropy": 5.578211927413941,
      "epoch": 0.6088725684795554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497065110402382,
      "loss": 5.5381,
      "mean_token_accuracy": 0.14146184399724007,
      "num_tokens": 13478217.0,
      "step": 6135
    },
    {
      "entropy": 5.654413175582886,
      "epoch": 0.6093687971417229,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970594035343846,
      "loss": 5.499,
      "mean_token_accuracy": 0.14559899866580964,
      "num_tokens": 13489169.0,
      "step": 6140
    },
    {
      "entropy": 5.629949617385864,
      "epoch": 0.6098650258038905,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970536911597926,
      "loss": 5.5276,
      "mean_token_accuracy": 0.13768994808197021,
      "num_tokens": 13499273.0,
      "step": 6145
    },
    {
      "entropy": 5.67266001701355,
      "epoch": 0.6103612544660579,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970479732787476,
      "loss": 5.5573,
      "mean_token_accuracy": 0.14647336602210997,
      "num_tokens": 13509664.0,
      "step": 6150
    },
    {
      "entropy": 5.592566394805909,
      "epoch": 0.6108574831282255,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004970422498913911,
      "loss": 5.454,
      "mean_token_accuracy": 0.14096739664673805,
      "num_tokens": 13520288.0,
      "step": 6155
    },
    {
      "entropy": 5.596761322021484,
      "epoch": 0.611353711790393,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970365209978655,
      "loss": 5.5425,
      "mean_token_accuracy": 0.14209363758563995,
      "num_tokens": 13531167.0,
      "step": 6160
    },
    {
      "entropy": 5.618218326568604,
      "epoch": 0.6118499404525606,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970307865983124,
      "loss": 5.587,
      "mean_token_accuracy": 0.1421313278377056,
      "num_tokens": 13544109.0,
      "step": 6165
    },
    {
      "entropy": 5.628209972381592,
      "epoch": 0.6123461691147281,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004970250466928744,
      "loss": 5.5061,
      "mean_token_accuracy": 0.14507102891802787,
      "num_tokens": 13554935.0,
      "step": 6170
    },
    {
      "entropy": 5.659920883178711,
      "epoch": 0.6128423977768956,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004970193012816936,
      "loss": 5.5525,
      "mean_token_accuracy": 0.13894661515951157,
      "num_tokens": 13565959.0,
      "step": 6175
    },
    {
      "entropy": 5.5555932998657225,
      "epoch": 0.6133386264390631,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970135503649125,
      "loss": 5.5399,
      "mean_token_accuracy": 0.14264004975557326,
      "num_tokens": 13577883.0,
      "step": 6180
    },
    {
      "entropy": 5.603068208694458,
      "epoch": 0.6138348551012306,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004970077939426738,
      "loss": 5.5131,
      "mean_token_accuracy": 0.14211745262145997,
      "num_tokens": 13588597.0,
      "step": 6185
    },
    {
      "entropy": 5.581861543655395,
      "epoch": 0.6143310837633982,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970020320151202,
      "loss": 5.4515,
      "mean_token_accuracy": 0.1445506379008293,
      "num_tokens": 13598772.0,
      "step": 6190
    },
    {
      "entropy": 5.610093355178833,
      "epoch": 0.6148273124255657,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004969962645823946,
      "loss": 5.5432,
      "mean_token_accuracy": 0.14223300963640212,
      "num_tokens": 13609696.0,
      "step": 6195
    },
    {
      "entropy": 5.590139579772949,
      "epoch": 0.6153235410877332,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049699049164464,
      "loss": 5.5157,
      "mean_token_accuracy": 0.14494847655296325,
      "num_tokens": 13619389.0,
      "step": 6200
    },
    {
      "entropy": 5.613835859298706,
      "epoch": 0.6158197697499007,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969847132019995,
      "loss": 5.502,
      "mean_token_accuracy": 0.14289411902427673,
      "num_tokens": 13632133.0,
      "step": 6205
    },
    {
      "entropy": 5.604021263122559,
      "epoch": 0.6163159984120683,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969789292546164,
      "loss": 5.4636,
      "mean_token_accuracy": 0.1498689293861389,
      "num_tokens": 13642715.0,
      "step": 6210
    },
    {
      "entropy": 5.614312314987183,
      "epoch": 0.6168122270742358,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004969731398026343,
      "loss": 5.5582,
      "mean_token_accuracy": 0.14249989092350007,
      "num_tokens": 13653223.0,
      "step": 6215
    },
    {
      "entropy": 5.554715967178344,
      "epoch": 0.6173084557364034,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004969673448461966,
      "loss": 5.5415,
      "mean_token_accuracy": 0.14299920946359634,
      "num_tokens": 13664206.0,
      "step": 6220
    },
    {
      "entropy": 5.633339786529541,
      "epoch": 0.6178046843985708,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000496961544385447,
      "loss": 5.5598,
      "mean_token_accuracy": 0.1444960840046406,
      "num_tokens": 13673974.0,
      "step": 6225
    },
    {
      "entropy": 5.575097608566284,
      "epoch": 0.6183009130607384,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004969557384205295,
      "loss": 5.4709,
      "mean_token_accuracy": 0.1455431781709194,
      "num_tokens": 13685189.0,
      "step": 6230
    },
    {
      "entropy": 5.6610228538513185,
      "epoch": 0.6187971417229059,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004969499269515878,
      "loss": 5.4967,
      "mean_token_accuracy": 0.14609827399253844,
      "num_tokens": 13696181.0,
      "step": 6235
    },
    {
      "entropy": 5.684266853332519,
      "epoch": 0.6192933703850735,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004969441099787663,
      "loss": 5.5927,
      "mean_token_accuracy": 0.1392730802297592,
      "num_tokens": 13706719.0,
      "step": 6240
    },
    {
      "entropy": 5.664065837860107,
      "epoch": 0.619789599047241,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000496938287502209,
      "loss": 5.5756,
      "mean_token_accuracy": 0.1419723927974701,
      "num_tokens": 13716653.0,
      "step": 6245
    },
    {
      "entropy": 5.563004827499389,
      "epoch": 0.6202858277094085,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004969324595220605,
      "loss": 5.4726,
      "mean_token_accuracy": 0.14750099033117295,
      "num_tokens": 13728220.0,
      "step": 6250
    },
    {
      "entropy": 5.51563024520874,
      "epoch": 0.620782056371576,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004969266260384652,
      "loss": 5.4468,
      "mean_token_accuracy": 0.1471132755279541,
      "num_tokens": 13738103.0,
      "step": 6255
    },
    {
      "entropy": 5.598431777954102,
      "epoch": 0.6212782850337435,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004969207870515678,
      "loss": 5.5047,
      "mean_token_accuracy": 0.14320664554834367,
      "num_tokens": 13749845.0,
      "step": 6260
    },
    {
      "entropy": 5.5818838596344,
      "epoch": 0.6217745136959111,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000496914942561513,
      "loss": 5.5525,
      "mean_token_accuracy": 0.14325065910816193,
      "num_tokens": 13759983.0,
      "step": 6265
    },
    {
      "entropy": 5.676664972305298,
      "epoch": 0.6222707423580786,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969090925684459,
      "loss": 5.6394,
      "mean_token_accuracy": 0.13805949687957764,
      "num_tokens": 13770045.0,
      "step": 6270
    },
    {
      "entropy": 5.596723985671997,
      "epoch": 0.6227669710202461,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004969032370725114,
      "loss": 5.4883,
      "mean_token_accuracy": 0.14576412439346315,
      "num_tokens": 13779823.0,
      "step": 6275
    },
    {
      "entropy": 5.565663766860962,
      "epoch": 0.6232631996824136,
      "grad_norm": 1.0,
      "learning_rate": 0.0004968973760738548,
      "loss": 5.4772,
      "mean_token_accuracy": 0.14060586988925933,
      "num_tokens": 13791165.0,
      "step": 6280
    },
    {
      "entropy": 5.689731645584106,
      "epoch": 0.6237594283445812,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968915095726216,
      "loss": 5.5968,
      "mean_token_accuracy": 0.13638106882572174,
      "num_tokens": 13802872.0,
      "step": 6285
    },
    {
      "entropy": 5.6198687076568605,
      "epoch": 0.6242556570067487,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004968856375689568,
      "loss": 5.5273,
      "mean_token_accuracy": 0.14362960159778596,
      "num_tokens": 13813893.0,
      "step": 6290
    },
    {
      "entropy": 5.634898900985718,
      "epoch": 0.6247518856689163,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004968797600630065,
      "loss": 5.582,
      "mean_token_accuracy": 0.1434788689017296,
      "num_tokens": 13823613.0,
      "step": 6295
    },
    {
      "entropy": 5.69560604095459,
      "epoch": 0.6252481143310837,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968738770549162,
      "loss": 5.586,
      "mean_token_accuracy": 0.14546991884708405,
      "num_tokens": 13835413.0,
      "step": 6300
    },
    {
      "entropy": 5.606576251983642,
      "epoch": 0.6257443429932513,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004968679885448318,
      "loss": 5.5263,
      "mean_token_accuracy": 0.1430337108671665,
      "num_tokens": 13845806.0,
      "step": 6305
    },
    {
      "entropy": 5.662768268585205,
      "epoch": 0.6262405716554188,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004968620945328995,
      "loss": 5.6218,
      "mean_token_accuracy": 0.13542826250195503,
      "num_tokens": 13856540.0,
      "step": 6310
    },
    {
      "entropy": 5.639750671386719,
      "epoch": 0.6267368003175864,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968561950192652,
      "loss": 5.486,
      "mean_token_accuracy": 0.1482410080730915,
      "num_tokens": 13867609.0,
      "step": 6315
    },
    {
      "entropy": 5.636210203170776,
      "epoch": 0.6272330289797539,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004968502900040753,
      "loss": 5.5381,
      "mean_token_accuracy": 0.13960634022951127,
      "num_tokens": 13878299.0,
      "step": 6320
    },
    {
      "entropy": 5.610741329193115,
      "epoch": 0.6277292576419214,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968443794874764,
      "loss": 5.5296,
      "mean_token_accuracy": 0.14378902912139893,
      "num_tokens": 13889625.0,
      "step": 6325
    },
    {
      "entropy": 5.584087276458741,
      "epoch": 0.6282254863040889,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004968384634696149,
      "loss": 5.5096,
      "mean_token_accuracy": 0.13866836875677108,
      "num_tokens": 13899680.0,
      "step": 6330
    },
    {
      "entropy": 5.551987266540527,
      "epoch": 0.6287217149662565,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968325419506375,
      "loss": 5.5031,
      "mean_token_accuracy": 0.14664536938071251,
      "num_tokens": 13910883.0,
      "step": 6335
    },
    {
      "entropy": 5.601029396057129,
      "epoch": 0.629217943628424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004968266149306909,
      "loss": 5.5219,
      "mean_token_accuracy": 0.14437941014766692,
      "num_tokens": 13921442.0,
      "step": 6340
    },
    {
      "entropy": 5.620112514495849,
      "epoch": 0.6297141722905915,
      "grad_norm": 1.0,
      "learning_rate": 0.0004968206824099224,
      "loss": 5.5479,
      "mean_token_accuracy": 0.1442171186208725,
      "num_tokens": 13930309.0,
      "step": 6345
    },
    {
      "entropy": 5.642942094802857,
      "epoch": 0.630210400952759,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004968147443884789,
      "loss": 5.5282,
      "mean_token_accuracy": 0.1514594227075577,
      "num_tokens": 13941346.0,
      "step": 6350
    },
    {
      "entropy": 5.577560377120972,
      "epoch": 0.6307066296149265,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004968088008665078,
      "loss": 5.5107,
      "mean_token_accuracy": 0.1457039326429367,
      "num_tokens": 13951933.0,
      "step": 6355
    },
    {
      "entropy": 5.576596117019653,
      "epoch": 0.6312028582770941,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004968028518441563,
      "loss": 5.5692,
      "mean_token_accuracy": 0.13854328766465188,
      "num_tokens": 13963112.0,
      "step": 6360
    },
    {
      "entropy": 5.717953062057495,
      "epoch": 0.6316990869392616,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004967968973215721,
      "loss": 5.5942,
      "mean_token_accuracy": 0.14378464072942734,
      "num_tokens": 13974218.0,
      "step": 6365
    },
    {
      "entropy": 5.644935369491577,
      "epoch": 0.6321953156014292,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004967909372989027,
      "loss": 5.5243,
      "mean_token_accuracy": 0.1479162111878395,
      "num_tokens": 13984841.0,
      "step": 6370
    },
    {
      "entropy": 5.564591789245606,
      "epoch": 0.6326915442635966,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004967849717762959,
      "loss": 5.5281,
      "mean_token_accuracy": 0.144460928440094,
      "num_tokens": 13997504.0,
      "step": 6375
    },
    {
      "entropy": 5.582908678054809,
      "epoch": 0.6331877729257642,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004967790007538999,
      "loss": 5.4724,
      "mean_token_accuracy": 0.1487226143479347,
      "num_tokens": 14009244.0,
      "step": 6380
    },
    {
      "entropy": 5.575730466842652,
      "epoch": 0.6336840015879317,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004967730242318624,
      "loss": 5.4902,
      "mean_token_accuracy": 0.14656757712364196,
      "num_tokens": 14021190.0,
      "step": 6385
    },
    {
      "entropy": 5.546020221710205,
      "epoch": 0.6341802302500993,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000496767042210332,
      "loss": 5.4622,
      "mean_token_accuracy": 0.14225331023335458,
      "num_tokens": 14032191.0,
      "step": 6390
    },
    {
      "entropy": 5.506117486953736,
      "epoch": 0.6346764589122668,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004967610546894566,
      "loss": 5.4428,
      "mean_token_accuracy": 0.15072381645441055,
      "num_tokens": 14042817.0,
      "step": 6395
    },
    {
      "entropy": 5.561826610565186,
      "epoch": 0.6351726875744343,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000496755061669385,
      "loss": 5.4516,
      "mean_token_accuracy": 0.15164192169904708,
      "num_tokens": 14052725.0,
      "step": 6400
    },
    {
      "entropy": 5.633494424819946,
      "epoch": 0.6356689162366018,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004967490631502656,
      "loss": 5.5252,
      "mean_token_accuracy": 0.13410304263234138,
      "num_tokens": 14063621.0,
      "step": 6405
    },
    {
      "entropy": 5.731800985336304,
      "epoch": 0.6361651448987694,
      "grad_norm": 1.0,
      "learning_rate": 0.0004967430591322474,
      "loss": 5.652,
      "mean_token_accuracy": 0.1400383397936821,
      "num_tokens": 14074350.0,
      "step": 6410
    },
    {
      "entropy": 5.664568090438843,
      "epoch": 0.6366613735609369,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004967370496154791,
      "loss": 5.5803,
      "mean_token_accuracy": 0.1434120237827301,
      "num_tokens": 14084600.0,
      "step": 6415
    },
    {
      "entropy": 5.651761722564697,
      "epoch": 0.6371576022231044,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004967310346001098,
      "loss": 5.4962,
      "mean_token_accuracy": 0.1455589085817337,
      "num_tokens": 14094739.0,
      "step": 6420
    },
    {
      "entropy": 5.707870054244995,
      "epoch": 0.6376538308852719,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004967250140862885,
      "loss": 5.6757,
      "mean_token_accuracy": 0.13760743364691735,
      "num_tokens": 14106020.0,
      "step": 6425
    },
    {
      "entropy": 5.71898136138916,
      "epoch": 0.6381500595474394,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004967189880741647,
      "loss": 5.6487,
      "mean_token_accuracy": 0.13639736026525498,
      "num_tokens": 14117699.0,
      "step": 6430
    },
    {
      "entropy": 5.61235032081604,
      "epoch": 0.638646288209607,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004967129565638878,
      "loss": 5.5105,
      "mean_token_accuracy": 0.14647904634475709,
      "num_tokens": 14128901.0,
      "step": 6435
    },
    {
      "entropy": 5.5687188625335695,
      "epoch": 0.6391425168717745,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004967069195556073,
      "loss": 5.5505,
      "mean_token_accuracy": 0.14266734719276428,
      "num_tokens": 14140832.0,
      "step": 6440
    },
    {
      "entropy": 5.635690498352051,
      "epoch": 0.6396387455339421,
      "grad_norm": 1.078125,
      "learning_rate": 0.000496700877049473,
      "loss": 5.5751,
      "mean_token_accuracy": 0.1452000133693218,
      "num_tokens": 14151936.0,
      "step": 6445
    },
    {
      "entropy": 5.673252534866333,
      "epoch": 0.6401349741961095,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966948290456345,
      "loss": 5.5823,
      "mean_token_accuracy": 0.14218578189611436,
      "num_tokens": 14160945.0,
      "step": 6450
    },
    {
      "entropy": 5.569803762435913,
      "epoch": 0.6406312028582771,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004966887755442421,
      "loss": 5.4853,
      "mean_token_accuracy": 0.14295473396778108,
      "num_tokens": 14172052.0,
      "step": 6455
    },
    {
      "entropy": 5.641026020050049,
      "epoch": 0.6411274315204446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004966827165454456,
      "loss": 5.5444,
      "mean_token_accuracy": 0.14384291544556618,
      "num_tokens": 14181788.0,
      "step": 6460
    },
    {
      "entropy": 5.727074527740479,
      "epoch": 0.6416236601826122,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004966766520493956,
      "loss": 5.6324,
      "mean_token_accuracy": 0.14106333404779434,
      "num_tokens": 14193668.0,
      "step": 6465
    },
    {
      "entropy": 5.673287296295166,
      "epoch": 0.6421198888447797,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004966705820562422,
      "loss": 5.4976,
      "mean_token_accuracy": 0.14693205803632736,
      "num_tokens": 14203665.0,
      "step": 6470
    },
    {
      "entropy": 5.545781660079956,
      "epoch": 0.6426161175069472,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966645065661361,
      "loss": 5.5505,
      "mean_token_accuracy": 0.13993965536355973,
      "num_tokens": 14214213.0,
      "step": 6475
    },
    {
      "entropy": 5.637283897399902,
      "epoch": 0.6431123461691147,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004966584255792279,
      "loss": 5.6285,
      "mean_token_accuracy": 0.14050951600074768,
      "num_tokens": 14225754.0,
      "step": 6480
    },
    {
      "entropy": 5.657700634002685,
      "epoch": 0.6436085748312823,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966523390956683,
      "loss": 5.5243,
      "mean_token_accuracy": 0.1451603278517723,
      "num_tokens": 14236508.0,
      "step": 6485
    },
    {
      "entropy": 5.707139730453491,
      "epoch": 0.6441048034934498,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966462471156083,
      "loss": 5.5574,
      "mean_token_accuracy": 0.1475798264145851,
      "num_tokens": 14247719.0,
      "step": 6490
    },
    {
      "entropy": 5.656018304824829,
      "epoch": 0.6446010321556173,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966401496391989,
      "loss": 5.6087,
      "mean_token_accuracy": 0.13209823817014693,
      "num_tokens": 14258667.0,
      "step": 6495
    },
    {
      "entropy": 5.595034408569336,
      "epoch": 0.6450972608177848,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004966340466665915,
      "loss": 5.6074,
      "mean_token_accuracy": 0.14210932180285454,
      "num_tokens": 14269666.0,
      "step": 6500
    },
    {
      "entropy": 5.575317525863648,
      "epoch": 0.6455934894799523,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004966279381979373,
      "loss": 5.5205,
      "mean_token_accuracy": 0.14046879336237908,
      "num_tokens": 14280862.0,
      "step": 6505
    },
    {
      "entropy": 5.697528982162476,
      "epoch": 0.6460897181421199,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004966218242333879,
      "loss": 5.6582,
      "mean_token_accuracy": 0.13086239099502564,
      "num_tokens": 14292505.0,
      "step": 6510
    },
    {
      "entropy": 5.666721248626709,
      "epoch": 0.6465859468042874,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004966157047730945,
      "loss": 5.4578,
      "mean_token_accuracy": 0.1494349032640457,
      "num_tokens": 14302547.0,
      "step": 6515
    },
    {
      "entropy": 5.59673810005188,
      "epoch": 0.647082175466455,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004966095798172095,
      "loss": 5.5955,
      "mean_token_accuracy": 0.14004509821534156,
      "num_tokens": 14313585.0,
      "step": 6520
    },
    {
      "entropy": 5.579741811752319,
      "epoch": 0.6475784041286224,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004966034493658843,
      "loss": 5.4695,
      "mean_token_accuracy": 0.13805497884750367,
      "num_tokens": 14323665.0,
      "step": 6525
    },
    {
      "entropy": 5.61812310218811,
      "epoch": 0.64807463279079,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004965973134192711,
      "loss": 5.49,
      "mean_token_accuracy": 0.14620914235711097,
      "num_tokens": 14334257.0,
      "step": 6530
    },
    {
      "entropy": 5.652993440628052,
      "epoch": 0.6485708614529575,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000496591171977522,
      "loss": 5.5177,
      "mean_token_accuracy": 0.14237041026353836,
      "num_tokens": 14345350.0,
      "step": 6535
    },
    {
      "entropy": 5.659733533859253,
      "epoch": 0.6490670901151251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965850250407893,
      "loss": 5.5352,
      "mean_token_accuracy": 0.14441534951329232,
      "num_tokens": 14355593.0,
      "step": 6540
    },
    {
      "entropy": 5.564375162124634,
      "epoch": 0.6495633187772926,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004965788726092256,
      "loss": 5.4171,
      "mean_token_accuracy": 0.14944821298122407,
      "num_tokens": 14366405.0,
      "step": 6545
    },
    {
      "entropy": 5.552011728286743,
      "epoch": 0.6500595474394602,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004965727146829832,
      "loss": 5.5082,
      "mean_token_accuracy": 0.14282310679554938,
      "num_tokens": 14377228.0,
      "step": 6550
    },
    {
      "entropy": 5.731308555603027,
      "epoch": 0.6505557761016276,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004965665512622148,
      "loss": 5.6496,
      "mean_token_accuracy": 0.13893002048134803,
      "num_tokens": 14387113.0,
      "step": 6555
    },
    {
      "entropy": 5.6539770603179935,
      "epoch": 0.6510520047637952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004965603823470735,
      "loss": 5.5216,
      "mean_token_accuracy": 0.14390210211277008,
      "num_tokens": 14397980.0,
      "step": 6560
    },
    {
      "entropy": 5.571946239471435,
      "epoch": 0.6515482334259627,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004965542079377122,
      "loss": 5.4395,
      "mean_token_accuracy": 0.1514829620718956,
      "num_tokens": 14409063.0,
      "step": 6565
    },
    {
      "entropy": 5.5214122295379635,
      "epoch": 0.6520444620881302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004965480280342839,
      "loss": 5.4252,
      "mean_token_accuracy": 0.1550879493355751,
      "num_tokens": 14421056.0,
      "step": 6570
    },
    {
      "entropy": 5.573328733444214,
      "epoch": 0.6525406907502977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004965418426369419,
      "loss": 5.415,
      "mean_token_accuracy": 0.15354251787066459,
      "num_tokens": 14430778.0,
      "step": 6575
    },
    {
      "entropy": 5.567942762374878,
      "epoch": 0.6530369194124652,
      "grad_norm": 1.0,
      "learning_rate": 0.0004965356517458395,
      "loss": 5.4393,
      "mean_token_accuracy": 0.15072097182273864,
      "num_tokens": 14441544.0,
      "step": 6580
    },
    {
      "entropy": 5.580814933776855,
      "epoch": 0.6535331480746328,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004965294553611303,
      "loss": 5.6063,
      "mean_token_accuracy": 0.13923998326063156,
      "num_tokens": 14453062.0,
      "step": 6585
    },
    {
      "entropy": 5.6467822074890135,
      "epoch": 0.6540293767368003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496523253482968,
      "loss": 5.5057,
      "mean_token_accuracy": 0.1438075214624405,
      "num_tokens": 14464402.0,
      "step": 6590
    },
    {
      "entropy": 5.6507617950439455,
      "epoch": 0.6545256053989679,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004965170461115065,
      "loss": 5.6078,
      "mean_token_accuracy": 0.14371883869171143,
      "num_tokens": 14476595.0,
      "step": 6595
    },
    {
      "entropy": 5.695961856842041,
      "epoch": 0.6550218340611353,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004965108332468994,
      "loss": 5.5966,
      "mean_token_accuracy": 0.14018445014953612,
      "num_tokens": 14487298.0,
      "step": 6600
    },
    {
      "entropy": 5.626483964920044,
      "epoch": 0.6555180627233029,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004965046148893011,
      "loss": 5.4926,
      "mean_token_accuracy": 0.1490776374936104,
      "num_tokens": 14499708.0,
      "step": 6605
    },
    {
      "entropy": 5.602768516540527,
      "epoch": 0.6560142913854704,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004964983910388654,
      "loss": 5.5404,
      "mean_token_accuracy": 0.14275878965854644,
      "num_tokens": 14511773.0,
      "step": 6610
    },
    {
      "entropy": 5.5468052387237545,
      "epoch": 0.656510520047638,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004964921616957471,
      "loss": 5.4507,
      "mean_token_accuracy": 0.14010079652071,
      "num_tokens": 14523024.0,
      "step": 6615
    },
    {
      "entropy": 5.600555467605591,
      "epoch": 0.6570067487098055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004964859268601005,
      "loss": 5.5158,
      "mean_token_accuracy": 0.14727514013648033,
      "num_tokens": 14532915.0,
      "step": 6620
    },
    {
      "entropy": 5.660110569000244,
      "epoch": 0.657502977371973,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004964796865320801,
      "loss": 5.5015,
      "mean_token_accuracy": 0.1462259352207184,
      "num_tokens": 14543627.0,
      "step": 6625
    },
    {
      "entropy": 5.529415416717529,
      "epoch": 0.6579992060341405,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004964734407118406,
      "loss": 5.4729,
      "mean_token_accuracy": 0.14952642172574998,
      "num_tokens": 14555181.0,
      "step": 6630
    },
    {
      "entropy": 5.591297817230225,
      "epoch": 0.6584954346963081,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004964671893995372,
      "loss": 5.4826,
      "mean_token_accuracy": 0.14319788217544555,
      "num_tokens": 14565542.0,
      "step": 6635
    },
    {
      "entropy": 5.651774787902832,
      "epoch": 0.6589916633584756,
      "grad_norm": 1.125,
      "learning_rate": 0.0004964609325953245,
      "loss": 5.5318,
      "mean_token_accuracy": 0.14688521325588227,
      "num_tokens": 14575395.0,
      "step": 6640
    },
    {
      "entropy": 5.646538591384887,
      "epoch": 0.6594878920206431,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004964546702993581,
      "loss": 5.5734,
      "mean_token_accuracy": 0.1409864418208599,
      "num_tokens": 14587206.0,
      "step": 6645
    },
    {
      "entropy": 5.614670085906982,
      "epoch": 0.6599841206828106,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000496448402511793,
      "loss": 5.5849,
      "mean_token_accuracy": 0.14147609621286392,
      "num_tokens": 14599048.0,
      "step": 6650
    },
    {
      "entropy": 5.705883121490478,
      "epoch": 0.6604803493449781,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004964421292327847,
      "loss": 5.5409,
      "mean_token_accuracy": 0.14338165149092674,
      "num_tokens": 14609633.0,
      "step": 6655
    },
    {
      "entropy": 5.5681374073028564,
      "epoch": 0.6609765780071457,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004964358504624887,
      "loss": 5.4479,
      "mean_token_accuracy": 0.14635639190673827,
      "num_tokens": 14620373.0,
      "step": 6660
    },
    {
      "entropy": 5.572950983047486,
      "epoch": 0.6614728066693132,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004964295662010608,
      "loss": 5.383,
      "mean_token_accuracy": 0.1540623798966408,
      "num_tokens": 14630463.0,
      "step": 6665
    },
    {
      "entropy": 5.713060665130615,
      "epoch": 0.6619690353314808,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004964232764486568,
      "loss": 5.6363,
      "mean_token_accuracy": 0.13606127724051476,
      "num_tokens": 14641523.0,
      "step": 6670
    },
    {
      "entropy": 5.563221883773804,
      "epoch": 0.6624652639936482,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004964169812054326,
      "loss": 5.4533,
      "mean_token_accuracy": 0.1445788249373436,
      "num_tokens": 14651812.0,
      "step": 6675
    },
    {
      "entropy": 5.582422399520874,
      "epoch": 0.6629614926558158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004964106804715444,
      "loss": 5.4611,
      "mean_token_accuracy": 0.15091832280158995,
      "num_tokens": 14661982.0,
      "step": 6680
    },
    {
      "entropy": 5.548552227020264,
      "epoch": 0.6634577213179833,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004964043742471486,
      "loss": 5.4737,
      "mean_token_accuracy": 0.15126288831233978,
      "num_tokens": 14671981.0,
      "step": 6685
    },
    {
      "entropy": 5.577681398391723,
      "epoch": 0.6639539499801509,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004963980625324012,
      "loss": 5.4928,
      "mean_token_accuracy": 0.1391047164797783,
      "num_tokens": 14682478.0,
      "step": 6690
    },
    {
      "entropy": 5.637539911270141,
      "epoch": 0.6644501786423184,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496391745327459,
      "loss": 5.5652,
      "mean_token_accuracy": 0.13711804449558257,
      "num_tokens": 14695197.0,
      "step": 6695
    },
    {
      "entropy": 5.5980311870574955,
      "epoch": 0.664946407304486,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963854226324785,
      "loss": 5.4865,
      "mean_token_accuracy": 0.14268676713109016,
      "num_tokens": 14705702.0,
      "step": 6700
    },
    {
      "entropy": 5.591587066650391,
      "epoch": 0.6654426359666534,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963790944476167,
      "loss": 5.4876,
      "mean_token_accuracy": 0.1489054873585701,
      "num_tokens": 14715370.0,
      "step": 6705
    },
    {
      "entropy": 5.634316682815552,
      "epoch": 0.665938864628821,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004963727607730303,
      "loss": 5.5133,
      "mean_token_accuracy": 0.14755212515592575,
      "num_tokens": 14725481.0,
      "step": 6710
    },
    {
      "entropy": 5.56249508857727,
      "epoch": 0.6664350932909885,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963664216088764,
      "loss": 5.4339,
      "mean_token_accuracy": 0.1443425089120865,
      "num_tokens": 14735250.0,
      "step": 6715
    },
    {
      "entropy": 5.559953498840332,
      "epoch": 0.666931321953156,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004963600769553124,
      "loss": 5.4509,
      "mean_token_accuracy": 0.1510240003466606,
      "num_tokens": 14744741.0,
      "step": 6720
    },
    {
      "entropy": 5.63917293548584,
      "epoch": 0.6674275506153236,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004963537268124953,
      "loss": 5.4844,
      "mean_token_accuracy": 0.1444581426680088,
      "num_tokens": 14755865.0,
      "step": 6725
    },
    {
      "entropy": 5.509986686706543,
      "epoch": 0.667923779277491,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004963473711805829,
      "loss": 5.4376,
      "mean_token_accuracy": 0.14917942732572556,
      "num_tokens": 14766154.0,
      "step": 6730
    },
    {
      "entropy": 5.575598049163818,
      "epoch": 0.6684200079396586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004963410100597326,
      "loss": 5.4714,
      "mean_token_accuracy": 0.14657177031040192,
      "num_tokens": 14776940.0,
      "step": 6735
    },
    {
      "entropy": 5.566277170181275,
      "epoch": 0.6689162366018261,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004963346434501021,
      "loss": 5.5909,
      "mean_token_accuracy": 0.14163725599646568,
      "num_tokens": 14788367.0,
      "step": 6740
    },
    {
      "entropy": 5.626369524002075,
      "epoch": 0.6694124652639937,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004963282713518495,
      "loss": 5.5053,
      "mean_token_accuracy": 0.1468518152832985,
      "num_tokens": 14798751.0,
      "step": 6745
    },
    {
      "entropy": 5.531801795959472,
      "epoch": 0.6699086939261611,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004963218937651326,
      "loss": 5.4357,
      "mean_token_accuracy": 0.14639147222042084,
      "num_tokens": 14808748.0,
      "step": 6750
    },
    {
      "entropy": 5.633822202682495,
      "epoch": 0.6704049225883287,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004963155106901098,
      "loss": 5.5947,
      "mean_token_accuracy": 0.13615550994873046,
      "num_tokens": 14820524.0,
      "step": 6755
    },
    {
      "entropy": 5.714738845825195,
      "epoch": 0.6709011512504962,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004963091221269391,
      "loss": 5.6253,
      "mean_token_accuracy": 0.13395797833800316,
      "num_tokens": 14832485.0,
      "step": 6760
    },
    {
      "entropy": 5.5704278469085695,
      "epoch": 0.6713973799126638,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496302728075779,
      "loss": 5.4436,
      "mean_token_accuracy": 0.14065870195627211,
      "num_tokens": 14843951.0,
      "step": 6765
    },
    {
      "entropy": 5.5867705821990965,
      "epoch": 0.6718936085748313,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004962963285367882,
      "loss": 5.5361,
      "mean_token_accuracy": 0.14580765068531037,
      "num_tokens": 14855502.0,
      "step": 6770
    },
    {
      "entropy": 5.5551587581634525,
      "epoch": 0.6723898372369989,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962899235101253,
      "loss": 5.4446,
      "mean_token_accuracy": 0.15309773683547973,
      "num_tokens": 14866305.0,
      "step": 6775
    },
    {
      "entropy": 5.589531517028808,
      "epoch": 0.6728860658991663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962835129959491,
      "loss": 5.4612,
      "mean_token_accuracy": 0.1452888086438179,
      "num_tokens": 14877222.0,
      "step": 6780
    },
    {
      "entropy": 5.551893663406372,
      "epoch": 0.6733822945613339,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962770969944186,
      "loss": 5.5446,
      "mean_token_accuracy": 0.1419015571475029,
      "num_tokens": 14888099.0,
      "step": 6785
    },
    {
      "entropy": 5.593816041946411,
      "epoch": 0.6738785232235014,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496270675505693,
      "loss": 5.4728,
      "mean_token_accuracy": 0.1512457899749279,
      "num_tokens": 14899042.0,
      "step": 6790
    },
    {
      "entropy": 5.6144002914428714,
      "epoch": 0.6743747518856689,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004962642485299314,
      "loss": 5.5286,
      "mean_token_accuracy": 0.13888545632362365,
      "num_tokens": 14908837.0,
      "step": 6795
    },
    {
      "entropy": 5.595649099349975,
      "epoch": 0.6748709805478365,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004962578160672931,
      "loss": 5.4622,
      "mean_token_accuracy": 0.14449624419212342,
      "num_tokens": 14920773.0,
      "step": 6800
    },
    {
      "entropy": 5.580962610244751,
      "epoch": 0.6753672092100039,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004962513781179379,
      "loss": 5.4557,
      "mean_token_accuracy": 0.14727083146572112,
      "num_tokens": 14930970.0,
      "step": 6805
    },
    {
      "entropy": 5.595052814483642,
      "epoch": 0.6758634378721715,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962449346820251,
      "loss": 5.5097,
      "mean_token_accuracy": 0.13841721042990685,
      "num_tokens": 14942378.0,
      "step": 6810
    },
    {
      "entropy": 5.548154401779175,
      "epoch": 0.676359666534339,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004962384857597148,
      "loss": 5.5531,
      "mean_token_accuracy": 0.13887142166495323,
      "num_tokens": 14953020.0,
      "step": 6815
    },
    {
      "entropy": 5.68001389503479,
      "epoch": 0.6768558951965066,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962320313511668,
      "loss": 5.6222,
      "mean_token_accuracy": 0.14384716451168061,
      "num_tokens": 14962340.0,
      "step": 6820
    },
    {
      "entropy": 5.664641857147217,
      "epoch": 0.677352123858674,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004962255714565409,
      "loss": 5.4957,
      "mean_token_accuracy": 0.14904532730579376,
      "num_tokens": 14972792.0,
      "step": 6825
    },
    {
      "entropy": 5.5904662132263185,
      "epoch": 0.6778483525208416,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962191060759978,
      "loss": 5.484,
      "mean_token_accuracy": 0.14596888422966003,
      "num_tokens": 14983698.0,
      "step": 6830
    },
    {
      "entropy": 5.569444370269776,
      "epoch": 0.6783445811830091,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962126352096973,
      "loss": 5.4555,
      "mean_token_accuracy": 0.14673737734556197,
      "num_tokens": 14994583.0,
      "step": 6835
    },
    {
      "entropy": 5.536163473129273,
      "epoch": 0.6788408098451767,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004962061588578004,
      "loss": 5.4146,
      "mean_token_accuracy": 0.14648718684911727,
      "num_tokens": 15005325.0,
      "step": 6840
    },
    {
      "entropy": 5.625168895721435,
      "epoch": 0.6793370385073442,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004961996770204673,
      "loss": 5.474,
      "mean_token_accuracy": 0.1451435223221779,
      "num_tokens": 15015155.0,
      "step": 6845
    },
    {
      "entropy": 5.571862363815308,
      "epoch": 0.6798332671695118,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004961931896978588,
      "loss": 5.5464,
      "mean_token_accuracy": 0.14163526073098182,
      "num_tokens": 15026529.0,
      "step": 6850
    },
    {
      "entropy": 5.526344156265258,
      "epoch": 0.6803294958316792,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961866968901359,
      "loss": 5.4724,
      "mean_token_accuracy": 0.1455441638827324,
      "num_tokens": 15037632.0,
      "step": 6855
    },
    {
      "entropy": 5.588477230072021,
      "epoch": 0.6808257244938468,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004961801985974595,
      "loss": 5.5442,
      "mean_token_accuracy": 0.14573636651039124,
      "num_tokens": 15050397.0,
      "step": 6860
    },
    {
      "entropy": 5.587202835083008,
      "epoch": 0.6813219531560143,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961736948199907,
      "loss": 5.46,
      "mean_token_accuracy": 0.14760808348655702,
      "num_tokens": 15061505.0,
      "step": 6865
    },
    {
      "entropy": 5.705941772460937,
      "epoch": 0.6818181818181818,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961671855578909,
      "loss": 5.6454,
      "mean_token_accuracy": 0.1393208235502243,
      "num_tokens": 15073752.0,
      "step": 6870
    },
    {
      "entropy": 5.62511043548584,
      "epoch": 0.6823144104803494,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004961606708113216,
      "loss": 5.5402,
      "mean_token_accuracy": 0.1458849363029003,
      "num_tokens": 15084730.0,
      "step": 6875
    },
    {
      "entropy": 5.579836177825928,
      "epoch": 0.6828106391425168,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004961541505804441,
      "loss": 5.4377,
      "mean_token_accuracy": 0.14643568843603133,
      "num_tokens": 15095811.0,
      "step": 6880
    },
    {
      "entropy": 5.5513428211212155,
      "epoch": 0.6833068678046844,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004961476248654202,
      "loss": 5.4378,
      "mean_token_accuracy": 0.1498889833688736,
      "num_tokens": 15106102.0,
      "step": 6885
    },
    {
      "entropy": 5.627523517608642,
      "epoch": 0.6838030964668519,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004961410936664118,
      "loss": 5.5211,
      "mean_token_accuracy": 0.14715197756886483,
      "num_tokens": 15115907.0,
      "step": 6890
    },
    {
      "entropy": 5.602481651306152,
      "epoch": 0.6842993251290195,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004961345569835807,
      "loss": 5.5182,
      "mean_token_accuracy": 0.14342393204569817,
      "num_tokens": 15126788.0,
      "step": 6895
    },
    {
      "entropy": 5.560369825363159,
      "epoch": 0.684795553791187,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004961280148170892,
      "loss": 5.4677,
      "mean_token_accuracy": 0.14440471604466437,
      "num_tokens": 15137814.0,
      "step": 6900
    },
    {
      "entropy": 5.619000768661499,
      "epoch": 0.6852917824533545,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004961214671670994,
      "loss": 5.6767,
      "mean_token_accuracy": 0.1368719905614853,
      "num_tokens": 15148597.0,
      "step": 6905
    },
    {
      "entropy": 5.6778661727905275,
      "epoch": 0.685788011115522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961149140337737,
      "loss": 5.5038,
      "mean_token_accuracy": 0.14402559995651246,
      "num_tokens": 15158985.0,
      "step": 6910
    },
    {
      "entropy": 5.589445543289185,
      "epoch": 0.6862842397776896,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004961083554172745,
      "loss": 5.4882,
      "mean_token_accuracy": 0.14529428258538246,
      "num_tokens": 15170242.0,
      "step": 6915
    },
    {
      "entropy": 5.524306869506836,
      "epoch": 0.6867804684398571,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004961017913177646,
      "loss": 5.4384,
      "mean_token_accuracy": 0.15010960698127745,
      "num_tokens": 15180353.0,
      "step": 6920
    },
    {
      "entropy": 5.527584218978882,
      "epoch": 0.6872766971020247,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004960952217354066,
      "loss": 5.3934,
      "mean_token_accuracy": 0.15160997658967973,
      "num_tokens": 15191955.0,
      "step": 6925
    },
    {
      "entropy": 5.556703281402588,
      "epoch": 0.6877729257641921,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004960886466703636,
      "loss": 5.491,
      "mean_token_accuracy": 0.14613229781389236,
      "num_tokens": 15202088.0,
      "step": 6930
    },
    {
      "entropy": 5.5396201610565186,
      "epoch": 0.6882691544263597,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004960820661227987,
      "loss": 5.4297,
      "mean_token_accuracy": 0.1497945010662079,
      "num_tokens": 15212291.0,
      "step": 6935
    },
    {
      "entropy": 5.57582836151123,
      "epoch": 0.6887653830885272,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004960754800928748,
      "loss": 5.4634,
      "mean_token_accuracy": 0.1505399987101555,
      "num_tokens": 15223315.0,
      "step": 6940
    },
    {
      "entropy": 5.6802342414855955,
      "epoch": 0.6892616117506947,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004960688885807553,
      "loss": 5.5819,
      "mean_token_accuracy": 0.14442477524280548,
      "num_tokens": 15233416.0,
      "step": 6945
    },
    {
      "entropy": 5.600401973724365,
      "epoch": 0.6897578404128623,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004960622915866038,
      "loss": 5.4255,
      "mean_token_accuracy": 0.15087948441505433,
      "num_tokens": 15242951.0,
      "step": 6950
    },
    {
      "entropy": 5.517944860458374,
      "epoch": 0.6902540690750297,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960556891105839,
      "loss": 5.5088,
      "mean_token_accuracy": 0.14747381955385208,
      "num_tokens": 15253145.0,
      "step": 6955
    },
    {
      "entropy": 5.569548845291138,
      "epoch": 0.6907502977371973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004960490811528591,
      "loss": 5.4572,
      "mean_token_accuracy": 0.15006899386644362,
      "num_tokens": 15263712.0,
      "step": 6960
    },
    {
      "entropy": 5.678875684738159,
      "epoch": 0.6912465263993648,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004960424677135935,
      "loss": 5.561,
      "mean_token_accuracy": 0.1464661553502083,
      "num_tokens": 15273609.0,
      "step": 6965
    },
    {
      "entropy": 5.554692029953003,
      "epoch": 0.6917427550615324,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960358487929511,
      "loss": 5.4329,
      "mean_token_accuracy": 0.15622831135988235,
      "num_tokens": 15283934.0,
      "step": 6970
    },
    {
      "entropy": 5.58153510093689,
      "epoch": 0.6922389837236999,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004960292243910958,
      "loss": 5.5085,
      "mean_token_accuracy": 0.14504391551017762,
      "num_tokens": 15294394.0,
      "step": 6975
    },
    {
      "entropy": 5.572801828384399,
      "epoch": 0.6927352123858674,
      "grad_norm": 1.15625,
      "learning_rate": 0.000496022594508192,
      "loss": 5.4541,
      "mean_token_accuracy": 0.1531597927212715,
      "num_tokens": 15304622.0,
      "step": 6980
    },
    {
      "entropy": 5.57880802154541,
      "epoch": 0.6932314410480349,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004960159591444041,
      "loss": 5.3827,
      "mean_token_accuracy": 0.1477219358086586,
      "num_tokens": 15314094.0,
      "step": 6985
    },
    {
      "entropy": 5.57395076751709,
      "epoch": 0.6937276697102025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004960093182998968,
      "loss": 5.4941,
      "mean_token_accuracy": 0.14785784631967544,
      "num_tokens": 15324313.0,
      "step": 6990
    },
    {
      "entropy": 5.520534229278565,
      "epoch": 0.69422389837237,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004960026719748346,
      "loss": 5.3753,
      "mean_token_accuracy": 0.1674494668841362,
      "num_tokens": 15335195.0,
      "step": 6995
    },
    {
      "entropy": 5.664771175384521,
      "epoch": 0.6947201270345376,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004959960201693822,
      "loss": 5.6062,
      "mean_token_accuracy": 0.13749497383832932,
      "num_tokens": 15345798.0,
      "step": 7000
    },
    {
      "entropy": 5.587755107879639,
      "epoch": 0.695216355696705,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004959893628837048,
      "loss": 5.4379,
      "mean_token_accuracy": 0.14814721941947936,
      "num_tokens": 15356340.0,
      "step": 7005
    },
    {
      "entropy": 5.606631278991699,
      "epoch": 0.6957125843588726,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004959827001179675,
      "loss": 5.5326,
      "mean_token_accuracy": 0.1406660333275795,
      "num_tokens": 15367795.0,
      "step": 7010
    },
    {
      "entropy": 5.639504098892212,
      "epoch": 0.6962088130210401,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004959760318723353,
      "loss": 5.6548,
      "mean_token_accuracy": 0.1354363091289997,
      "num_tokens": 15379214.0,
      "step": 7015
    },
    {
      "entropy": 5.631503438949585,
      "epoch": 0.6967050416832076,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004959693581469736,
      "loss": 5.505,
      "mean_token_accuracy": 0.14959356635808946,
      "num_tokens": 15390242.0,
      "step": 7020
    },
    {
      "entropy": 5.565413761138916,
      "epoch": 0.6972012703453752,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004959626789420481,
      "loss": 5.4551,
      "mean_token_accuracy": 0.14655417650938035,
      "num_tokens": 15402600.0,
      "step": 7025
    },
    {
      "entropy": 5.559041261672974,
      "epoch": 0.6976974990075426,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004959559942577243,
      "loss": 5.4892,
      "mean_token_accuracy": 0.1493976816534996,
      "num_tokens": 15412883.0,
      "step": 7030
    },
    {
      "entropy": 5.667265462875366,
      "epoch": 0.6981937276697102,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004959493040941678,
      "loss": 5.5184,
      "mean_token_accuracy": 0.1440194934606552,
      "num_tokens": 15425919.0,
      "step": 7035
    },
    {
      "entropy": 5.536500644683838,
      "epoch": 0.6986899563318777,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959426084515448,
      "loss": 5.4518,
      "mean_token_accuracy": 0.14711084216833115,
      "num_tokens": 15437604.0,
      "step": 7040
    },
    {
      "entropy": 5.563373947143555,
      "epoch": 0.6991861849940453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959359073300212,
      "loss": 5.5189,
      "mean_token_accuracy": 0.14214091524481773,
      "num_tokens": 15449023.0,
      "step": 7045
    },
    {
      "entropy": 5.6633744716644285,
      "epoch": 0.6996824136562128,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004959292007297631,
      "loss": 5.5746,
      "mean_token_accuracy": 0.14451055973768234,
      "num_tokens": 15459567.0,
      "step": 7050
    },
    {
      "entropy": 5.642871713638305,
      "epoch": 0.7001786423183803,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495922488650937,
      "loss": 5.488,
      "mean_token_accuracy": 0.14581795483827592,
      "num_tokens": 15470950.0,
      "step": 7055
    },
    {
      "entropy": 5.532980060577392,
      "epoch": 0.7006748709805478,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004959157710937092,
      "loss": 5.4706,
      "mean_token_accuracy": 0.14553122073411942,
      "num_tokens": 15481258.0,
      "step": 7060
    },
    {
      "entropy": 5.583482265472412,
      "epoch": 0.7011710996427154,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004959090480582462,
      "loss": 5.4169,
      "mean_token_accuracy": 0.14957983493804933,
      "num_tokens": 15491672.0,
      "step": 7065
    },
    {
      "entropy": 5.501806259155273,
      "epoch": 0.7016673283048829,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004959023195447148,
      "loss": 5.48,
      "mean_token_accuracy": 0.1457020379602909,
      "num_tokens": 15502342.0,
      "step": 7070
    },
    {
      "entropy": 5.569126892089844,
      "epoch": 0.7021635569670505,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495895585553282,
      "loss": 5.5124,
      "mean_token_accuracy": 0.1426779270172119,
      "num_tokens": 15513430.0,
      "step": 7075
    },
    {
      "entropy": 5.792016124725341,
      "epoch": 0.7026597856292179,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004958888460841147,
      "loss": 5.6195,
      "mean_token_accuracy": 0.14699716717004777,
      "num_tokens": 15524889.0,
      "step": 7080
    },
    {
      "entropy": 5.56859073638916,
      "epoch": 0.7031560142913855,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004958821011373798,
      "loss": 5.4337,
      "mean_token_accuracy": 0.1466015636920929,
      "num_tokens": 15534998.0,
      "step": 7085
    },
    {
      "entropy": 5.485933828353882,
      "epoch": 0.703652242953553,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004958753507132449,
      "loss": 5.4375,
      "mean_token_accuracy": 0.15051031112670898,
      "num_tokens": 15544752.0,
      "step": 7090
    },
    {
      "entropy": 5.594445991516113,
      "epoch": 0.7041484716157205,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958685948118772,
      "loss": 5.4555,
      "mean_token_accuracy": 0.1496884658932686,
      "num_tokens": 15555490.0,
      "step": 7095
    },
    {
      "entropy": 5.535046195983886,
      "epoch": 0.7046447002778881,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004958618334334443,
      "loss": 5.4018,
      "mean_token_accuracy": 0.14933619946241378,
      "num_tokens": 15565752.0,
      "step": 7100
    },
    {
      "entropy": 5.572567415237427,
      "epoch": 0.7051409289400555,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004958550665781138,
      "loss": 5.5298,
      "mean_token_accuracy": 0.14128798097372056,
      "num_tokens": 15576368.0,
      "step": 7105
    },
    {
      "entropy": 5.687538719177246,
      "epoch": 0.7056371576022231,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958482942460536,
      "loss": 5.5128,
      "mean_token_accuracy": 0.1431007981300354,
      "num_tokens": 15587972.0,
      "step": 7110
    },
    {
      "entropy": 5.689526462554932,
      "epoch": 0.7061333862643906,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004958415164374316,
      "loss": 5.5604,
      "mean_token_accuracy": 0.14275309294462205,
      "num_tokens": 15599951.0,
      "step": 7115
    },
    {
      "entropy": 5.537745666503906,
      "epoch": 0.7066296149265582,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004958347331524158,
      "loss": 5.54,
      "mean_token_accuracy": 0.14321425855159758,
      "num_tokens": 15610183.0,
      "step": 7120
    },
    {
      "entropy": 5.607424926757813,
      "epoch": 0.7071258435887257,
      "grad_norm": 1.0,
      "learning_rate": 0.0004958279443911745,
      "loss": 5.551,
      "mean_token_accuracy": 0.13691192343831063,
      "num_tokens": 15621074.0,
      "step": 7125
    },
    {
      "entropy": 5.685508966445923,
      "epoch": 0.7076220722508932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000495821150153876,
      "loss": 5.543,
      "mean_token_accuracy": 0.14305271357297897,
      "num_tokens": 15632752.0,
      "step": 7130
    },
    {
      "entropy": 5.541546249389649,
      "epoch": 0.7081183009130607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000495814350440689,
      "loss": 5.5282,
      "mean_token_accuracy": 0.1388225868344307,
      "num_tokens": 15644703.0,
      "step": 7135
    },
    {
      "entropy": 5.580911540985108,
      "epoch": 0.7086145295752283,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004958075452517816,
      "loss": 5.4836,
      "mean_token_accuracy": 0.1460116907954216,
      "num_tokens": 15655278.0,
      "step": 7140
    },
    {
      "entropy": 5.636132049560547,
      "epoch": 0.7091107582373958,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004958007345873232,
      "loss": 5.5173,
      "mean_token_accuracy": 0.1469028428196907,
      "num_tokens": 15665998.0,
      "step": 7145
    },
    {
      "entropy": 5.5625019550323485,
      "epoch": 0.7096069868995634,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004957939184474821,
      "loss": 5.4081,
      "mean_token_accuracy": 0.1479800283908844,
      "num_tokens": 15676343.0,
      "step": 7150
    },
    {
      "entropy": 5.537234210968018,
      "epoch": 0.7101032155617308,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004957870968324277,
      "loss": 5.5032,
      "mean_token_accuracy": 0.14563940092921257,
      "num_tokens": 15686520.0,
      "step": 7155
    },
    {
      "entropy": 5.593419790267944,
      "epoch": 0.7105994442238984,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004957802697423291,
      "loss": 5.6037,
      "mean_token_accuracy": 0.13748548775911332,
      "num_tokens": 15698302.0,
      "step": 7160
    },
    {
      "entropy": 5.577815103530884,
      "epoch": 0.7110956728860659,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004957734371773554,
      "loss": 5.3747,
      "mean_token_accuracy": 0.1502794161438942,
      "num_tokens": 15708462.0,
      "step": 7165
    },
    {
      "entropy": 5.5939477443695065,
      "epoch": 0.7115919015482334,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004957665991376763,
      "loss": 5.4793,
      "mean_token_accuracy": 0.15064630806446075,
      "num_tokens": 15719234.0,
      "step": 7170
    },
    {
      "entropy": 5.518609142303466,
      "epoch": 0.712088130210401,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004957597556234613,
      "loss": 5.5232,
      "mean_token_accuracy": 0.14414323642849922,
      "num_tokens": 15732245.0,
      "step": 7175
    },
    {
      "entropy": 5.6174146175384525,
      "epoch": 0.7125843588725684,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004957529066348799,
      "loss": 5.4878,
      "mean_token_accuracy": 0.14517530798912048,
      "num_tokens": 15742403.0,
      "step": 7180
    },
    {
      "entropy": 5.532131814956665,
      "epoch": 0.713080587534736,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957460521721022,
      "loss": 5.373,
      "mean_token_accuracy": 0.15445320755243303,
      "num_tokens": 15752100.0,
      "step": 7185
    },
    {
      "entropy": 5.550415802001953,
      "epoch": 0.7135768161969035,
      "grad_norm": 1.0,
      "learning_rate": 0.000495739192235298,
      "loss": 5.4638,
      "mean_token_accuracy": 0.14823665097355843,
      "num_tokens": 15762315.0,
      "step": 7190
    },
    {
      "entropy": 5.541091012954712,
      "epoch": 0.7140730448590711,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004957323268246376,
      "loss": 5.4958,
      "mean_token_accuracy": 0.1444174513220787,
      "num_tokens": 15773841.0,
      "step": 7195
    },
    {
      "entropy": 5.566438150405884,
      "epoch": 0.7145692735212386,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004957254559402911,
      "loss": 5.4595,
      "mean_token_accuracy": 0.14450706094503402,
      "num_tokens": 15785511.0,
      "step": 7200
    },
    {
      "entropy": 5.565189695358276,
      "epoch": 0.7150655021834061,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004957185795824288,
      "loss": 5.4775,
      "mean_token_accuracy": 0.1433386817574501,
      "num_tokens": 15795394.0,
      "step": 7205
    },
    {
      "entropy": 5.55120997428894,
      "epoch": 0.7155617308455736,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957116977512214,
      "loss": 5.4381,
      "mean_token_accuracy": 0.14540694802999496,
      "num_tokens": 15805894.0,
      "step": 7210
    },
    {
      "entropy": 5.631142330169678,
      "epoch": 0.7160579595077412,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004957048104468395,
      "loss": 5.4908,
      "mean_token_accuracy": 0.14104633033275604,
      "num_tokens": 15816595.0,
      "step": 7215
    },
    {
      "entropy": 5.513956308364868,
      "epoch": 0.7165541881699087,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956979176694539,
      "loss": 5.4312,
      "mean_token_accuracy": 0.1509195327758789,
      "num_tokens": 15827638.0,
      "step": 7220
    },
    {
      "entropy": 5.5239582538604735,
      "epoch": 0.7170504168320763,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004956910194192355,
      "loss": 5.5155,
      "mean_token_accuracy": 0.14243088215589522,
      "num_tokens": 15838437.0,
      "step": 7225
    },
    {
      "entropy": 5.717722320556641,
      "epoch": 0.7175466454942437,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956841156963553,
      "loss": 5.6132,
      "mean_token_accuracy": 0.14455166310071946,
      "num_tokens": 15850110.0,
      "step": 7230
    },
    {
      "entropy": 5.607511520385742,
      "epoch": 0.7180428741564113,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004956772065009846,
      "loss": 5.4942,
      "mean_token_accuracy": 0.13964518457651137,
      "num_tokens": 15861786.0,
      "step": 7235
    },
    {
      "entropy": 5.594943046569824,
      "epoch": 0.7185391028185788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004956702918332948,
      "loss": 5.461,
      "mean_token_accuracy": 0.14653570726513862,
      "num_tokens": 15871574.0,
      "step": 7240
    },
    {
      "entropy": 5.576704740524292,
      "epoch": 0.7190353314807463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004956633716934572,
      "loss": 5.4517,
      "mean_token_accuracy": 0.15220170170068742,
      "num_tokens": 15881944.0,
      "step": 7245
    },
    {
      "entropy": 5.491304397583008,
      "epoch": 0.7195315601429139,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956564460816433,
      "loss": 5.4464,
      "mean_token_accuracy": 0.14957428425550462,
      "num_tokens": 15892511.0,
      "step": 7250
    },
    {
      "entropy": 5.530431032180786,
      "epoch": 0.7200277888050813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004956495149980252,
      "loss": 5.4702,
      "mean_token_accuracy": 0.1412671133875847,
      "num_tokens": 15903387.0,
      "step": 7255
    },
    {
      "entropy": 5.62625732421875,
      "epoch": 0.7205240174672489,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956425784427745,
      "loss": 5.4957,
      "mean_token_accuracy": 0.14424169510602952,
      "num_tokens": 15915117.0,
      "step": 7260
    },
    {
      "entropy": 5.551268863677978,
      "epoch": 0.7210202461294164,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004956356364160634,
      "loss": 5.4486,
      "mean_token_accuracy": 0.14467760920524597,
      "num_tokens": 15925163.0,
      "step": 7265
    },
    {
      "entropy": 5.435801792144775,
      "epoch": 0.721516474791584,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004956286889180639,
      "loss": 5.373,
      "mean_token_accuracy": 0.1516958013176918,
      "num_tokens": 15935021.0,
      "step": 7270
    },
    {
      "entropy": 5.51355504989624,
      "epoch": 0.7220127034537515,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004956217359489483,
      "loss": 5.4702,
      "mean_token_accuracy": 0.1447344556450844,
      "num_tokens": 15945416.0,
      "step": 7275
    },
    {
      "entropy": 5.719403553009033,
      "epoch": 0.722508932115919,
      "grad_norm": 1.0625,
      "learning_rate": 0.000495614777508889,
      "loss": 5.593,
      "mean_token_accuracy": 0.1420062780380249,
      "num_tokens": 15955388.0,
      "step": 7280
    },
    {
      "entropy": 5.678073215484619,
      "epoch": 0.7230051607780865,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004956078135980586,
      "loss": 5.5941,
      "mean_token_accuracy": 0.14293039590120316,
      "num_tokens": 15966721.0,
      "step": 7285
    },
    {
      "entropy": 5.564630651473999,
      "epoch": 0.7235013894402541,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004956008442166299,
      "loss": 5.5117,
      "mean_token_accuracy": 0.1447232998907566,
      "num_tokens": 15976388.0,
      "step": 7290
    },
    {
      "entropy": 5.609762859344483,
      "epoch": 0.7239976181024216,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004955938693647756,
      "loss": 5.4202,
      "mean_token_accuracy": 0.14851459115743637,
      "num_tokens": 15986632.0,
      "step": 7295
    },
    {
      "entropy": 5.584561443328857,
      "epoch": 0.7244938467645892,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004955868890426686,
      "loss": 5.4605,
      "mean_token_accuracy": 0.14389446377754211,
      "num_tokens": 15996626.0,
      "step": 7300
    },
    {
      "entropy": 5.50442066192627,
      "epoch": 0.7249900754267566,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004955799032504821,
      "loss": 5.4439,
      "mean_token_accuracy": 0.14867453575134276,
      "num_tokens": 16007055.0,
      "step": 7305
    },
    {
      "entropy": 5.685265064239502,
      "epoch": 0.7254863040889242,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004955729119883893,
      "loss": 5.505,
      "mean_token_accuracy": 0.14218003749847413,
      "num_tokens": 16019087.0,
      "step": 7310
    },
    {
      "entropy": 5.5734282493591305,
      "epoch": 0.7259825327510917,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004955659152565636,
      "loss": 5.5136,
      "mean_token_accuracy": 0.14022271633148192,
      "num_tokens": 16029980.0,
      "step": 7315
    },
    {
      "entropy": 5.56582989692688,
      "epoch": 0.7264787614132592,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004955589130551784,
      "loss": 5.4902,
      "mean_token_accuracy": 0.14813822954893113,
      "num_tokens": 16040932.0,
      "step": 7320
    },
    {
      "entropy": 5.504037523269654,
      "epoch": 0.7269749900754268,
      "grad_norm": 1.0,
      "learning_rate": 0.0004955519053844076,
      "loss": 5.4475,
      "mean_token_accuracy": 0.1465040400624275,
      "num_tokens": 16051809.0,
      "step": 7325
    },
    {
      "entropy": 5.584791994094848,
      "epoch": 0.7274712187375942,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004955448922444246,
      "loss": 5.4936,
      "mean_token_accuracy": 0.14538883343338965,
      "num_tokens": 16062957.0,
      "step": 7330
    },
    {
      "entropy": 5.6081489562988285,
      "epoch": 0.7279674473997618,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955378736354036,
      "loss": 5.5306,
      "mean_token_accuracy": 0.1465644970536232,
      "num_tokens": 16073592.0,
      "step": 7335
    },
    {
      "entropy": 5.5162529945373535,
      "epoch": 0.7284636760619293,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955308495575185,
      "loss": 5.4731,
      "mean_token_accuracy": 0.1491437241435051,
      "num_tokens": 16085054.0,
      "step": 7340
    },
    {
      "entropy": 5.6472797870635985,
      "epoch": 0.7289599047240969,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004955238200109436,
      "loss": 5.5623,
      "mean_token_accuracy": 0.14247652441263198,
      "num_tokens": 16096936.0,
      "step": 7345
    },
    {
      "entropy": 5.651778364181519,
      "epoch": 0.7294561333862644,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004955167849958531,
      "loss": 5.5733,
      "mean_token_accuracy": 0.1442542314529419,
      "num_tokens": 16107673.0,
      "step": 7350
    },
    {
      "entropy": 5.611193323135376,
      "epoch": 0.729952362048432,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004955097445124215,
      "loss": 5.4315,
      "mean_token_accuracy": 0.15109449476003647,
      "num_tokens": 16117941.0,
      "step": 7355
    },
    {
      "entropy": 5.522215890884399,
      "epoch": 0.7304485907105994,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955026985608235,
      "loss": 5.3946,
      "mean_token_accuracy": 0.15170413255691528,
      "num_tokens": 16128737.0,
      "step": 7360
    },
    {
      "entropy": 5.679868459701538,
      "epoch": 0.730944819372767,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004954956471412336,
      "loss": 5.5523,
      "mean_token_accuracy": 0.14851489588618277,
      "num_tokens": 16139599.0,
      "step": 7365
    },
    {
      "entropy": 5.467503213882447,
      "epoch": 0.7314410480349345,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004954885902538269,
      "loss": 5.3892,
      "mean_token_accuracy": 0.1426581025123596,
      "num_tokens": 16150796.0,
      "step": 7370
    },
    {
      "entropy": 5.492081499099731,
      "epoch": 0.7319372766971021,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004954815278987781,
      "loss": 5.461,
      "mean_token_accuracy": 0.15160699039697648,
      "num_tokens": 16161236.0,
      "step": 7375
    },
    {
      "entropy": 5.619355916976929,
      "epoch": 0.7324335053592695,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004954744600762627,
      "loss": 5.5711,
      "mean_token_accuracy": 0.14571552500128745,
      "num_tokens": 16172210.0,
      "step": 7380
    },
    {
      "entropy": 5.515890884399414,
      "epoch": 0.7329297340214371,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004954673867864555,
      "loss": 5.4306,
      "mean_token_accuracy": 0.15293084233999252,
      "num_tokens": 16182950.0,
      "step": 7385
    },
    {
      "entropy": 5.545074415206909,
      "epoch": 0.7334259626836046,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004954603080295323,
      "loss": 5.3393,
      "mean_token_accuracy": 0.15570643097162246,
      "num_tokens": 16193971.0,
      "step": 7390
    },
    {
      "entropy": 5.575541734695435,
      "epoch": 0.7339221913457721,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004954532238056686,
      "loss": 5.4653,
      "mean_token_accuracy": 0.14673469215631485,
      "num_tokens": 16205774.0,
      "step": 7395
    },
    {
      "entropy": 5.530745601654052,
      "epoch": 0.7344184200079397,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004954461341150399,
      "loss": 5.4464,
      "mean_token_accuracy": 0.1477830305695534,
      "num_tokens": 16216626.0,
      "step": 7400
    },
    {
      "entropy": 5.559146928787231,
      "epoch": 0.7349146486701071,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004954390389578222,
      "loss": 5.524,
      "mean_token_accuracy": 0.14277490824460984,
      "num_tokens": 16227748.0,
      "step": 7405
    },
    {
      "entropy": 5.603081703186035,
      "epoch": 0.7354108773322747,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004954319383341912,
      "loss": 5.5074,
      "mean_token_accuracy": 0.14815373197197915,
      "num_tokens": 16240390.0,
      "step": 7410
    },
    {
      "entropy": 5.56106972694397,
      "epoch": 0.7359071059944422,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004954248322443232,
      "loss": 5.4562,
      "mean_token_accuracy": 0.15491658002138137,
      "num_tokens": 16250563.0,
      "step": 7415
    },
    {
      "entropy": 5.577298974990844,
      "epoch": 0.7364033346566098,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004954177206883943,
      "loss": 5.5244,
      "mean_token_accuracy": 0.13953978568315506,
      "num_tokens": 16260722.0,
      "step": 7420
    },
    {
      "entropy": 5.703858375549316,
      "epoch": 0.7368995633187773,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004954106036665809,
      "loss": 5.5176,
      "mean_token_accuracy": 0.1494123511016369,
      "num_tokens": 16270258.0,
      "step": 7425
    },
    {
      "entropy": 5.604127550125122,
      "epoch": 0.7373957919809448,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004954034811790594,
      "loss": 5.3918,
      "mean_token_accuracy": 0.15408581346273423,
      "num_tokens": 16279504.0,
      "step": 7430
    },
    {
      "entropy": 5.475208711624146,
      "epoch": 0.7378920206431123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004953963532260067,
      "loss": 5.4177,
      "mean_token_accuracy": 0.14885843843221663,
      "num_tokens": 16291327.0,
      "step": 7435
    },
    {
      "entropy": 5.565517568588257,
      "epoch": 0.7383882493052799,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004953892198075991,
      "loss": 5.5401,
      "mean_token_accuracy": 0.14473335593938827,
      "num_tokens": 16302778.0,
      "step": 7440
    },
    {
      "entropy": 5.637333536148072,
      "epoch": 0.7388844779674474,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004953820809240138,
      "loss": 5.4975,
      "mean_token_accuracy": 0.1401463508605957,
      "num_tokens": 16314373.0,
      "step": 7445
    },
    {
      "entropy": 5.55228214263916,
      "epoch": 0.739380706629615,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004953749365754279,
      "loss": 5.4584,
      "mean_token_accuracy": 0.1441833794116974,
      "num_tokens": 16324518.0,
      "step": 7450
    },
    {
      "entropy": 5.55927357673645,
      "epoch": 0.7398769352917824,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004953677867620184,
      "loss": 5.4023,
      "mean_token_accuracy": 0.1472688525915146,
      "num_tokens": 16335657.0,
      "step": 7455
    },
    {
      "entropy": 5.614712810516357,
      "epoch": 0.74037316395395,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953606314839626,
      "loss": 5.5972,
      "mean_token_accuracy": 0.13894294947385788,
      "num_tokens": 16346497.0,
      "step": 7460
    },
    {
      "entropy": 5.543916463851929,
      "epoch": 0.7408693926161175,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000495353470741438,
      "loss": 5.4452,
      "mean_token_accuracy": 0.14747430235147477,
      "num_tokens": 16358175.0,
      "step": 7465
    },
    {
      "entropy": 5.608590364456177,
      "epoch": 0.741365621278285,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000495346304534622,
      "loss": 5.5875,
      "mean_token_accuracy": 0.13845830634236336,
      "num_tokens": 16369803.0,
      "step": 7470
    },
    {
      "entropy": 5.500791311264038,
      "epoch": 0.7418618499404526,
      "grad_norm": 1.0,
      "learning_rate": 0.0004953391328636926,
      "loss": 5.2683,
      "mean_token_accuracy": 0.15424345731735228,
      "num_tokens": 16380535.0,
      "step": 7475
    },
    {
      "entropy": 5.485128498077392,
      "epoch": 0.74235807860262,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004953319557288275,
      "loss": 5.4493,
      "mean_token_accuracy": 0.14647470265626908,
      "num_tokens": 16390518.0,
      "step": 7480
    },
    {
      "entropy": 5.5716887474060055,
      "epoch": 0.7428543072647876,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004953247731302046,
      "loss": 5.4838,
      "mean_token_accuracy": 0.13988323733210564,
      "num_tokens": 16400627.0,
      "step": 7485
    },
    {
      "entropy": 5.58498363494873,
      "epoch": 0.7433505359269551,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004953175850680022,
      "loss": 5.5356,
      "mean_token_accuracy": 0.14903751388192177,
      "num_tokens": 16411136.0,
      "step": 7490
    },
    {
      "entropy": 5.606583786010742,
      "epoch": 0.7438467645891227,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004953103915423983,
      "loss": 5.475,
      "mean_token_accuracy": 0.1458166606724262,
      "num_tokens": 16422830.0,
      "step": 7495
    },
    {
      "entropy": 5.504616928100586,
      "epoch": 0.7443429932512902,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004953031925535715,
      "loss": 5.4084,
      "mean_token_accuracy": 0.15125985890626908,
      "num_tokens": 16434158.0,
      "step": 7500
    },
    {
      "entropy": 5.522843360900879,
      "epoch": 0.7448392219134577,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004952959881017003,
      "loss": 5.4494,
      "mean_token_accuracy": 0.15473447889089584,
      "num_tokens": 16444562.0,
      "step": 7505
    },
    {
      "entropy": 5.558233737945557,
      "epoch": 0.7453354505756252,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004952887781869631,
      "loss": 5.4901,
      "mean_token_accuracy": 0.1495634898543358,
      "num_tokens": 16456353.0,
      "step": 7510
    },
    {
      "entropy": 5.569683790206909,
      "epoch": 0.7458316792377928,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004952815628095391,
      "loss": 5.5054,
      "mean_token_accuracy": 0.13602832183241845,
      "num_tokens": 16465756.0,
      "step": 7515
    },
    {
      "entropy": 5.557487821578979,
      "epoch": 0.7463279078999603,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004952743419696069,
      "loss": 5.4586,
      "mean_token_accuracy": 0.14768147766590117,
      "num_tokens": 16477562.0,
      "step": 7520
    },
    {
      "entropy": 5.588579940795898,
      "epoch": 0.7468241365621279,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004952671156673456,
      "loss": 5.4687,
      "mean_token_accuracy": 0.14742666482925415,
      "num_tokens": 16488134.0,
      "step": 7525
    },
    {
      "entropy": 5.590447807312012,
      "epoch": 0.7473203652242953,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952598839029345,
      "loss": 5.468,
      "mean_token_accuracy": 0.1477799743413925,
      "num_tokens": 16499146.0,
      "step": 7530
    },
    {
      "entropy": 5.580975914001465,
      "epoch": 0.7478165938864629,
      "grad_norm": 1.0,
      "learning_rate": 0.0004952526466765529,
      "loss": 5.5158,
      "mean_token_accuracy": 0.14711739048361777,
      "num_tokens": 16510334.0,
      "step": 7535
    },
    {
      "entropy": 5.558179759979248,
      "epoch": 0.7483128225486304,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004952454039883804,
      "loss": 5.4042,
      "mean_token_accuracy": 0.1516387343406677,
      "num_tokens": 16520574.0,
      "step": 7540
    },
    {
      "entropy": 5.598708391189575,
      "epoch": 0.7488090512107979,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004952381558385963,
      "loss": 5.5447,
      "mean_token_accuracy": 0.1414144791662693,
      "num_tokens": 16532246.0,
      "step": 7545
    },
    {
      "entropy": 5.6961428165435795,
      "epoch": 0.7493052798729655,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004952309022273805,
      "loss": 5.5114,
      "mean_token_accuracy": 0.13977369368076326,
      "num_tokens": 16541756.0,
      "step": 7550
    },
    {
      "entropy": 5.575630235671997,
      "epoch": 0.7498015085351329,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004952236431549129,
      "loss": 5.5111,
      "mean_token_accuracy": 0.14872334748506547,
      "num_tokens": 16551061.0,
      "step": 7555
    },
    {
      "entropy": 5.503014755249024,
      "epoch": 0.7502977371973005,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004952163786213735,
      "loss": 5.4613,
      "mean_token_accuracy": 0.14346159622073174,
      "num_tokens": 16561155.0,
      "step": 7560
    },
    {
      "entropy": 5.554971742630005,
      "epoch": 0.750793965859468,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004952091086269425,
      "loss": 5.4318,
      "mean_token_accuracy": 0.14320596307516098,
      "num_tokens": 16571843.0,
      "step": 7565
    },
    {
      "entropy": 5.559373283386231,
      "epoch": 0.7512901945216356,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004952018331718001,
      "loss": 5.5078,
      "mean_token_accuracy": 0.14633592888712882,
      "num_tokens": 16583395.0,
      "step": 7570
    },
    {
      "entropy": 5.665293264389038,
      "epoch": 0.7517864231838031,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004951945522561268,
      "loss": 5.5512,
      "mean_token_accuracy": 0.1462844803929329,
      "num_tokens": 16594545.0,
      "step": 7575
    },
    {
      "entropy": 5.660638332366943,
      "epoch": 0.7522826518459707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004951872658801029,
      "loss": 5.6102,
      "mean_token_accuracy": 0.14177481532096864,
      "num_tokens": 16605901.0,
      "step": 7580
    },
    {
      "entropy": 5.59262375831604,
      "epoch": 0.7527788805081381,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951799740439093,
      "loss": 5.5436,
      "mean_token_accuracy": 0.1418263517320156,
      "num_tokens": 16616783.0,
      "step": 7585
    },
    {
      "entropy": 5.632286262512207,
      "epoch": 0.7532751091703057,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000495172676747727,
      "loss": 5.4819,
      "mean_token_accuracy": 0.14710267186164855,
      "num_tokens": 16628655.0,
      "step": 7590
    },
    {
      "entropy": 5.58706488609314,
      "epoch": 0.7537713378324732,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004951653739917366,
      "loss": 5.5011,
      "mean_token_accuracy": 0.1442305900156498,
      "num_tokens": 16639403.0,
      "step": 7595
    },
    {
      "entropy": 5.628613233566284,
      "epoch": 0.7542675664946408,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004951580657761192,
      "loss": 5.5222,
      "mean_token_accuracy": 0.15108118057250977,
      "num_tokens": 16649937.0,
      "step": 7600
    },
    {
      "entropy": 5.522719669342041,
      "epoch": 0.7547637951568082,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004951507521010563,
      "loss": 5.3958,
      "mean_token_accuracy": 0.1533457174897194,
      "num_tokens": 16661466.0,
      "step": 7605
    },
    {
      "entropy": 5.500709295272827,
      "epoch": 0.7552600238189758,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004951434329667293,
      "loss": 5.4394,
      "mean_token_accuracy": 0.14726634547114373,
      "num_tokens": 16673362.0,
      "step": 7610
    },
    {
      "entropy": 5.527061939239502,
      "epoch": 0.7557562524811433,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004951361083733192,
      "loss": 5.4726,
      "mean_token_accuracy": 0.1475066438317299,
      "num_tokens": 16685555.0,
      "step": 7615
    },
    {
      "entropy": 5.606302309036255,
      "epoch": 0.7562524811433108,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004951287783210082,
      "loss": 5.4514,
      "mean_token_accuracy": 0.14147108867764474,
      "num_tokens": 16696006.0,
      "step": 7620
    },
    {
      "entropy": 5.572164821624756,
      "epoch": 0.7567487098054784,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004951214428099778,
      "loss": 5.4554,
      "mean_token_accuracy": 0.1504653662443161,
      "num_tokens": 16706784.0,
      "step": 7625
    },
    {
      "entropy": 5.488680219650268,
      "epoch": 0.7572449384676458,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951141018404099,
      "loss": 5.4104,
      "mean_token_accuracy": 0.15233126506209374,
      "num_tokens": 16718492.0,
      "step": 7630
    },
    {
      "entropy": 5.560507678985596,
      "epoch": 0.7577411671298134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004951067554124865,
      "loss": 5.4932,
      "mean_token_accuracy": 0.1460588678717613,
      "num_tokens": 16729711.0,
      "step": 7635
    },
    {
      "entropy": 5.608654451370239,
      "epoch": 0.7582373957919809,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00049509940352639,
      "loss": 5.4861,
      "mean_token_accuracy": 0.14390336126089096,
      "num_tokens": 16739913.0,
      "step": 7640
    },
    {
      "entropy": 5.609306383132934,
      "epoch": 0.7587336244541485,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004950920461823024,
      "loss": 5.5022,
      "mean_token_accuracy": 0.1434624195098877,
      "num_tokens": 16752273.0,
      "step": 7645
    },
    {
      "entropy": 5.552107429504394,
      "epoch": 0.759229853116316,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004950846833804064,
      "loss": 5.4391,
      "mean_token_accuracy": 0.14713409841060637,
      "num_tokens": 16762050.0,
      "step": 7650
    },
    {
      "entropy": 5.552213716506958,
      "epoch": 0.7597260817784836,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004950773151208845,
      "loss": 5.5134,
      "mean_token_accuracy": 0.14788412153720856,
      "num_tokens": 16772376.0,
      "step": 7655
    },
    {
      "entropy": 5.565105199813843,
      "epoch": 0.760222310440651,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004950699414039194,
      "loss": 5.4616,
      "mean_token_accuracy": 0.14564137831330298,
      "num_tokens": 16783307.0,
      "step": 7660
    },
    {
      "entropy": 5.600151109695434,
      "epoch": 0.7607185391028186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000495062562229694,
      "loss": 5.357,
      "mean_token_accuracy": 0.1529240094125271,
      "num_tokens": 16793582.0,
      "step": 7665
    },
    {
      "entropy": 5.512470245361328,
      "epoch": 0.7612147677649861,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004950551775983912,
      "loss": 5.4621,
      "mean_token_accuracy": 0.14833781272172927,
      "num_tokens": 16805160.0,
      "step": 7670
    },
    {
      "entropy": 5.5589155673980715,
      "epoch": 0.7617109964271537,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004950477875101942,
      "loss": 5.5202,
      "mean_token_accuracy": 0.1466103360056877,
      "num_tokens": 16816108.0,
      "step": 7675
    },
    {
      "entropy": 5.566047430038452,
      "epoch": 0.7622072250893211,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004950403919652863,
      "loss": 5.3965,
      "mean_token_accuracy": 0.15156176090240478,
      "num_tokens": 16828126.0,
      "step": 7680
    },
    {
      "entropy": 5.585244083404541,
      "epoch": 0.7627034537514887,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004950329909638507,
      "loss": 5.4794,
      "mean_token_accuracy": 0.14587293043732644,
      "num_tokens": 16838186.0,
      "step": 7685
    },
    {
      "entropy": 5.529992771148682,
      "epoch": 0.7631996824136562,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495025584506071,
      "loss": 5.4157,
      "mean_token_accuracy": 0.14930623769760132,
      "num_tokens": 16848588.0,
      "step": 7690
    },
    {
      "entropy": 5.579004240036011,
      "epoch": 0.7636959110758237,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004950181725921311,
      "loss": 5.5312,
      "mean_token_accuracy": 0.1479268953204155,
      "num_tokens": 16860269.0,
      "step": 7695
    },
    {
      "entropy": 5.54180908203125,
      "epoch": 0.7641921397379913,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004950107552222145,
      "loss": 5.5276,
      "mean_token_accuracy": 0.14610255509614944,
      "num_tokens": 16872221.0,
      "step": 7700
    },
    {
      "entropy": 5.57837405204773,
      "epoch": 0.7646883684001587,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004950033323965052,
      "loss": 5.3975,
      "mean_token_accuracy": 0.1441833920776844,
      "num_tokens": 16882588.0,
      "step": 7705
    },
    {
      "entropy": 5.625712299346924,
      "epoch": 0.7651845970623263,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004949959041151875,
      "loss": 5.4715,
      "mean_token_accuracy": 0.14966340512037277,
      "num_tokens": 16894278.0,
      "step": 7710
    },
    {
      "entropy": 5.554607963562011,
      "epoch": 0.7656808257244938,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004949884703784453,
      "loss": 5.4477,
      "mean_token_accuracy": 0.14578246995806693,
      "num_tokens": 16905448.0,
      "step": 7715
    },
    {
      "entropy": 5.520160293579101,
      "epoch": 0.7661770543866614,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000494981031186463,
      "loss": 5.4885,
      "mean_token_accuracy": 0.1419324740767479,
      "num_tokens": 16914960.0,
      "step": 7720
    },
    {
      "entropy": 5.563627433776856,
      "epoch": 0.7666732830488289,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949735865394253,
      "loss": 5.5382,
      "mean_token_accuracy": 0.14516519978642464,
      "num_tokens": 16926839.0,
      "step": 7725
    },
    {
      "entropy": 5.583517599105835,
      "epoch": 0.7671695117109965,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004949661364375167,
      "loss": 5.4112,
      "mean_token_accuracy": 0.15338422507047653,
      "num_tokens": 16938553.0,
      "step": 7730
    },
    {
      "entropy": 5.586976146697998,
      "epoch": 0.7676657403731639,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004949586808809217,
      "loss": 5.4843,
      "mean_token_accuracy": 0.14464979320764543,
      "num_tokens": 16947987.0,
      "step": 7735
    },
    {
      "entropy": 5.478079700469971,
      "epoch": 0.7681619690353315,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004949512198698256,
      "loss": 5.4405,
      "mean_token_accuracy": 0.1481389209628105,
      "num_tokens": 16959359.0,
      "step": 7740
    },
    {
      "entropy": 5.568147277832031,
      "epoch": 0.768658197697499,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000494943753404413,
      "loss": 5.4855,
      "mean_token_accuracy": 0.14885537028312684,
      "num_tokens": 16970731.0,
      "step": 7745
    },
    {
      "entropy": 5.543462038040161,
      "epoch": 0.7691544263596666,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004949362814848693,
      "loss": 5.4336,
      "mean_token_accuracy": 0.1499241515994072,
      "num_tokens": 16981868.0,
      "step": 7750
    },
    {
      "entropy": 5.57402868270874,
      "epoch": 0.769650655021834,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004949288041113799,
      "loss": 5.4812,
      "mean_token_accuracy": 0.1407908983528614,
      "num_tokens": 16994639.0,
      "step": 7755
    },
    {
      "entropy": 5.55854287147522,
      "epoch": 0.7701468836840016,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004949213212841299,
      "loss": 5.4733,
      "mean_token_accuracy": 0.14500055760145186,
      "num_tokens": 17005067.0,
      "step": 7760
    },
    {
      "entropy": 5.544031572341919,
      "epoch": 0.7706431123461691,
      "grad_norm": 1.140625,
      "learning_rate": 0.000494913833003305,
      "loss": 5.3901,
      "mean_token_accuracy": 0.15047784000635148,
      "num_tokens": 17015102.0,
      "step": 7765
    },
    {
      "entropy": 5.550044584274292,
      "epoch": 0.7711393410083366,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494906339269091,
      "loss": 5.4302,
      "mean_token_accuracy": 0.14870319589972497,
      "num_tokens": 17026290.0,
      "step": 7770
    },
    {
      "entropy": 5.521420955657959,
      "epoch": 0.7716355696705042,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004948988400816736,
      "loss": 5.3934,
      "mean_token_accuracy": 0.14590171426534654,
      "num_tokens": 17035445.0,
      "step": 7775
    },
    {
      "entropy": 5.607032442092896,
      "epoch": 0.7721317983326716,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004948913354412388,
      "loss": 5.4983,
      "mean_token_accuracy": 0.1492329940199852,
      "num_tokens": 17046515.0,
      "step": 7780
    },
    {
      "entropy": 5.558110332489013,
      "epoch": 0.7726280269948392,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004948838253479727,
      "loss": 5.4648,
      "mean_token_accuracy": 0.14597852155566216,
      "num_tokens": 17056988.0,
      "step": 7785
    },
    {
      "entropy": 5.578532314300537,
      "epoch": 0.7731242556570067,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004948763098020616,
      "loss": 5.5149,
      "mean_token_accuracy": 0.13886399269104005,
      "num_tokens": 17068304.0,
      "step": 7790
    },
    {
      "entropy": 5.566470050811768,
      "epoch": 0.7736204843191743,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004948687888036918,
      "loss": 5.4372,
      "mean_token_accuracy": 0.14790554717183113,
      "num_tokens": 17078405.0,
      "step": 7795
    },
    {
      "entropy": 5.483156538009643,
      "epoch": 0.7741167129813418,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004948612623530497,
      "loss": 5.3343,
      "mean_token_accuracy": 0.15858649760484694,
      "num_tokens": 17087839.0,
      "step": 7800
    },
    {
      "entropy": 5.47665548324585,
      "epoch": 0.7746129416435094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004948537304503221,
      "loss": 5.4172,
      "mean_token_accuracy": 0.15127672851085663,
      "num_tokens": 17099386.0,
      "step": 7805
    },
    {
      "entropy": 5.623297452926636,
      "epoch": 0.7751091703056768,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004948461930956959,
      "loss": 5.4956,
      "mean_token_accuracy": 0.1418422505259514,
      "num_tokens": 17111754.0,
      "step": 7810
    },
    {
      "entropy": 5.589095163345337,
      "epoch": 0.7756053989678444,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948386502893577,
      "loss": 5.374,
      "mean_token_accuracy": 0.15193604826927185,
      "num_tokens": 17122882.0,
      "step": 7815
    },
    {
      "entropy": 5.510799741744995,
      "epoch": 0.7761016276300119,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004948311020314946,
      "loss": 5.4916,
      "mean_token_accuracy": 0.1442389689385891,
      "num_tokens": 17133986.0,
      "step": 7820
    },
    {
      "entropy": 5.557250308990478,
      "epoch": 0.7765978562921795,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004948235483222941,
      "loss": 5.4659,
      "mean_token_accuracy": 0.15105831027030944,
      "num_tokens": 17144784.0,
      "step": 7825
    },
    {
      "entropy": 5.5940186977386475,
      "epoch": 0.777094084954347,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004948159891619431,
      "loss": 5.5658,
      "mean_token_accuracy": 0.1403675876557827,
      "num_tokens": 17156328.0,
      "step": 7830
    },
    {
      "entropy": 5.615364456176758,
      "epoch": 0.7775903136165145,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948084245506293,
      "loss": 5.3996,
      "mean_token_accuracy": 0.14310784488916398,
      "num_tokens": 17168062.0,
      "step": 7835
    },
    {
      "entropy": 5.56660680770874,
      "epoch": 0.778086542278682,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00049480085448854,
      "loss": 5.4203,
      "mean_token_accuracy": 0.14798627644777299,
      "num_tokens": 17178872.0,
      "step": 7840
    },
    {
      "entropy": 5.551462888717651,
      "epoch": 0.7785827709408495,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004947932789758635,
      "loss": 5.4814,
      "mean_token_accuracy": 0.1472082629799843,
      "num_tokens": 17189445.0,
      "step": 7845
    },
    {
      "entropy": 5.516396951675415,
      "epoch": 0.7790789996030171,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000494785698012787,
      "loss": 5.4254,
      "mean_token_accuracy": 0.15329916179180145,
      "num_tokens": 17199454.0,
      "step": 7850
    },
    {
      "entropy": 5.536207389831543,
      "epoch": 0.7795752282651845,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004947781115994989,
      "loss": 5.4681,
      "mean_token_accuracy": 0.1425333321094513,
      "num_tokens": 17210729.0,
      "step": 7855
    },
    {
      "entropy": 5.560697221755982,
      "epoch": 0.7800714569273521,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004947705197361872,
      "loss": 5.5383,
      "mean_token_accuracy": 0.14402690678834915,
      "num_tokens": 17222961.0,
      "step": 7860
    },
    {
      "entropy": 5.557329511642456,
      "epoch": 0.7805676855895196,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947629224230401,
      "loss": 5.4261,
      "mean_token_accuracy": 0.15382362008094788,
      "num_tokens": 17234026.0,
      "step": 7865
    },
    {
      "entropy": 5.652938413619995,
      "epoch": 0.7810639142516872,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004947553196602461,
      "loss": 5.5461,
      "mean_token_accuracy": 0.14615664333105088,
      "num_tokens": 17244643.0,
      "step": 7870
    },
    {
      "entropy": 5.60826473236084,
      "epoch": 0.7815601429138547,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004947477114479939,
      "loss": 5.4717,
      "mean_token_accuracy": 0.15020737200975418,
      "num_tokens": 17255382.0,
      "step": 7875
    },
    {
      "entropy": 5.526484537124634,
      "epoch": 0.7820563715760223,
      "grad_norm": 1.125,
      "learning_rate": 0.0004947400977864718,
      "loss": 5.4502,
      "mean_token_accuracy": 0.1419450044631958,
      "num_tokens": 17265860.0,
      "step": 7880
    },
    {
      "entropy": 5.573448562622071,
      "epoch": 0.7825526002381897,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947324786758688,
      "loss": 5.5273,
      "mean_token_accuracy": 0.14007658809423446,
      "num_tokens": 17277570.0,
      "step": 7885
    },
    {
      "entropy": 5.653734016418457,
      "epoch": 0.7830488289003573,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004947248541163738,
      "loss": 5.5811,
      "mean_token_accuracy": 0.14135265946388245,
      "num_tokens": 17288684.0,
      "step": 7890
    },
    {
      "entropy": 5.610208892822266,
      "epoch": 0.7835450575625248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004947172241081759,
      "loss": 5.4208,
      "mean_token_accuracy": 0.15115006715059282,
      "num_tokens": 17299541.0,
      "step": 7895
    },
    {
      "entropy": 5.543471145629883,
      "epoch": 0.7840412862246924,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004947095886514643,
      "loss": 5.3931,
      "mean_token_accuracy": 0.1523099422454834,
      "num_tokens": 17310243.0,
      "step": 7900
    },
    {
      "entropy": 5.554357624053955,
      "epoch": 0.7845375148868599,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004947019477464283,
      "loss": 5.4811,
      "mean_token_accuracy": 0.14539020135998726,
      "num_tokens": 17321167.0,
      "step": 7905
    },
    {
      "entropy": 5.451063537597657,
      "epoch": 0.7850337435490274,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004946943013932574,
      "loss": 5.4601,
      "mean_token_accuracy": 0.1496385008096695,
      "num_tokens": 17331076.0,
      "step": 7910
    },
    {
      "entropy": 5.571023988723755,
      "epoch": 0.7855299722111949,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004946866495921413,
      "loss": 5.4468,
      "mean_token_accuracy": 0.14041875228285788,
      "num_tokens": 17341337.0,
      "step": 7915
    },
    {
      "entropy": 5.610820722579956,
      "epoch": 0.7860262008733624,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004946789923432695,
      "loss": 5.463,
      "mean_token_accuracy": 0.14773353189229965,
      "num_tokens": 17353055.0,
      "step": 7920
    },
    {
      "entropy": 5.4558464050292965,
      "epoch": 0.78652242953553,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004946713296468322,
      "loss": 5.3712,
      "mean_token_accuracy": 0.14940313547849654,
      "num_tokens": 17364442.0,
      "step": 7925
    },
    {
      "entropy": 5.534676885604858,
      "epoch": 0.7870186581976975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946636615030192,
      "loss": 5.4792,
      "mean_token_accuracy": 0.1457243114709854,
      "num_tokens": 17376237.0,
      "step": 7930
    },
    {
      "entropy": 5.593193912506104,
      "epoch": 0.787514886859865,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004946559879120208,
      "loss": 5.4166,
      "mean_token_accuracy": 0.15014394968748093,
      "num_tokens": 17386188.0,
      "step": 7935
    },
    {
      "entropy": 5.449300241470337,
      "epoch": 0.7880111155220325,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004946483088740273,
      "loss": 5.4089,
      "mean_token_accuracy": 0.1506262131035328,
      "num_tokens": 17397058.0,
      "step": 7940
    },
    {
      "entropy": 5.610685873031616,
      "epoch": 0.7885073441842001,
      "grad_norm": 1.0,
      "learning_rate": 0.0004946406243892289,
      "loss": 5.4774,
      "mean_token_accuracy": 0.14903090745210648,
      "num_tokens": 17408191.0,
      "step": 7945
    },
    {
      "entropy": 5.607821130752564,
      "epoch": 0.7890035728463676,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004946329344578163,
      "loss": 5.4299,
      "mean_token_accuracy": 0.14606284201145173,
      "num_tokens": 17418472.0,
      "step": 7950
    },
    {
      "entropy": 5.518605899810791,
      "epoch": 0.7894998015085352,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004946252390799803,
      "loss": 5.3628,
      "mean_token_accuracy": 0.14880176782608032,
      "num_tokens": 17429556.0,
      "step": 7955
    },
    {
      "entropy": 5.548377943038941,
      "epoch": 0.7899960301707026,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004946175382559116,
      "loss": 5.5368,
      "mean_token_accuracy": 0.13931523784995078,
      "num_tokens": 17440950.0,
      "step": 7960
    },
    {
      "entropy": 5.533775711059571,
      "epoch": 0.7904922588328702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000494609831985801,
      "loss": 5.4136,
      "mean_token_accuracy": 0.1522986635565758,
      "num_tokens": 17453724.0,
      "step": 7965
    },
    {
      "entropy": 5.531882953643799,
      "epoch": 0.7909884874950377,
      "grad_norm": 1.125,
      "learning_rate": 0.00049460212026984,
      "loss": 5.4484,
      "mean_token_accuracy": 0.14882000163197517,
      "num_tokens": 17464623.0,
      "step": 7970
    },
    {
      "entropy": 5.482162141799927,
      "epoch": 0.7914847161572053,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004945944031082195,
      "loss": 5.4494,
      "mean_token_accuracy": 0.14701810628175735,
      "num_tokens": 17474919.0,
      "step": 7975
    },
    {
      "entropy": 5.639609909057617,
      "epoch": 0.7919809448193728,
      "grad_norm": 1.109375,
      "learning_rate": 0.000494586680501131,
      "loss": 5.4445,
      "mean_token_accuracy": 0.14752373993396758,
      "num_tokens": 17484747.0,
      "step": 7980
    },
    {
      "entropy": 5.5790173530578615,
      "epoch": 0.7924771734815403,
      "grad_norm": 1.03125,
      "learning_rate": 0.000494578952448766,
      "loss": 5.3792,
      "mean_token_accuracy": 0.1509535565972328,
      "num_tokens": 17496299.0,
      "step": 7985
    },
    {
      "entropy": 5.60280556678772,
      "epoch": 0.7929734021437078,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004945712189513161,
      "loss": 5.5326,
      "mean_token_accuracy": 0.14063167870044707,
      "num_tokens": 17507448.0,
      "step": 7990
    },
    {
      "entropy": 5.610171747207642,
      "epoch": 0.7934696308058753,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004945634800089733,
      "loss": 5.5165,
      "mean_token_accuracy": 0.14304168969392778,
      "num_tokens": 17520112.0,
      "step": 7995
    },
    {
      "entropy": 5.611541891098023,
      "epoch": 0.7939658594680429,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945557356219291,
      "loss": 5.5041,
      "mean_token_accuracy": 0.14014820531010627,
      "num_tokens": 17531368.0,
      "step": 8000
    },
    {
      "entropy": 5.533007669448852,
      "epoch": 0.7944620881302104,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004945479857903758,
      "loss": 5.4611,
      "mean_token_accuracy": 0.15101756602525712,
      "num_tokens": 17542082.0,
      "step": 8005
    },
    {
      "entropy": 5.657965850830078,
      "epoch": 0.7949583167923779,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004945402305145056,
      "loss": 5.482,
      "mean_token_accuracy": 0.1450247697532177,
      "num_tokens": 17555518.0,
      "step": 8010
    },
    {
      "entropy": 5.600556516647339,
      "epoch": 0.7954545454545454,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004945324697945107,
      "loss": 5.4497,
      "mean_token_accuracy": 0.15078138411045075,
      "num_tokens": 17566110.0,
      "step": 8015
    },
    {
      "entropy": 5.399105501174927,
      "epoch": 0.795950774116713,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004945247036305836,
      "loss": 5.3537,
      "mean_token_accuracy": 0.15356534123420715,
      "num_tokens": 17575942.0,
      "step": 8020
    },
    {
      "entropy": 5.520645666122436,
      "epoch": 0.7964470027788805,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004945169320229169,
      "loss": 5.4377,
      "mean_token_accuracy": 0.15009109377861024,
      "num_tokens": 17586001.0,
      "step": 8025
    },
    {
      "entropy": 5.6623146533966064,
      "epoch": 0.7969432314410481,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004945091549717033,
      "loss": 5.5948,
      "mean_token_accuracy": 0.13641601875424386,
      "num_tokens": 17596774.0,
      "step": 8030
    },
    {
      "entropy": 5.598223638534546,
      "epoch": 0.7974394601032155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945013724771356,
      "loss": 5.4549,
      "mean_token_accuracy": 0.15231607258319854,
      "num_tokens": 17606557.0,
      "step": 8035
    },
    {
      "entropy": 5.503456401824951,
      "epoch": 0.7979356887653831,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004944935845394069,
      "loss": 5.3334,
      "mean_token_accuracy": 0.15167354941368102,
      "num_tokens": 17616465.0,
      "step": 8040
    },
    {
      "entropy": 5.595588731765747,
      "epoch": 0.7984319174275506,
      "grad_norm": 1.125,
      "learning_rate": 0.0004944857911587103,
      "loss": 5.4866,
      "mean_token_accuracy": 0.14506380409002304,
      "num_tokens": 17626907.0,
      "step": 8045
    },
    {
      "entropy": 5.615129137039185,
      "epoch": 0.7989281460897182,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004944779923352389,
      "loss": 5.508,
      "mean_token_accuracy": 0.1475403293967247,
      "num_tokens": 17638660.0,
      "step": 8050
    },
    {
      "entropy": 5.5215531349182125,
      "epoch": 0.7994243747518857,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944701880691864,
      "loss": 5.4159,
      "mean_token_accuracy": 0.1503916338086128,
      "num_tokens": 17650135.0,
      "step": 8055
    },
    {
      "entropy": 5.509420347213745,
      "epoch": 0.7999206034140532,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000494462378360746,
      "loss": 5.445,
      "mean_token_accuracy": 0.15311212837696075,
      "num_tokens": 17660987.0,
      "step": 8060
    },
    {
      "entropy": 5.5887596130371096,
      "epoch": 0.8004168320762207,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004944545632101117,
      "loss": 5.3688,
      "mean_token_accuracy": 0.15472581386566162,
      "num_tokens": 17672520.0,
      "step": 8065
    },
    {
      "entropy": 5.577902507781983,
      "epoch": 0.8009130607383882,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944467426174769,
      "loss": 5.4422,
      "mean_token_accuracy": 0.15027025192975998,
      "num_tokens": 17683688.0,
      "step": 8070
    },
    {
      "entropy": 5.475872898101807,
      "epoch": 0.8014092894005558,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000494438916583036,
      "loss": 5.358,
      "mean_token_accuracy": 0.15467156320810319,
      "num_tokens": 17695213.0,
      "step": 8075
    },
    {
      "entropy": 5.578530216217041,
      "epoch": 0.8019055180627233,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004944310851069826,
      "loss": 5.444,
      "mean_token_accuracy": 0.14726410955190658,
      "num_tokens": 17705545.0,
      "step": 8080
    },
    {
      "entropy": 5.604553365707398,
      "epoch": 0.8024017467248908,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004944232481895112,
      "loss": 5.4355,
      "mean_token_accuracy": 0.15486004054546357,
      "num_tokens": 17715855.0,
      "step": 8085
    },
    {
      "entropy": 5.578699064254761,
      "epoch": 0.8028979753870583,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000494415405830816,
      "loss": 5.5134,
      "mean_token_accuracy": 0.14788013696670532,
      "num_tokens": 17727764.0,
      "step": 8090
    },
    {
      "entropy": 5.54158763885498,
      "epoch": 0.8033942040492259,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004944075580310916,
      "loss": 5.5002,
      "mean_token_accuracy": 0.14260387569665908,
      "num_tokens": 17738464.0,
      "step": 8095
    },
    {
      "entropy": 5.630135869979858,
      "epoch": 0.8038904327113934,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004943997047905325,
      "loss": 5.4749,
      "mean_token_accuracy": 0.15533988922834396,
      "num_tokens": 17747502.0,
      "step": 8100
    },
    {
      "entropy": 5.568667459487915,
      "epoch": 0.804386661373561,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004943918461093335,
      "loss": 5.4752,
      "mean_token_accuracy": 0.14455884397029878,
      "num_tokens": 17759520.0,
      "step": 8105
    },
    {
      "entropy": 5.560026597976685,
      "epoch": 0.8048828900357284,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004943839819876894,
      "loss": 5.4391,
      "mean_token_accuracy": 0.14611605256795884,
      "num_tokens": 17770631.0,
      "step": 8110
    },
    {
      "entropy": 5.607076644897461,
      "epoch": 0.805379118697896,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004943761124257954,
      "loss": 5.4041,
      "mean_token_accuracy": 0.1444598488509655,
      "num_tokens": 17781320.0,
      "step": 8115
    },
    {
      "entropy": 5.548258066177368,
      "epoch": 0.8058753473600635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004943682374238464,
      "loss": 5.462,
      "mean_token_accuracy": 0.14913394600152968,
      "num_tokens": 17792627.0,
      "step": 8120
    },
    {
      "entropy": 5.563149118423462,
      "epoch": 0.8063715760222311,
      "grad_norm": 1.015625,
      "learning_rate": 0.000494360356982038,
      "loss": 5.495,
      "mean_token_accuracy": 0.14413844794034958,
      "num_tokens": 17803286.0,
      "step": 8125
    },
    {
      "entropy": 5.513989353179932,
      "epoch": 0.8068678046843986,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004943524711005653,
      "loss": 5.3111,
      "mean_token_accuracy": 0.1494537889957428,
      "num_tokens": 17813631.0,
      "step": 8130
    },
    {
      "entropy": 5.513043069839478,
      "epoch": 0.8073640333465661,
      "grad_norm": 1.0625,
      "learning_rate": 0.000494344579779624,
      "loss": 5.4301,
      "mean_token_accuracy": 0.1515509933233261,
      "num_tokens": 17823467.0,
      "step": 8135
    },
    {
      "entropy": 5.525391912460327,
      "epoch": 0.8078602620087336,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943366830194098,
      "loss": 5.4188,
      "mean_token_accuracy": 0.15036700367927552,
      "num_tokens": 17833600.0,
      "step": 8140
    },
    {
      "entropy": 5.527675914764404,
      "epoch": 0.8083564906709012,
      "grad_norm": 1.125,
      "learning_rate": 0.0004943287808201183,
      "loss": 5.4048,
      "mean_token_accuracy": 0.1431363545358181,
      "num_tokens": 17843112.0,
      "step": 8145
    },
    {
      "entropy": 5.5455128192901615,
      "epoch": 0.8088527193330687,
      "grad_norm": 1.125,
      "learning_rate": 0.0004943208731819459,
      "loss": 5.414,
      "mean_token_accuracy": 0.14957288801670074,
      "num_tokens": 17853678.0,
      "step": 8150
    },
    {
      "entropy": 5.4899797439575195,
      "epoch": 0.8093489479952362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004943129601050883,
      "loss": 5.3936,
      "mean_token_accuracy": 0.14986039325594902,
      "num_tokens": 17863537.0,
      "step": 8155
    },
    {
      "entropy": 5.540888833999634,
      "epoch": 0.8098451766574037,
      "grad_norm": 1.03125,
      "learning_rate": 0.000494305041589742,
      "loss": 5.4799,
      "mean_token_accuracy": 0.14610693603754044,
      "num_tokens": 17873511.0,
      "step": 8160
    },
    {
      "entropy": 5.5646607875823975,
      "epoch": 0.8103414053195712,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004942971176361031,
      "loss": 5.4061,
      "mean_token_accuracy": 0.1530952051281929,
      "num_tokens": 17884546.0,
      "step": 8165
    },
    {
      "entropy": 5.490582704544067,
      "epoch": 0.8108376339817388,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004942891882443682,
      "loss": 5.3607,
      "mean_token_accuracy": 0.15139609575271606,
      "num_tokens": 17895331.0,
      "step": 8170
    },
    {
      "entropy": 5.558597183227539,
      "epoch": 0.8113338626439063,
      "grad_norm": 1.078125,
      "learning_rate": 0.000494281253414734,
      "loss": 5.5344,
      "mean_token_accuracy": 0.14457477480173112,
      "num_tokens": 17906465.0,
      "step": 8175
    },
    {
      "entropy": 5.538271236419678,
      "epoch": 0.8118300913060739,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004942733131473973,
      "loss": 5.4422,
      "mean_token_accuracy": 0.14445995688438415,
      "num_tokens": 17916936.0,
      "step": 8180
    },
    {
      "entropy": 5.60152416229248,
      "epoch": 0.8123263199682413,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004942653674425548,
      "loss": 5.474,
      "mean_token_accuracy": 0.1432388760149479,
      "num_tokens": 17927522.0,
      "step": 8185
    },
    {
      "entropy": 5.6055480480194095,
      "epoch": 0.8128225486304089,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004942574163004038,
      "loss": 5.5087,
      "mean_token_accuracy": 0.14854711368680001,
      "num_tokens": 17938216.0,
      "step": 8190
    },
    {
      "entropy": 5.545066881179809,
      "epoch": 0.8133187772925764,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004942494597211412,
      "loss": 5.4187,
      "mean_token_accuracy": 0.14836825728416442,
      "num_tokens": 17949333.0,
      "step": 8195
    },
    {
      "entropy": 5.59882345199585,
      "epoch": 0.813815005954744,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004942414977049646,
      "loss": 5.4861,
      "mean_token_accuracy": 0.1397691398859024,
      "num_tokens": 17960178.0,
      "step": 8200
    },
    {
      "entropy": 5.547542381286621,
      "epoch": 0.8143112346169115,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000494233530252071,
      "loss": 5.4412,
      "mean_token_accuracy": 0.1462123915553093,
      "num_tokens": 17971180.0,
      "step": 8205
    },
    {
      "entropy": 5.540498208999634,
      "epoch": 0.814807463279079,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004942255573626585,
      "loss": 5.4745,
      "mean_token_accuracy": 0.14544690623879433,
      "num_tokens": 17982572.0,
      "step": 8210
    },
    {
      "entropy": 5.569422864913941,
      "epoch": 0.8153036919412465,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004942175790369244,
      "loss": 5.41,
      "mean_token_accuracy": 0.1544869288802147,
      "num_tokens": 17993236.0,
      "step": 8215
    },
    {
      "entropy": 5.539626598358154,
      "epoch": 0.8157999206034141,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004942095952750669,
      "loss": 5.5114,
      "mean_token_accuracy": 0.14735156148672104,
      "num_tokens": 18004215.0,
      "step": 8220
    },
    {
      "entropy": 5.522426128387451,
      "epoch": 0.8162961492655816,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004942016060772837,
      "loss": 5.4079,
      "mean_token_accuracy": 0.15141918361186982,
      "num_tokens": 18013880.0,
      "step": 8225
    },
    {
      "entropy": 5.610029172897339,
      "epoch": 0.8167923779277491,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004941936114437731,
      "loss": 5.4755,
      "mean_token_accuracy": 0.14427937865257262,
      "num_tokens": 18024928.0,
      "step": 8230
    },
    {
      "entropy": 5.573900365829468,
      "epoch": 0.8172886065899166,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004941856113747331,
      "loss": 5.4296,
      "mean_token_accuracy": 0.14206683039665222,
      "num_tokens": 18035857.0,
      "step": 8235
    },
    {
      "entropy": 5.625568866729736,
      "epoch": 0.8177848352520841,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004941776058703623,
      "loss": 5.5384,
      "mean_token_accuracy": 0.14434613585472106,
      "num_tokens": 18046776.0,
      "step": 8240
    },
    {
      "entropy": 5.554650688171387,
      "epoch": 0.8182810639142517,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004941695949308593,
      "loss": 5.484,
      "mean_token_accuracy": 0.1481780603528023,
      "num_tokens": 18057167.0,
      "step": 8245
    },
    {
      "entropy": 5.558407306671143,
      "epoch": 0.8187772925764192,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004941615785564224,
      "loss": 5.5073,
      "mean_token_accuracy": 0.14432247206568719,
      "num_tokens": 18066953.0,
      "step": 8250
    },
    {
      "entropy": 5.544755315780639,
      "epoch": 0.8192735212385868,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941535567472506,
      "loss": 5.3298,
      "mean_token_accuracy": 0.15345962345600128,
      "num_tokens": 18077977.0,
      "step": 8255
    },
    {
      "entropy": 5.529454135894776,
      "epoch": 0.8197697499007542,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000494145529503543,
      "loss": 5.4451,
      "mean_token_accuracy": 0.1482371523976326,
      "num_tokens": 18087581.0,
      "step": 8260
    },
    {
      "entropy": 5.450969552993774,
      "epoch": 0.8202659785629218,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004941374968254985,
      "loss": 5.3424,
      "mean_token_accuracy": 0.1521090164780617,
      "num_tokens": 18098193.0,
      "step": 8265
    },
    {
      "entropy": 5.487133264541626,
      "epoch": 0.8207622072250893,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941294587133161,
      "loss": 5.456,
      "mean_token_accuracy": 0.14906689673662185,
      "num_tokens": 18110830.0,
      "step": 8270
    },
    {
      "entropy": 5.648927116394043,
      "epoch": 0.8212584358872569,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004941214151671954,
      "loss": 5.48,
      "mean_token_accuracy": 0.14420426189899443,
      "num_tokens": 18122073.0,
      "step": 8275
    },
    {
      "entropy": 5.533260726928711,
      "epoch": 0.8217546645494244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004941133661873357,
      "loss": 5.4611,
      "mean_token_accuracy": 0.15090519785881043,
      "num_tokens": 18132777.0,
      "step": 8280
    },
    {
      "entropy": 5.482033395767212,
      "epoch": 0.822250893211592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004941053117739367,
      "loss": 5.3754,
      "mean_token_accuracy": 0.15093337446451188,
      "num_tokens": 18143232.0,
      "step": 8285
    },
    {
      "entropy": 5.55641393661499,
      "epoch": 0.8227471218737594,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004940972519271982,
      "loss": 5.5498,
      "mean_token_accuracy": 0.14168024063110352,
      "num_tokens": 18155225.0,
      "step": 8290
    },
    {
      "entropy": 5.60756664276123,
      "epoch": 0.823243350535927,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004940891866473199,
      "loss": 5.468,
      "mean_token_accuracy": 0.14743804112076758,
      "num_tokens": 18165945.0,
      "step": 8295
    },
    {
      "entropy": 5.633839225769043,
      "epoch": 0.8237395791980945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004940811159345017,
      "loss": 5.5081,
      "mean_token_accuracy": 0.1480460211634636,
      "num_tokens": 18177271.0,
      "step": 8300
    },
    {
      "entropy": 5.62118353843689,
      "epoch": 0.824235807860262,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004940730397889441,
      "loss": 5.4551,
      "mean_token_accuracy": 0.14795344471931457,
      "num_tokens": 18189732.0,
      "step": 8305
    },
    {
      "entropy": 5.536730480194092,
      "epoch": 0.8247320365224295,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004940649582108472,
      "loss": 5.5457,
      "mean_token_accuracy": 0.14492482617497443,
      "num_tokens": 18200731.0,
      "step": 8310
    },
    {
      "entropy": 5.602692079544068,
      "epoch": 0.825228265184597,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004940568712004113,
      "loss": 5.4623,
      "mean_token_accuracy": 0.1492995612323284,
      "num_tokens": 18210873.0,
      "step": 8315
    },
    {
      "entropy": 5.643162059783935,
      "epoch": 0.8257244938467646,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004940487787578369,
      "loss": 5.459,
      "mean_token_accuracy": 0.15006872564554213,
      "num_tokens": 18222072.0,
      "step": 8320
    },
    {
      "entropy": 5.421371603012085,
      "epoch": 0.8262207225089321,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000494040680883325,
      "loss": 5.3608,
      "mean_token_accuracy": 0.14893197864294053,
      "num_tokens": 18232522.0,
      "step": 8325
    },
    {
      "entropy": 5.58615026473999,
      "epoch": 0.8267169511710997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004940325775770762,
      "loss": 5.4666,
      "mean_token_accuracy": 0.14552596360445022,
      "num_tokens": 18243587.0,
      "step": 8330
    },
    {
      "entropy": 5.544474792480469,
      "epoch": 0.8272131798332671,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004940244688392913,
      "loss": 5.3893,
      "mean_token_accuracy": 0.15559020787477493,
      "num_tokens": 18254965.0,
      "step": 8335
    },
    {
      "entropy": 5.505191802978516,
      "epoch": 0.8277094084954347,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004940163546701716,
      "loss": 5.3937,
      "mean_token_accuracy": 0.14967342764139174,
      "num_tokens": 18265892.0,
      "step": 8340
    },
    {
      "entropy": 5.535961580276489,
      "epoch": 0.8282056371576022,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004940082350699183,
      "loss": 5.3955,
      "mean_token_accuracy": 0.14893627017736435,
      "num_tokens": 18276713.0,
      "step": 8345
    },
    {
      "entropy": 5.600809907913208,
      "epoch": 0.8287018658197698,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004940001100387325,
      "loss": 5.449,
      "mean_token_accuracy": 0.14580243080854416,
      "num_tokens": 18288846.0,
      "step": 8350
    },
    {
      "entropy": 5.591676425933838,
      "epoch": 0.8291980944819373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939919795768161,
      "loss": 5.4672,
      "mean_token_accuracy": 0.15377946943044662,
      "num_tokens": 18298750.0,
      "step": 8355
    },
    {
      "entropy": 5.560976791381836,
      "epoch": 0.8296943231441049,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004939838436843704,
      "loss": 5.4597,
      "mean_token_accuracy": 0.14587485194206237,
      "num_tokens": 18310065.0,
      "step": 8360
    },
    {
      "entropy": 5.538149452209472,
      "epoch": 0.8301905518062723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004939757023615973,
      "loss": 5.3948,
      "mean_token_accuracy": 0.15021784156560897,
      "num_tokens": 18319498.0,
      "step": 8365
    },
    {
      "entropy": 5.491366386413574,
      "epoch": 0.8306867804684399,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004939675556086986,
      "loss": 5.3804,
      "mean_token_accuracy": 0.15116882249712943,
      "num_tokens": 18329718.0,
      "step": 8370
    },
    {
      "entropy": 5.485607576370239,
      "epoch": 0.8311830091306074,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004939594034258763,
      "loss": 5.3923,
      "mean_token_accuracy": 0.15384181141853331,
      "num_tokens": 18341628.0,
      "step": 8375
    },
    {
      "entropy": 5.571638584136963,
      "epoch": 0.8316792377927749,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004939512458133326,
      "loss": 5.4393,
      "mean_token_accuracy": 0.15046600103378296,
      "num_tokens": 18352869.0,
      "step": 8380
    },
    {
      "entropy": 5.531062889099121,
      "epoch": 0.8321754664549424,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004939430827712698,
      "loss": 5.4336,
      "mean_token_accuracy": 0.14440379217267035,
      "num_tokens": 18363664.0,
      "step": 8385
    },
    {
      "entropy": 5.564842891693115,
      "epoch": 0.8326716951171099,
      "grad_norm": 1.125,
      "learning_rate": 0.0004939349142998905,
      "loss": 5.4226,
      "mean_token_accuracy": 0.15019726902246475,
      "num_tokens": 18374325.0,
      "step": 8390
    },
    {
      "entropy": 5.538015413284302,
      "epoch": 0.8331679237792775,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004939267403993968,
      "loss": 5.4177,
      "mean_token_accuracy": 0.14942388236522675,
      "num_tokens": 18384547.0,
      "step": 8395
    },
    {
      "entropy": 5.562836790084839,
      "epoch": 0.833664152441445,
      "grad_norm": 1.171875,
      "learning_rate": 0.000493918561069992,
      "loss": 5.3841,
      "mean_token_accuracy": 0.14861270934343337,
      "num_tokens": 18395141.0,
      "step": 8400
    },
    {
      "entropy": 5.4846254825592045,
      "epoch": 0.8341603811036126,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004939103763118784,
      "loss": 5.3903,
      "mean_token_accuracy": 0.1516875520348549,
      "num_tokens": 18406326.0,
      "step": 8405
    },
    {
      "entropy": 5.5294005393981935,
      "epoch": 0.83465660976578,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004939021861252593,
      "loss": 5.4656,
      "mean_token_accuracy": 0.1521238550543785,
      "num_tokens": 18417521.0,
      "step": 8410
    },
    {
      "entropy": 5.567513704299927,
      "epoch": 0.8351528384279476,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004938939905103376,
      "loss": 5.3844,
      "mean_token_accuracy": 0.15346331894397736,
      "num_tokens": 18426813.0,
      "step": 8415
    },
    {
      "entropy": 5.502229976654053,
      "epoch": 0.8356490670901151,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004938857894673166,
      "loss": 5.3438,
      "mean_token_accuracy": 0.15055184662342072,
      "num_tokens": 18438227.0,
      "step": 8420
    },
    {
      "entropy": 5.529934406280518,
      "epoch": 0.8361452957522827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004938775829963997,
      "loss": 5.4945,
      "mean_token_accuracy": 0.1485433906316757,
      "num_tokens": 18448915.0,
      "step": 8425
    },
    {
      "entropy": 5.551042270660401,
      "epoch": 0.8366415244144502,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004938693710977904,
      "loss": 5.3622,
      "mean_token_accuracy": 0.14919409155845642,
      "num_tokens": 18459330.0,
      "step": 8430
    },
    {
      "entropy": 5.526158666610717,
      "epoch": 0.8371377530766178,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004938611537716922,
      "loss": 5.3448,
      "mean_token_accuracy": 0.15527479201555253,
      "num_tokens": 18469813.0,
      "step": 8435
    },
    {
      "entropy": 5.520946979522705,
      "epoch": 0.8376339817387852,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004938529310183091,
      "loss": 5.417,
      "mean_token_accuracy": 0.14860693961381913,
      "num_tokens": 18479783.0,
      "step": 8440
    },
    {
      "entropy": 5.555679607391357,
      "epoch": 0.8381302104009528,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000493844702837845,
      "loss": 5.4246,
      "mean_token_accuracy": 0.15056385099887848,
      "num_tokens": 18489640.0,
      "step": 8445
    },
    {
      "entropy": 5.556673383712768,
      "epoch": 0.8386264390631203,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004938364692305036,
      "loss": 5.4921,
      "mean_token_accuracy": 0.15282512456178665,
      "num_tokens": 18500744.0,
      "step": 8450
    },
    {
      "entropy": 5.538465404510498,
      "epoch": 0.8391226677252878,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004938282301964896,
      "loss": 5.3799,
      "mean_token_accuracy": 0.14792532324790955,
      "num_tokens": 18513286.0,
      "step": 8455
    },
    {
      "entropy": 5.553688097000122,
      "epoch": 0.8396188963874553,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004938199857360068,
      "loss": 5.3636,
      "mean_token_accuracy": 0.15171041935682297,
      "num_tokens": 18522934.0,
      "step": 8460
    },
    {
      "entropy": 5.469200897216797,
      "epoch": 0.8401151250496228,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00049381173584926,
      "loss": 5.4211,
      "mean_token_accuracy": 0.1483452409505844,
      "num_tokens": 18534828.0,
      "step": 8465
    },
    {
      "entropy": 5.461701250076294,
      "epoch": 0.8406113537117904,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004938034805364535,
      "loss": 5.3716,
      "mean_token_accuracy": 0.1469104364514351,
      "num_tokens": 18544733.0,
      "step": 8470
    },
    {
      "entropy": 5.531709671020508,
      "epoch": 0.8411075823739579,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004937952197977923,
      "loss": 5.432,
      "mean_token_accuracy": 0.1486746460199356,
      "num_tokens": 18554583.0,
      "step": 8475
    },
    {
      "entropy": 5.655738592147827,
      "epoch": 0.8416038110361255,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000493786953633481,
      "loss": 5.4825,
      "mean_token_accuracy": 0.14451623558998108,
      "num_tokens": 18566503.0,
      "step": 8480
    },
    {
      "entropy": 5.538340473175049,
      "epoch": 0.842100039698293,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004937786820437248,
      "loss": 5.4325,
      "mean_token_accuracy": 0.14823201596736907,
      "num_tokens": 18578161.0,
      "step": 8485
    },
    {
      "entropy": 5.512860107421875,
      "epoch": 0.8425962683604605,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004937704050287286,
      "loss": 5.4887,
      "mean_token_accuracy": 0.1507240816950798,
      "num_tokens": 18588212.0,
      "step": 8490
    },
    {
      "entropy": 5.668630456924438,
      "epoch": 0.843092497022628,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004937621225886978,
      "loss": 5.4753,
      "mean_token_accuracy": 0.1439543306827545,
      "num_tokens": 18601293.0,
      "step": 8495
    },
    {
      "entropy": 5.561993741989136,
      "epoch": 0.8435887256847956,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004937538347238378,
      "loss": 5.5101,
      "mean_token_accuracy": 0.14477006569504738,
      "num_tokens": 18613316.0,
      "step": 8500
    },
    {
      "entropy": 5.535253572463989,
      "epoch": 0.8440849543469631,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000493745541434354,
      "loss": 5.4571,
      "mean_token_accuracy": 0.15068159103393555,
      "num_tokens": 18624461.0,
      "step": 8505
    },
    {
      "entropy": 5.583186674118042,
      "epoch": 0.8445811830091307,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004937372427204522,
      "loss": 5.3459,
      "mean_token_accuracy": 0.14902222603559495,
      "num_tokens": 18635165.0,
      "step": 8510
    },
    {
      "entropy": 5.550057601928711,
      "epoch": 0.8450774116712981,
      "grad_norm": 1.125,
      "learning_rate": 0.0004937289385823382,
      "loss": 5.4118,
      "mean_token_accuracy": 0.15501163080334662,
      "num_tokens": 18646649.0,
      "step": 8515
    },
    {
      "entropy": 5.476098823547363,
      "epoch": 0.8455736403334657,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004937206290202176,
      "loss": 5.4042,
      "mean_token_accuracy": 0.14524563401937485,
      "num_tokens": 18658527.0,
      "step": 8520
    },
    {
      "entropy": 5.548634004592896,
      "epoch": 0.8460698689956332,
      "grad_norm": 1.0,
      "learning_rate": 0.0004937123140342969,
      "loss": 5.4049,
      "mean_token_accuracy": 0.15190646648406983,
      "num_tokens": 18669463.0,
      "step": 8525
    },
    {
      "entropy": 5.588420677185058,
      "epoch": 0.8465660976578007,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000493703993624782,
      "loss": 5.4448,
      "mean_token_accuracy": 0.14451150447130204,
      "num_tokens": 18680806.0,
      "step": 8530
    },
    {
      "entropy": 5.4969908714294435,
      "epoch": 0.8470623263199683,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004936956677918794,
      "loss": 5.4485,
      "mean_token_accuracy": 0.14826684445142746,
      "num_tokens": 18692124.0,
      "step": 8535
    },
    {
      "entropy": 5.618858194351196,
      "epoch": 0.8475585549821357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004936873365357955,
      "loss": 5.4497,
      "mean_token_accuracy": 0.14569347053766252,
      "num_tokens": 18702017.0,
      "step": 8540
    },
    {
      "entropy": 5.498226022720337,
      "epoch": 0.8480547836443033,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004936789998567368,
      "loss": 5.3563,
      "mean_token_accuracy": 0.15395329743623734,
      "num_tokens": 18713205.0,
      "step": 8545
    },
    {
      "entropy": 5.49776701927185,
      "epoch": 0.8485510123064708,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004936706577549101,
      "loss": 5.3811,
      "mean_token_accuracy": 0.1490063972771168,
      "num_tokens": 18724165.0,
      "step": 8550
    },
    {
      "entropy": 5.514372205734253,
      "epoch": 0.8490472409686384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004936623102305223,
      "loss": 5.4561,
      "mean_token_accuracy": 0.1490431733429432,
      "num_tokens": 18735155.0,
      "step": 8555
    },
    {
      "entropy": 5.580429792404175,
      "epoch": 0.8495434696308058,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004936539572837804,
      "loss": 5.5243,
      "mean_token_accuracy": 0.14651782661676407,
      "num_tokens": 18746228.0,
      "step": 8560
    },
    {
      "entropy": 5.611409330368042,
      "epoch": 0.8500396982929734,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004936455989148916,
      "loss": 5.4662,
      "mean_token_accuracy": 0.14882544428110123,
      "num_tokens": 18756638.0,
      "step": 8565
    },
    {
      "entropy": 5.561304521560669,
      "epoch": 0.8505359269551409,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004936372351240629,
      "loss": 5.4428,
      "mean_token_accuracy": 0.14990992248058319,
      "num_tokens": 18768651.0,
      "step": 8570
    },
    {
      "entropy": 5.621553897857666,
      "epoch": 0.8510321556173085,
      "grad_norm": 1.03125,
      "learning_rate": 0.000493628865911502,
      "loss": 5.5393,
      "mean_token_accuracy": 0.14673913568258284,
      "num_tokens": 18780399.0,
      "step": 8575
    },
    {
      "entropy": 5.553244686126709,
      "epoch": 0.851528384279476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004936204912774163,
      "loss": 5.4138,
      "mean_token_accuracy": 0.14724221527576448,
      "num_tokens": 18791927.0,
      "step": 8580
    },
    {
      "entropy": 5.493992280960083,
      "epoch": 0.8520246129416436,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004936121112220135,
      "loss": 5.3857,
      "mean_token_accuracy": 0.14751749187707902,
      "num_tokens": 18802779.0,
      "step": 8585
    },
    {
      "entropy": 5.479309463500977,
      "epoch": 0.852520841603811,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004936037257455015,
      "loss": 5.39,
      "mean_token_accuracy": 0.1515676721930504,
      "num_tokens": 18813123.0,
      "step": 8590
    },
    {
      "entropy": 5.507645225524902,
      "epoch": 0.8530170702659786,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000493595334848088,
      "loss": 5.3851,
      "mean_token_accuracy": 0.15856402814388276,
      "num_tokens": 18823699.0,
      "step": 8595
    },
    {
      "entropy": 5.579003858566284,
      "epoch": 0.8535132989281461,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004935869385299813,
      "loss": 5.5581,
      "mean_token_accuracy": 0.14485259503126144,
      "num_tokens": 18835450.0,
      "step": 8600
    },
    {
      "entropy": 5.55187497138977,
      "epoch": 0.8540095275903136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004935785367913896,
      "loss": 5.3994,
      "mean_token_accuracy": 0.14939774870872496,
      "num_tokens": 18845919.0,
      "step": 8605
    },
    {
      "entropy": 5.613993549346924,
      "epoch": 0.8545057562524812,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004935701296325212,
      "loss": 5.5112,
      "mean_token_accuracy": 0.13831136450171472,
      "num_tokens": 18858376.0,
      "step": 8610
    },
    {
      "entropy": 5.609997844696045,
      "epoch": 0.8550019849146486,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004935617170535845,
      "loss": 5.4613,
      "mean_token_accuracy": 0.15094567239284515,
      "num_tokens": 18870456.0,
      "step": 8615
    },
    {
      "entropy": 5.5243676662445065,
      "epoch": 0.8554982135768162,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004935532990547882,
      "loss": 5.4714,
      "mean_token_accuracy": 0.14436968713998793,
      "num_tokens": 18882174.0,
      "step": 8620
    },
    {
      "entropy": 5.624684238433838,
      "epoch": 0.8559944422389837,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004935448756363409,
      "loss": 5.6001,
      "mean_token_accuracy": 0.1479609876871109,
      "num_tokens": 18892982.0,
      "step": 8625
    },
    {
      "entropy": 5.632325172424316,
      "epoch": 0.8564906709011513,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935364467984518,
      "loss": 5.418,
      "mean_token_accuracy": 0.15484932214021682,
      "num_tokens": 18903557.0,
      "step": 8630
    },
    {
      "entropy": 5.547096920013428,
      "epoch": 0.8569868995633187,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004935280125413295,
      "loss": 5.5524,
      "mean_token_accuracy": 0.14480863362550736,
      "num_tokens": 18914912.0,
      "step": 8635
    },
    {
      "entropy": 5.605364561080933,
      "epoch": 0.8574831282254863,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004935195728651836,
      "loss": 5.4847,
      "mean_token_accuracy": 0.14869843870401384,
      "num_tokens": 18927773.0,
      "step": 8640
    },
    {
      "entropy": 5.586603116989136,
      "epoch": 0.8579793568876538,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000493511127770223,
      "loss": 5.443,
      "mean_token_accuracy": 0.14573638141155243,
      "num_tokens": 18938313.0,
      "step": 8645
    },
    {
      "entropy": 5.536235809326172,
      "epoch": 0.8584755855498214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004935026772566574,
      "loss": 5.379,
      "mean_token_accuracy": 0.1523163437843323,
      "num_tokens": 18948600.0,
      "step": 8650
    },
    {
      "entropy": 5.589066839218139,
      "epoch": 0.8589718142119889,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004934942213246961,
      "loss": 5.4766,
      "mean_token_accuracy": 0.14389901608228683,
      "num_tokens": 18958927.0,
      "step": 8655
    },
    {
      "entropy": 5.551258134841919,
      "epoch": 0.8594680428741565,
      "grad_norm": 1.09375,
      "learning_rate": 0.000493485759974549,
      "loss": 5.416,
      "mean_token_accuracy": 0.14265228509902955,
      "num_tokens": 18968625.0,
      "step": 8660
    },
    {
      "entropy": 5.498787069320679,
      "epoch": 0.8599642715363239,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004934772932064259,
      "loss": 5.3609,
      "mean_token_accuracy": 0.150932240486145,
      "num_tokens": 18979772.0,
      "step": 8665
    },
    {
      "entropy": 5.584731817245483,
      "epoch": 0.8604605001984915,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004934688210205367,
      "loss": 5.506,
      "mean_token_accuracy": 0.1481923058629036,
      "num_tokens": 18991522.0,
      "step": 8670
    },
    {
      "entropy": 5.566020727157593,
      "epoch": 0.860956728860659,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934603434170913,
      "loss": 5.4574,
      "mean_token_accuracy": 0.14935235530138016,
      "num_tokens": 19002479.0,
      "step": 8675
    },
    {
      "entropy": 5.547944974899292,
      "epoch": 0.8614529575228265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004934518603963003,
      "loss": 5.5266,
      "mean_token_accuracy": 0.14217539951205255,
      "num_tokens": 19013254.0,
      "step": 8680
    },
    {
      "entropy": 5.501922273635865,
      "epoch": 0.861949186184994,
      "grad_norm": 1.0,
      "learning_rate": 0.0004934433719583739,
      "loss": 5.3586,
      "mean_token_accuracy": 0.1478867307305336,
      "num_tokens": 19023943.0,
      "step": 8685
    },
    {
      "entropy": 5.630067253112793,
      "epoch": 0.8624454148471615,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004934348781035226,
      "loss": 5.4878,
      "mean_token_accuracy": 0.14363348931074144,
      "num_tokens": 19034672.0,
      "step": 8690
    },
    {
      "entropy": 5.632747936248779,
      "epoch": 0.8629416435093291,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000493426378831957,
      "loss": 5.4712,
      "mean_token_accuracy": 0.15527579188346863,
      "num_tokens": 19045758.0,
      "step": 8695
    },
    {
      "entropy": 5.522514915466308,
      "epoch": 0.8634378721714966,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934178741438877,
      "loss": 5.4708,
      "mean_token_accuracy": 0.15148718208074569,
      "num_tokens": 19057526.0,
      "step": 8700
    },
    {
      "entropy": 5.467182874679565,
      "epoch": 0.8639341008336642,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000493409364039526,
      "loss": 5.3355,
      "mean_token_accuracy": 0.15133479982614517,
      "num_tokens": 19069657.0,
      "step": 8705
    },
    {
      "entropy": 5.51442699432373,
      "epoch": 0.8644303294958317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004934008485190825,
      "loss": 5.3855,
      "mean_token_accuracy": 0.15153009742498397,
      "num_tokens": 19080699.0,
      "step": 8710
    },
    {
      "entropy": 5.479568958282471,
      "epoch": 0.8649265581579992,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004933923275827687,
      "loss": 5.4161,
      "mean_token_accuracy": 0.14561865329742432,
      "num_tokens": 19093041.0,
      "step": 8715
    },
    {
      "entropy": 5.652049875259399,
      "epoch": 0.8654227868201667,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004933838012307957,
      "loss": 5.4983,
      "mean_token_accuracy": 0.1425291620194912,
      "num_tokens": 19104249.0,
      "step": 8720
    },
    {
      "entropy": 5.601692247390747,
      "epoch": 0.8659190154823343,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000493375269463375,
      "loss": 5.4375,
      "mean_token_accuracy": 0.14599128514528276,
      "num_tokens": 19114434.0,
      "step": 8725
    },
    {
      "entropy": 5.496259164810181,
      "epoch": 0.8664152441445018,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004933667322807182,
      "loss": 5.4524,
      "mean_token_accuracy": 0.14293356165289878,
      "num_tokens": 19124814.0,
      "step": 8730
    },
    {
      "entropy": 5.521295166015625,
      "epoch": 0.8669114728066694,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000493358189683037,
      "loss": 5.4685,
      "mean_token_accuracy": 0.14728949218988419,
      "num_tokens": 19135205.0,
      "step": 8735
    },
    {
      "entropy": 5.547979879379272,
      "epoch": 0.8674077014688368,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004933496416705432,
      "loss": 5.3441,
      "mean_token_accuracy": 0.1517267793416977,
      "num_tokens": 19147117.0,
      "step": 8740
    },
    {
      "entropy": 5.522016906738282,
      "epoch": 0.8679039301310044,
      "grad_norm": 1.0,
      "learning_rate": 0.0004933410882434488,
      "loss": 5.443,
      "mean_token_accuracy": 0.14760973304510117,
      "num_tokens": 19158131.0,
      "step": 8745
    },
    {
      "entropy": 5.483796739578247,
      "epoch": 0.8684001587931719,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004933325294019659,
      "loss": 5.3251,
      "mean_token_accuracy": 0.1540597304701805,
      "num_tokens": 19168820.0,
      "step": 8750
    },
    {
      "entropy": 5.587935829162598,
      "epoch": 0.8688963874553394,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004933239651463068,
      "loss": 5.4299,
      "mean_token_accuracy": 0.1531436026096344,
      "num_tokens": 19178845.0,
      "step": 8755
    },
    {
      "entropy": 5.589181232452392,
      "epoch": 0.869392616117507,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004933153954766837,
      "loss": 5.4062,
      "mean_token_accuracy": 0.1507592022418976,
      "num_tokens": 19190830.0,
      "step": 8760
    },
    {
      "entropy": 5.366233873367309,
      "epoch": 0.8698888447796744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004933068203933092,
      "loss": 5.2498,
      "mean_token_accuracy": 0.15559230297803878,
      "num_tokens": 19202056.0,
      "step": 8765
    },
    {
      "entropy": 5.492927980422974,
      "epoch": 0.870385073441842,
      "grad_norm": 1.03125,
      "learning_rate": 0.000493298239896396,
      "loss": 5.4584,
      "mean_token_accuracy": 0.1478209227323532,
      "num_tokens": 19213497.0,
      "step": 8770
    },
    {
      "entropy": 5.603393507003784,
      "epoch": 0.8708813021040095,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004932896539861568,
      "loss": 5.5068,
      "mean_token_accuracy": 0.14628409892320632,
      "num_tokens": 19226103.0,
      "step": 8775
    },
    {
      "entropy": 5.534812068939209,
      "epoch": 0.8713775307661771,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004932810626628047,
      "loss": 5.4965,
      "mean_token_accuracy": 0.149713883548975,
      "num_tokens": 19237297.0,
      "step": 8780
    },
    {
      "entropy": 5.5414466381073,
      "epoch": 0.8718737594283446,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004932724659265526,
      "loss": 5.4175,
      "mean_token_accuracy": 0.14820330440998078,
      "num_tokens": 19247537.0,
      "step": 8785
    },
    {
      "entropy": 5.4871766567230225,
      "epoch": 0.8723699880905121,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004932638637776134,
      "loss": 5.3593,
      "mean_token_accuracy": 0.15012865215539933,
      "num_tokens": 19260465.0,
      "step": 8790
    },
    {
      "entropy": 5.522654867172241,
      "epoch": 0.8728662167526796,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000493255256216201,
      "loss": 5.4207,
      "mean_token_accuracy": 0.14851754009723664,
      "num_tokens": 19270534.0,
      "step": 8795
    },
    {
      "entropy": 5.483912324905395,
      "epoch": 0.8733624454148472,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004932466432425284,
      "loss": 5.4491,
      "mean_token_accuracy": 0.14936281442642213,
      "num_tokens": 19281551.0,
      "step": 8800
    },
    {
      "entropy": 5.4793956756591795,
      "epoch": 0.8738586740770147,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004932380248568094,
      "loss": 5.374,
      "mean_token_accuracy": 0.15332597643136978,
      "num_tokens": 19293300.0,
      "step": 8805
    },
    {
      "entropy": 5.516641759872437,
      "epoch": 0.8743549027391823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004932294010592575,
      "loss": 5.318,
      "mean_token_accuracy": 0.15324191898107528,
      "num_tokens": 19302701.0,
      "step": 8810
    },
    {
      "entropy": 5.5574263572692875,
      "epoch": 0.8748511314013497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004932207718500868,
      "loss": 5.4092,
      "mean_token_accuracy": 0.14890351220965387,
      "num_tokens": 19313024.0,
      "step": 8815
    },
    {
      "entropy": 5.552362871170044,
      "epoch": 0.8753473600635173,
      "grad_norm": 1.125,
      "learning_rate": 0.0004932121372295112,
      "loss": 5.4116,
      "mean_token_accuracy": 0.14468025863170625,
      "num_tokens": 19324014.0,
      "step": 8820
    },
    {
      "entropy": 5.508990812301636,
      "epoch": 0.8758435887256848,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004932034971977448,
      "loss": 5.4406,
      "mean_token_accuracy": 0.14813138619065286,
      "num_tokens": 19334226.0,
      "step": 8825
    },
    {
      "entropy": 5.638391065597534,
      "epoch": 0.8763398173878523,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004931948517550019,
      "loss": 5.4889,
      "mean_token_accuracy": 0.14102088436484336,
      "num_tokens": 19344363.0,
      "step": 8830
    },
    {
      "entropy": 5.5545642375946045,
      "epoch": 0.8768360460500199,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004931862009014968,
      "loss": 5.4892,
      "mean_token_accuracy": 0.14179774522781372,
      "num_tokens": 19354915.0,
      "step": 8835
    },
    {
      "entropy": 5.5872047424316404,
      "epoch": 0.8773322747121873,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493177544637444,
      "loss": 5.4947,
      "mean_token_accuracy": 0.14904706478118895,
      "num_tokens": 19365570.0,
      "step": 8840
    },
    {
      "entropy": 5.6439306259155275,
      "epoch": 0.8778285033743549,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004931688829630582,
      "loss": 5.4294,
      "mean_token_accuracy": 0.1553065836429596,
      "num_tokens": 19375731.0,
      "step": 8845
    },
    {
      "entropy": 5.49164400100708,
      "epoch": 0.8783247320365224,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004931602158785544,
      "loss": 5.3687,
      "mean_token_accuracy": 0.14904445856809617,
      "num_tokens": 19386248.0,
      "step": 8850
    },
    {
      "entropy": 5.498191785812378,
      "epoch": 0.87882096069869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004931515433841472,
      "loss": 5.3676,
      "mean_token_accuracy": 0.15414228290319443,
      "num_tokens": 19397593.0,
      "step": 8855
    },
    {
      "entropy": 5.539998388290405,
      "epoch": 0.8793171893608575,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004931428654800518,
      "loss": 5.4645,
      "mean_token_accuracy": 0.14690039679408073,
      "num_tokens": 19409701.0,
      "step": 8860
    },
    {
      "entropy": 5.5156224250793455,
      "epoch": 0.879813418023025,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931341821664834,
      "loss": 5.4468,
      "mean_token_accuracy": 0.1461575746536255,
      "num_tokens": 19420280.0,
      "step": 8865
    },
    {
      "entropy": 5.524348640441895,
      "epoch": 0.8803096466851925,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004931254934436573,
      "loss": 5.4251,
      "mean_token_accuracy": 0.14768008142709732,
      "num_tokens": 19432208.0,
      "step": 8870
    },
    {
      "entropy": 5.623254108428955,
      "epoch": 0.8808058753473601,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004931167993117891,
      "loss": 5.545,
      "mean_token_accuracy": 0.14630000442266464,
      "num_tokens": 19443333.0,
      "step": 8875
    },
    {
      "entropy": 5.4848377227783205,
      "epoch": 0.8813021040095276,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004931080997710942,
      "loss": 5.2968,
      "mean_token_accuracy": 0.15597349405288696,
      "num_tokens": 19452867.0,
      "step": 8880
    },
    {
      "entropy": 5.423505592346191,
      "epoch": 0.8817983326716952,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930993948217884,
      "loss": 5.3324,
      "mean_token_accuracy": 0.16214994639158248,
      "num_tokens": 19463021.0,
      "step": 8885
    },
    {
      "entropy": 5.5309906005859375,
      "epoch": 0.8822945613338626,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930906844640876,
      "loss": 5.4688,
      "mean_token_accuracy": 0.14767196476459504,
      "num_tokens": 19474092.0,
      "step": 8890
    },
    {
      "entropy": 5.549347925186157,
      "epoch": 0.8827907899960302,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930819686982077,
      "loss": 5.4396,
      "mean_token_accuracy": 0.14400598481297494,
      "num_tokens": 19486035.0,
      "step": 8895
    },
    {
      "entropy": 5.529491281509399,
      "epoch": 0.8832870186581977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000493073247524365,
      "loss": 5.375,
      "mean_token_accuracy": 0.15078936517238617,
      "num_tokens": 19497484.0,
      "step": 8900
    },
    {
      "entropy": 5.550891494750976,
      "epoch": 0.8837832473203652,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004930645209427756,
      "loss": 5.4853,
      "mean_token_accuracy": 0.1420919269323349,
      "num_tokens": 19508463.0,
      "step": 8905
    },
    {
      "entropy": 5.499340581893921,
      "epoch": 0.8842794759825328,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000493055788953656,
      "loss": 5.4013,
      "mean_token_accuracy": 0.14937035143375396,
      "num_tokens": 19520260.0,
      "step": 8910
    },
    {
      "entropy": 5.538986587524414,
      "epoch": 0.8847757046447002,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004930470515572227,
      "loss": 5.3684,
      "mean_token_accuracy": 0.15278342366218567,
      "num_tokens": 19531621.0,
      "step": 8915
    },
    {
      "entropy": 5.54758882522583,
      "epoch": 0.8852719333068678,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004930383087536923,
      "loss": 5.3413,
      "mean_token_accuracy": 0.1496839180588722,
      "num_tokens": 19543271.0,
      "step": 8920
    },
    {
      "entropy": 5.558869028091431,
      "epoch": 0.8857681619690353,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004930295605432818,
      "loss": 5.4994,
      "mean_token_accuracy": 0.1523902475833893,
      "num_tokens": 19555178.0,
      "step": 8925
    },
    {
      "entropy": 5.5412476539611815,
      "epoch": 0.8862643906312029,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004930208069262079,
      "loss": 5.4347,
      "mean_token_accuracy": 0.15091503113508226,
      "num_tokens": 19565153.0,
      "step": 8930
    },
    {
      "entropy": 5.5265477180480955,
      "epoch": 0.8867606192933704,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004930120479026878,
      "loss": 5.3863,
      "mean_token_accuracy": 0.1530556008219719,
      "num_tokens": 19575608.0,
      "step": 8935
    },
    {
      "entropy": 5.5148841381073,
      "epoch": 0.8872568479555379,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004930032834729388,
      "loss": 5.3348,
      "mean_token_accuracy": 0.15864698439836503,
      "num_tokens": 19587556.0,
      "step": 8940
    },
    {
      "entropy": 5.4669452667236325,
      "epoch": 0.8877530766177054,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000492994513637178,
      "loss": 5.3775,
      "mean_token_accuracy": 0.14984949976205825,
      "num_tokens": 19597341.0,
      "step": 8945
    },
    {
      "entropy": 5.52333025932312,
      "epoch": 0.888249305279873,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929857383956229,
      "loss": 5.4213,
      "mean_token_accuracy": 0.14854597449302673,
      "num_tokens": 19608431.0,
      "step": 8950
    },
    {
      "entropy": 5.5654700756072994,
      "epoch": 0.8887455339420405,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004929769577484914,
      "loss": 5.3908,
      "mean_token_accuracy": 0.15749148428440093,
      "num_tokens": 19619645.0,
      "step": 8955
    },
    {
      "entropy": 5.4893574714660645,
      "epoch": 0.8892417626042081,
      "grad_norm": 1.1875,
      "learning_rate": 0.000492968171696001,
      "loss": 5.3473,
      "mean_token_accuracy": 0.15694154500961305,
      "num_tokens": 19630412.0,
      "step": 8960
    },
    {
      "entropy": 5.554796934127808,
      "epoch": 0.8897379912663755,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004929593802383696,
      "loss": 5.429,
      "mean_token_accuracy": 0.15158904939889908,
      "num_tokens": 19640055.0,
      "step": 8965
    },
    {
      "entropy": 5.505509281158448,
      "epoch": 0.8902342199285431,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004929505833758152,
      "loss": 5.4152,
      "mean_token_accuracy": 0.1454661712050438,
      "num_tokens": 19650814.0,
      "step": 8970
    },
    {
      "entropy": 5.5592933177948,
      "epoch": 0.8907304485907106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929417811085561,
      "loss": 5.4048,
      "mean_token_accuracy": 0.14785738065838813,
      "num_tokens": 19663669.0,
      "step": 8975
    },
    {
      "entropy": 5.5864850044250485,
      "epoch": 0.8912266772528781,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004929329734368103,
      "loss": 5.4903,
      "mean_token_accuracy": 0.1435614138841629,
      "num_tokens": 19675851.0,
      "step": 8980
    },
    {
      "entropy": 5.525371980667114,
      "epoch": 0.8917229059150457,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929241603607964,
      "loss": 5.3707,
      "mean_token_accuracy": 0.1562560260295868,
      "num_tokens": 19686613.0,
      "step": 8985
    },
    {
      "entropy": 5.493849658966065,
      "epoch": 0.8922191345772131,
      "grad_norm": 1.125,
      "learning_rate": 0.000492915341880733,
      "loss": 5.311,
      "mean_token_accuracy": 0.15805860012769699,
      "num_tokens": 19696419.0,
      "step": 8990
    },
    {
      "entropy": 5.502732229232788,
      "epoch": 0.8927153632393807,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929065179968387,
      "loss": 5.4009,
      "mean_token_accuracy": 0.14759214073419571,
      "num_tokens": 19707269.0,
      "step": 8995
    },
    {
      "entropy": 5.614929151535034,
      "epoch": 0.8932115919015482,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004928976887093321,
      "loss": 5.4544,
      "mean_token_accuracy": 0.15342584997415543,
      "num_tokens": 19716099.0,
      "step": 9000
    },
    {
      "epoch": 0.8932115919015482,
      "eval_entropy": 5.442668267229895,
      "eval_loss": 5.497566223144531,
      "eval_mean_token_accuracy": 0.15083033391923936,
      "eval_num_tokens": 19716099.0,
      "eval_runtime": 26.7173,
      "eval_samples_per_second": 1260.195,
      "eval_steps_per_second": 157.538,
      "step": 9000
    },
    {
      "entropy": 5.530213499069214,
      "epoch": 0.8937078205637158,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004928888540184325,
      "loss": 5.4381,
      "mean_token_accuracy": 0.1489228054881096,
      "num_tokens": 19726632.0,
      "step": 9005
    },
    {
      "entropy": 5.4784809112548825,
      "epoch": 0.8942040492258833,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004928800139243587,
      "loss": 5.341,
      "mean_token_accuracy": 0.15205781459808348,
      "num_tokens": 19737929.0,
      "step": 9010
    },
    {
      "entropy": 5.498447561264038,
      "epoch": 0.8947002778880508,
      "grad_norm": 1.125,
      "learning_rate": 0.0004928711684273302,
      "loss": 5.3677,
      "mean_token_accuracy": 0.1516488678753376,
      "num_tokens": 19748191.0,
      "step": 9015
    },
    {
      "entropy": 5.5268003940582275,
      "epoch": 0.8951965065502183,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928623175275661,
      "loss": 5.4165,
      "mean_token_accuracy": 0.15140914618968965,
      "num_tokens": 19758892.0,
      "step": 9020
    },
    {
      "entropy": 5.49551305770874,
      "epoch": 0.8956927352123859,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928534612252862,
      "loss": 5.3054,
      "mean_token_accuracy": 0.15685733109712602,
      "num_tokens": 19769120.0,
      "step": 9025
    },
    {
      "entropy": 5.448373079299927,
      "epoch": 0.8961889638745534,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004928445995207095,
      "loss": 5.3799,
      "mean_token_accuracy": 0.15353552550077437,
      "num_tokens": 19779484.0,
      "step": 9030
    },
    {
      "entropy": 5.54969630241394,
      "epoch": 0.896685192536721,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004928357324140564,
      "loss": 5.4036,
      "mean_token_accuracy": 0.14872636198997496,
      "num_tokens": 19790438.0,
      "step": 9035
    },
    {
      "entropy": 5.6278149604797365,
      "epoch": 0.8971814211988884,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004928268599055465,
      "loss": 5.4902,
      "mean_token_accuracy": 0.15090776085853577,
      "num_tokens": 19801646.0,
      "step": 9040
    },
    {
      "entropy": 5.514270210266114,
      "epoch": 0.897677649861056,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004928179819953998,
      "loss": 5.358,
      "mean_token_accuracy": 0.15458858758211136,
      "num_tokens": 19812887.0,
      "step": 9045
    },
    {
      "entropy": 5.494898128509521,
      "epoch": 0.8981738785232235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004928090986838366,
      "loss": 5.4372,
      "mean_token_accuracy": 0.1474384568631649,
      "num_tokens": 19822938.0,
      "step": 9050
    },
    {
      "entropy": 5.500031280517578,
      "epoch": 0.898670107185391,
      "grad_norm": 1.15625,
      "learning_rate": 0.000492800209971077,
      "loss": 5.3612,
      "mean_token_accuracy": 0.15316907465457916,
      "num_tokens": 19833988.0,
      "step": 9055
    },
    {
      "entropy": 5.594955253601074,
      "epoch": 0.8991663358475586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004927913158573416,
      "loss": 5.4408,
      "mean_token_accuracy": 0.14832141995429993,
      "num_tokens": 19845112.0,
      "step": 9060
    },
    {
      "entropy": 5.497157335281372,
      "epoch": 0.899662564509726,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004927824163428508,
      "loss": 5.3994,
      "mean_token_accuracy": 0.1488441988825798,
      "num_tokens": 19856304.0,
      "step": 9065
    },
    {
      "entropy": 5.501319169998169,
      "epoch": 0.9001587931718936,
      "grad_norm": 1.125,
      "learning_rate": 0.0004927735114278254,
      "loss": 5.3937,
      "mean_token_accuracy": 0.15133703500032425,
      "num_tokens": 19866776.0,
      "step": 9070
    },
    {
      "entropy": 5.550291299819946,
      "epoch": 0.9006550218340611,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004927646011124861,
      "loss": 5.3767,
      "mean_token_accuracy": 0.152619419246912,
      "num_tokens": 19876392.0,
      "step": 9075
    },
    {
      "entropy": 5.511345624923706,
      "epoch": 0.9011512504962287,
      "grad_norm": 1.0,
      "learning_rate": 0.0004927556853970539,
      "loss": 5.4129,
      "mean_token_accuracy": 0.1482156328856945,
      "num_tokens": 19887362.0,
      "step": 9080
    },
    {
      "entropy": 5.523296880722046,
      "epoch": 0.9016474791583962,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004927467642817499,
      "loss": 5.3885,
      "mean_token_accuracy": 0.1500864863395691,
      "num_tokens": 19898989.0,
      "step": 9085
    },
    {
      "entropy": 5.550860595703125,
      "epoch": 0.9021437078205637,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004927378377667955,
      "loss": 5.3955,
      "mean_token_accuracy": 0.14326826184988023,
      "num_tokens": 19910112.0,
      "step": 9090
    },
    {
      "entropy": 5.579243946075439,
      "epoch": 0.9026399364827312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492728905852412,
      "loss": 5.5027,
      "mean_token_accuracy": 0.15027060359716415,
      "num_tokens": 19922601.0,
      "step": 9095
    },
    {
      "entropy": 5.518284749984741,
      "epoch": 0.9031361651448988,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004927199685388205,
      "loss": 5.4169,
      "mean_token_accuracy": 0.15244396179914474,
      "num_tokens": 19934153.0,
      "step": 9100
    },
    {
      "entropy": 5.498987197875977,
      "epoch": 0.9036323938070663,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000492711025826243,
      "loss": 5.3785,
      "mean_token_accuracy": 0.14863183945417405,
      "num_tokens": 19944572.0,
      "step": 9105
    },
    {
      "entropy": 5.549972820281982,
      "epoch": 0.9041286224692339,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004927020777149014,
      "loss": 5.4934,
      "mean_token_accuracy": 0.14485350027680396,
      "num_tokens": 19956431.0,
      "step": 9110
    },
    {
      "entropy": 5.617206478118897,
      "epoch": 0.9046248511314013,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926931242050171,
      "loss": 5.4132,
      "mean_token_accuracy": 0.14434127807617186,
      "num_tokens": 19968118.0,
      "step": 9115
    },
    {
      "entropy": 5.558873844146729,
      "epoch": 0.9051210797935689,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926841652968125,
      "loss": 5.4883,
      "mean_token_accuracy": 0.14938305988907813,
      "num_tokens": 19979545.0,
      "step": 9120
    },
    {
      "entropy": 5.530987071990967,
      "epoch": 0.9056173084557364,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926752009905099,
      "loss": 5.4394,
      "mean_token_accuracy": 0.1413062706589699,
      "num_tokens": 19989832.0,
      "step": 9125
    },
    {
      "entropy": 5.537513017654419,
      "epoch": 0.9061135371179039,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926662312863311,
      "loss": 5.4288,
      "mean_token_accuracy": 0.14608306661248208,
      "num_tokens": 20000935.0,
      "step": 9130
    },
    {
      "entropy": 5.546253633499146,
      "epoch": 0.9066097657800715,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926572561844988,
      "loss": 5.446,
      "mean_token_accuracy": 0.14424679577350616,
      "num_tokens": 20011485.0,
      "step": 9135
    },
    {
      "entropy": 5.525794839859008,
      "epoch": 0.9071059944422389,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004926482756852356,
      "loss": 5.3088,
      "mean_token_accuracy": 0.15703367590904235,
      "num_tokens": 20023043.0,
      "step": 9140
    },
    {
      "entropy": 5.531298303604126,
      "epoch": 0.9076022231044065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926392897887642,
      "loss": 5.4202,
      "mean_token_accuracy": 0.157535183429718,
      "num_tokens": 20032437.0,
      "step": 9145
    },
    {
      "entropy": 5.519662761688233,
      "epoch": 0.908098451766574,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004926302984953073,
      "loss": 5.4348,
      "mean_token_accuracy": 0.14996563792228698,
      "num_tokens": 20044662.0,
      "step": 9150
    },
    {
      "entropy": 5.5613710403442385,
      "epoch": 0.9085946804287416,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004926213018050881,
      "loss": 5.4508,
      "mean_token_accuracy": 0.1501474305987358,
      "num_tokens": 20055924.0,
      "step": 9155
    },
    {
      "entropy": 5.53719425201416,
      "epoch": 0.9090909090909091,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004926122997183295,
      "loss": 5.4121,
      "mean_token_accuracy": 0.1464460700750351,
      "num_tokens": 20067909.0,
      "step": 9160
    },
    {
      "entropy": 5.490392017364502,
      "epoch": 0.9095871377530766,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004926032922352546,
      "loss": 5.3711,
      "mean_token_accuracy": 0.15813741236925125,
      "num_tokens": 20079333.0,
      "step": 9165
    },
    {
      "entropy": 5.517123651504517,
      "epoch": 0.9100833664152441,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004925942793560872,
      "loss": 5.3588,
      "mean_token_accuracy": 0.15189230740070342,
      "num_tokens": 20089741.0,
      "step": 9170
    },
    {
      "entropy": 5.464686012268066,
      "epoch": 0.9105795950774117,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004925852610810504,
      "loss": 5.3808,
      "mean_token_accuracy": 0.1502847895026207,
      "num_tokens": 20100267.0,
      "step": 9175
    },
    {
      "entropy": 5.486724615097046,
      "epoch": 0.9110758237395792,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000492576237410368,
      "loss": 5.3617,
      "mean_token_accuracy": 0.15151302367448807,
      "num_tokens": 20111964.0,
      "step": 9180
    },
    {
      "entropy": 5.571880292892456,
      "epoch": 0.9115720524017468,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925672083442638,
      "loss": 5.4247,
      "mean_token_accuracy": 0.14389239102602006,
      "num_tokens": 20123132.0,
      "step": 9185
    },
    {
      "entropy": 5.474943161010742,
      "epoch": 0.9120682810639142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004925581738829616,
      "loss": 5.3392,
      "mean_token_accuracy": 0.1548833131790161,
      "num_tokens": 20134050.0,
      "step": 9190
    },
    {
      "entropy": 5.517318487167358,
      "epoch": 0.9125645097260818,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004925491340266855,
      "loss": 5.4008,
      "mean_token_accuracy": 0.15220728218555452,
      "num_tokens": 20143814.0,
      "step": 9195
    },
    {
      "entropy": 5.585405540466309,
      "epoch": 0.9130607383882493,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004925400887756596,
      "loss": 5.48,
      "mean_token_accuracy": 0.1517788365483284,
      "num_tokens": 20157647.0,
      "step": 9200
    },
    {
      "entropy": 5.555648326873779,
      "epoch": 0.9135569670504168,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004925310381301083,
      "loss": 5.4215,
      "mean_token_accuracy": 0.15375025421380997,
      "num_tokens": 20168366.0,
      "step": 9205
    },
    {
      "entropy": 5.536394023895264,
      "epoch": 0.9140531957125844,
      "grad_norm": 1.078125,
      "learning_rate": 0.000492521982090256,
      "loss": 5.482,
      "mean_token_accuracy": 0.1524073101580143,
      "num_tokens": 20179362.0,
      "step": 9210
    },
    {
      "entropy": 5.4954001903533936,
      "epoch": 0.9145494243747518,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004925129206563273,
      "loss": 5.379,
      "mean_token_accuracy": 0.1546468809247017,
      "num_tokens": 20190320.0,
      "step": 9215
    },
    {
      "entropy": 5.4641612529754635,
      "epoch": 0.9150456530369194,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004925038538285467,
      "loss": 5.3904,
      "mean_token_accuracy": 0.1536310449242592,
      "num_tokens": 20201114.0,
      "step": 9220
    },
    {
      "entropy": 5.562612676620484,
      "epoch": 0.9155418816990869,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004924947816071394,
      "loss": 5.4108,
      "mean_token_accuracy": 0.15487478226423262,
      "num_tokens": 20213351.0,
      "step": 9225
    },
    {
      "entropy": 5.508360433578491,
      "epoch": 0.9160381103612545,
      "grad_norm": 1.109375,
      "learning_rate": 0.00049248570399233,
      "loss": 5.3237,
      "mean_token_accuracy": 0.1595936208963394,
      "num_tokens": 20223376.0,
      "step": 9230
    },
    {
      "entropy": 5.540471506118775,
      "epoch": 0.916534339023422,
      "grad_norm": 1.125,
      "learning_rate": 0.000492476620984344,
      "loss": 5.4423,
      "mean_token_accuracy": 0.15038547962903975,
      "num_tokens": 20234248.0,
      "step": 9235
    },
    {
      "entropy": 5.514523983001709,
      "epoch": 0.9170305676855895,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004924675325834063,
      "loss": 5.3793,
      "mean_token_accuracy": 0.14954861700534822,
      "num_tokens": 20245823.0,
      "step": 9240
    },
    {
      "entropy": 5.5835862159729,
      "epoch": 0.917526796347757,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924584387897424,
      "loss": 5.4365,
      "mean_token_accuracy": 0.14587546288967132,
      "num_tokens": 20257003.0,
      "step": 9245
    },
    {
      "entropy": 5.582525539398193,
      "epoch": 0.9180230250099246,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004924493396035777,
      "loss": 5.5636,
      "mean_token_accuracy": 0.14269972890615462,
      "num_tokens": 20268500.0,
      "step": 9250
    },
    {
      "entropy": 5.616550254821777,
      "epoch": 0.9185192536720921,
      "grad_norm": 1.140625,
      "learning_rate": 0.000492440235025138,
      "loss": 5.4353,
      "mean_token_accuracy": 0.1557466819882393,
      "num_tokens": 20279301.0,
      "step": 9255
    },
    {
      "entropy": 5.6475830078125,
      "epoch": 0.9190154823342597,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004924311250546491,
      "loss": 5.5342,
      "mean_token_accuracy": 0.139700785279274,
      "num_tokens": 20291263.0,
      "step": 9260
    },
    {
      "entropy": 5.519128894805908,
      "epoch": 0.9195117109964271,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004924220096923367,
      "loss": 5.4232,
      "mean_token_accuracy": 0.15304265469312667,
      "num_tokens": 20301554.0,
      "step": 9265
    },
    {
      "entropy": 5.620229291915893,
      "epoch": 0.9200079396585947,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004924128889384271,
      "loss": 5.4509,
      "mean_token_accuracy": 0.14695485085248947,
      "num_tokens": 20314299.0,
      "step": 9270
    },
    {
      "entropy": 5.49082350730896,
      "epoch": 0.9205041683207622,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004924037627931463,
      "loss": 5.3983,
      "mean_token_accuracy": 0.15196871608495713,
      "num_tokens": 20324595.0,
      "step": 9275
    },
    {
      "entropy": 5.502113914489746,
      "epoch": 0.9210003969829297,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004923946312567206,
      "loss": 5.3664,
      "mean_token_accuracy": 0.1510197177529335,
      "num_tokens": 20335744.0,
      "step": 9280
    },
    {
      "entropy": 5.473414039611816,
      "epoch": 0.9214966256450973,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004923854943293765,
      "loss": 5.2908,
      "mean_token_accuracy": 0.14862959533929826,
      "num_tokens": 20346072.0,
      "step": 9285
    },
    {
      "entropy": 5.524072456359863,
      "epoch": 0.9219928543072647,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004923763520113406,
      "loss": 5.3972,
      "mean_token_accuracy": 0.15875740647315978,
      "num_tokens": 20357747.0,
      "step": 9290
    },
    {
      "entropy": 5.5446373462677006,
      "epoch": 0.9224890829694323,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923672043028395,
      "loss": 5.3528,
      "mean_token_accuracy": 0.1558915928006172,
      "num_tokens": 20368838.0,
      "step": 9295
    },
    {
      "entropy": 5.505260276794433,
      "epoch": 0.9229853116315998,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004923580512041001,
      "loss": 5.4416,
      "mean_token_accuracy": 0.14829889237880706,
      "num_tokens": 20380844.0,
      "step": 9300
    },
    {
      "entropy": 5.570078039169312,
      "epoch": 0.9234815402937674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004923488927153495,
      "loss": 5.4426,
      "mean_token_accuracy": 0.14666374549269676,
      "num_tokens": 20391739.0,
      "step": 9305
    },
    {
      "entropy": 5.551726818084717,
      "epoch": 0.9239777689559349,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004923397288368146,
      "loss": 5.3383,
      "mean_token_accuracy": 0.159624320268631,
      "num_tokens": 20402571.0,
      "step": 9310
    },
    {
      "entropy": 5.491204929351807,
      "epoch": 0.9244739976181024,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004923305595687229,
      "loss": 5.3668,
      "mean_token_accuracy": 0.14744364395737647,
      "num_tokens": 20414766.0,
      "step": 9315
    },
    {
      "entropy": 5.491946268081665,
      "epoch": 0.9249702262802699,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004923213849113015,
      "loss": 5.334,
      "mean_token_accuracy": 0.1534505680203438,
      "num_tokens": 20425508.0,
      "step": 9320
    },
    {
      "entropy": 5.480807495117188,
      "epoch": 0.9254664549424375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004923122048647781,
      "loss": 5.3332,
      "mean_token_accuracy": 0.1511707827448845,
      "num_tokens": 20436128.0,
      "step": 9325
    },
    {
      "entropy": 5.487059688568115,
      "epoch": 0.925962683604605,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004923030194293802,
      "loss": 5.4375,
      "mean_token_accuracy": 0.15299820601940156,
      "num_tokens": 20447212.0,
      "step": 9330
    },
    {
      "entropy": 5.526946401596069,
      "epoch": 0.9264589122667726,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004922938286053358,
      "loss": 5.3501,
      "mean_token_accuracy": 0.15629509836435318,
      "num_tokens": 20457770.0,
      "step": 9335
    },
    {
      "entropy": 5.5582598686218265,
      "epoch": 0.92695514092894,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004922846323928726,
      "loss": 5.4313,
      "mean_token_accuracy": 0.15007508993148805,
      "num_tokens": 20468172.0,
      "step": 9340
    },
    {
      "entropy": 5.475312852859497,
      "epoch": 0.9274513695911076,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922754307922188,
      "loss": 5.3527,
      "mean_token_accuracy": 0.147872294485569,
      "num_tokens": 20479429.0,
      "step": 9345
    },
    {
      "entropy": 5.500398874282837,
      "epoch": 0.9279475982532751,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004922662238036024,
      "loss": 5.3265,
      "mean_token_accuracy": 0.15738919079303743,
      "num_tokens": 20490373.0,
      "step": 9350
    },
    {
      "entropy": 5.604583597183227,
      "epoch": 0.9284438269154426,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004922570114272518,
      "loss": 5.4542,
      "mean_token_accuracy": 0.14771705791354178,
      "num_tokens": 20500002.0,
      "step": 9355
    },
    {
      "entropy": 5.564875555038452,
      "epoch": 0.9289400555776102,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004922477936633955,
      "loss": 5.5036,
      "mean_token_accuracy": 0.14229383319616318,
      "num_tokens": 20512990.0,
      "step": 9360
    },
    {
      "entropy": 5.479275560379028,
      "epoch": 0.9294362842397776,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004922385705122621,
      "loss": 5.4217,
      "mean_token_accuracy": 0.15395166873931884,
      "num_tokens": 20522979.0,
      "step": 9365
    },
    {
      "entropy": 5.47132568359375,
      "epoch": 0.9299325129019452,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004922293419740802,
      "loss": 5.2973,
      "mean_token_accuracy": 0.15944019705057144,
      "num_tokens": 20534620.0,
      "step": 9370
    },
    {
      "entropy": 5.581718254089355,
      "epoch": 0.9304287415641127,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004922201080490787,
      "loss": 5.3918,
      "mean_token_accuracy": 0.14692213982343674,
      "num_tokens": 20545526.0,
      "step": 9375
    },
    {
      "entropy": 5.590382146835327,
      "epoch": 0.9309249702262803,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004922108687374867,
      "loss": 5.5592,
      "mean_token_accuracy": 0.14881354868412017,
      "num_tokens": 20556665.0,
      "step": 9380
    },
    {
      "entropy": 5.522130298614502,
      "epoch": 0.9314211988884478,
      "grad_norm": 1.0625,
      "learning_rate": 0.000492201624039533,
      "loss": 5.3979,
      "mean_token_accuracy": 0.1455645315349102,
      "num_tokens": 20566054.0,
      "step": 9385
    },
    {
      "entropy": 5.652860927581787,
      "epoch": 0.9319174275506154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004921923739554472,
      "loss": 5.494,
      "mean_token_accuracy": 0.1452071711421013,
      "num_tokens": 20577922.0,
      "step": 9390
    },
    {
      "entropy": 5.492084217071533,
      "epoch": 0.9324136562127828,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004921831184854585,
      "loss": 5.2691,
      "mean_token_accuracy": 0.1598840534687042,
      "num_tokens": 20588473.0,
      "step": 9395
    },
    {
      "entropy": 5.462988328933716,
      "epoch": 0.9329098848749504,
      "grad_norm": 1.125,
      "learning_rate": 0.0004921738576297965,
      "loss": 5.4701,
      "mean_token_accuracy": 0.1499202385544777,
      "num_tokens": 20599371.0,
      "step": 9400
    },
    {
      "entropy": 5.490965461730957,
      "epoch": 0.9334061135371179,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004921645913886907,
      "loss": 5.3559,
      "mean_token_accuracy": 0.14845976382493972,
      "num_tokens": 20610066.0,
      "step": 9405
    },
    {
      "entropy": 5.55334095954895,
      "epoch": 0.9339023421992855,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004921553197623711,
      "loss": 5.444,
      "mean_token_accuracy": 0.14659321457147598,
      "num_tokens": 20621804.0,
      "step": 9410
    },
    {
      "entropy": 5.541829586029053,
      "epoch": 0.934398570861453,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004921460427510674,
      "loss": 5.3303,
      "mean_token_accuracy": 0.1599414810538292,
      "num_tokens": 20632706.0,
      "step": 9415
    },
    {
      "entropy": 5.47169976234436,
      "epoch": 0.9348947995236205,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004921367603550098,
      "loss": 5.4321,
      "mean_token_accuracy": 0.14436158835887908,
      "num_tokens": 20643303.0,
      "step": 9420
    },
    {
      "entropy": 5.5641711235046385,
      "epoch": 0.935391028185788,
      "grad_norm": 1.125,
      "learning_rate": 0.0004921274725744283,
      "loss": 5.4047,
      "mean_token_accuracy": 0.15554872825741767,
      "num_tokens": 20653681.0,
      "step": 9425
    },
    {
      "entropy": 5.566141033172608,
      "epoch": 0.9358872568479555,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004921181794095536,
      "loss": 5.4137,
      "mean_token_accuracy": 0.15083324909210205,
      "num_tokens": 20665599.0,
      "step": 9430
    },
    {
      "entropy": 5.455755043029785,
      "epoch": 0.9363834855101231,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004921088808606157,
      "loss": 5.3072,
      "mean_token_accuracy": 0.15324898362159728,
      "num_tokens": 20675726.0,
      "step": 9435
    },
    {
      "entropy": 5.579277181625367,
      "epoch": 0.9368797141722905,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004920995769278454,
      "loss": 5.4768,
      "mean_token_accuracy": 0.147647275775671,
      "num_tokens": 20688083.0,
      "step": 9440
    },
    {
      "entropy": 5.579367542266846,
      "epoch": 0.9373759428344581,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004920902676114734,
      "loss": 5.3537,
      "mean_token_accuracy": 0.15420555770397187,
      "num_tokens": 20697642.0,
      "step": 9445
    },
    {
      "entropy": 5.55647988319397,
      "epoch": 0.9378721714966256,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004920809529117306,
      "loss": 5.4565,
      "mean_token_accuracy": 0.14341557919979095,
      "num_tokens": 20709172.0,
      "step": 9450
    },
    {
      "entropy": 5.605994462966919,
      "epoch": 0.9383684001587932,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004920716328288479,
      "loss": 5.4935,
      "mean_token_accuracy": 0.14637328386306764,
      "num_tokens": 20720563.0,
      "step": 9455
    },
    {
      "entropy": 5.5216569900512695,
      "epoch": 0.9388646288209607,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004920623073630566,
      "loss": 5.3626,
      "mean_token_accuracy": 0.15472846627235412,
      "num_tokens": 20730943.0,
      "step": 9460
    },
    {
      "entropy": 5.513111972808838,
      "epoch": 0.9393608574831283,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004920529765145876,
      "loss": 5.3504,
      "mean_token_accuracy": 0.15353218391537665,
      "num_tokens": 20740903.0,
      "step": 9465
    },
    {
      "entropy": 5.535007476806641,
      "epoch": 0.9398570861452957,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004920436402836726,
      "loss": 5.4571,
      "mean_token_accuracy": 0.15176392048597337,
      "num_tokens": 20750580.0,
      "step": 9470
    },
    {
      "entropy": 5.566563415527344,
      "epoch": 0.9403533148074633,
      "grad_norm": 1.015625,
      "learning_rate": 0.000492034298670543,
      "loss": 5.3658,
      "mean_token_accuracy": 0.1528700664639473,
      "num_tokens": 20761357.0,
      "step": 9475
    },
    {
      "entropy": 5.599583673477173,
      "epoch": 0.9408495434696308,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004920249516754304,
      "loss": 5.5084,
      "mean_token_accuracy": 0.1437133640050888,
      "num_tokens": 20773273.0,
      "step": 9480
    },
    {
      "entropy": 5.548130893707276,
      "epoch": 0.9413457721317984,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004920155992985667,
      "loss": 5.3528,
      "mean_token_accuracy": 0.15052724331617356,
      "num_tokens": 20784290.0,
      "step": 9485
    },
    {
      "entropy": 5.553037118911743,
      "epoch": 0.9418420007939659,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004920062415401837,
      "loss": 5.3814,
      "mean_token_accuracy": 0.15294163674116135,
      "num_tokens": 20794764.0,
      "step": 9490
    },
    {
      "entropy": 5.544193172454834,
      "epoch": 0.9423382294561334,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004919968784005136,
      "loss": 5.4309,
      "mean_token_accuracy": 0.14622164592146875,
      "num_tokens": 20804863.0,
      "step": 9495
    },
    {
      "entropy": 5.576853036880493,
      "epoch": 0.9428344581183009,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004919875098797886,
      "loss": 5.5194,
      "mean_token_accuracy": 0.14691849574446678,
      "num_tokens": 20816445.0,
      "step": 9500
    },
    {
      "entropy": 5.564183378219605,
      "epoch": 0.9433306867804684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004919781359782407,
      "loss": 5.3301,
      "mean_token_accuracy": 0.15829974561929702,
      "num_tokens": 20827141.0,
      "step": 9505
    },
    {
      "entropy": 5.519220304489136,
      "epoch": 0.943826915442636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004919687566961027,
      "loss": 5.3862,
      "mean_token_accuracy": 0.14479556381702424,
      "num_tokens": 20838291.0,
      "step": 9510
    },
    {
      "entropy": 5.592441129684448,
      "epoch": 0.9443231441048034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000491959372033607,
      "loss": 5.3983,
      "mean_token_accuracy": 0.15185801386833192,
      "num_tokens": 20848839.0,
      "step": 9515
    },
    {
      "entropy": 5.517627477645874,
      "epoch": 0.944819372766971,
      "grad_norm": 1.0,
      "learning_rate": 0.0004919499819909865,
      "loss": 5.3658,
      "mean_token_accuracy": 0.15196501910686494,
      "num_tokens": 20859961.0,
      "step": 9520
    },
    {
      "entropy": 5.420031595230102,
      "epoch": 0.9453156014291385,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004919405865684738,
      "loss": 5.309,
      "mean_token_accuracy": 0.15580219477415086,
      "num_tokens": 20869720.0,
      "step": 9525
    },
    {
      "entropy": 5.5042869567871096,
      "epoch": 0.9458118300913061,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000491931185766302,
      "loss": 5.4377,
      "mean_token_accuracy": 0.15631229132413865,
      "num_tokens": 20880513.0,
      "step": 9530
    },
    {
      "entropy": 5.570223808288574,
      "epoch": 0.9463080587534736,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004919217795847044,
      "loss": 5.3715,
      "mean_token_accuracy": 0.15462664365768433,
      "num_tokens": 20892456.0,
      "step": 9535
    },
    {
      "entropy": 5.551194477081299,
      "epoch": 0.9468042874156412,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004919123680239139,
      "loss": 5.3776,
      "mean_token_accuracy": 0.15113687217235566,
      "num_tokens": 20904500.0,
      "step": 9540
    },
    {
      "entropy": 5.528022289276123,
      "epoch": 0.9473005160778086,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004919029510841641,
      "loss": 5.3638,
      "mean_token_accuracy": 0.15525981187820434,
      "num_tokens": 20915133.0,
      "step": 9545
    },
    {
      "entropy": 5.446129274368286,
      "epoch": 0.9477967447399762,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004918935287656887,
      "loss": 5.3487,
      "mean_token_accuracy": 0.15198146402835847,
      "num_tokens": 20924779.0,
      "step": 9550
    },
    {
      "entropy": 5.456863117218018,
      "epoch": 0.9482929734021437,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000491884101068721,
      "loss": 5.3106,
      "mean_token_accuracy": 0.15263781547546387,
      "num_tokens": 20935387.0,
      "step": 9555
    },
    {
      "entropy": 5.594128608703613,
      "epoch": 0.9487892020643113,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004918746679934949,
      "loss": 5.4731,
      "mean_token_accuracy": 0.15020606368780137,
      "num_tokens": 20945432.0,
      "step": 9560
    },
    {
      "entropy": 5.506171655654907,
      "epoch": 0.9492854307264788,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004918652295402444,
      "loss": 5.358,
      "mean_token_accuracy": 0.15571733936667442,
      "num_tokens": 20956777.0,
      "step": 9565
    },
    {
      "entropy": 5.544921588897705,
      "epoch": 0.9497816593886463,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004918557857092036,
      "loss": 5.383,
      "mean_token_accuracy": 0.15328644067049027,
      "num_tokens": 20966057.0,
      "step": 9570
    },
    {
      "entropy": 5.563687801361084,
      "epoch": 0.9502778880508138,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004918463365006065,
      "loss": 5.4892,
      "mean_token_accuracy": 0.1471877232193947,
      "num_tokens": 20976838.0,
      "step": 9575
    },
    {
      "entropy": 5.6165854930877686,
      "epoch": 0.9507741167129813,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004918368819146875,
      "loss": 5.41,
      "mean_token_accuracy": 0.15309347212314606,
      "num_tokens": 20988553.0,
      "step": 9580
    },
    {
      "entropy": 5.514295864105224,
      "epoch": 0.9512703453751489,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004918274219516811,
      "loss": 5.3968,
      "mean_token_accuracy": 0.15224193334579467,
      "num_tokens": 20998983.0,
      "step": 9585
    },
    {
      "entropy": 5.561841630935669,
      "epoch": 0.9517665740373163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004918179566118218,
      "loss": 5.4822,
      "mean_token_accuracy": 0.14829333275556564,
      "num_tokens": 21010008.0,
      "step": 9590
    },
    {
      "entropy": 5.537058687210083,
      "epoch": 0.9522628026994839,
      "grad_norm": 1.125,
      "learning_rate": 0.0004918084858953445,
      "loss": 5.4763,
      "mean_token_accuracy": 0.15138718038797377,
      "num_tokens": 21022826.0,
      "step": 9595
    },
    {
      "entropy": 5.582567930221558,
      "epoch": 0.9527590313616514,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004917990098024839,
      "loss": 5.3933,
      "mean_token_accuracy": 0.15087516903877257,
      "num_tokens": 21033457.0,
      "step": 9600
    },
    {
      "entropy": 5.5221943855285645,
      "epoch": 0.953255260023819,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004917895283334749,
      "loss": 5.414,
      "mean_token_accuracy": 0.1480000376701355,
      "num_tokens": 21044640.0,
      "step": 9605
    },
    {
      "entropy": 5.570703983306885,
      "epoch": 0.9537514886859865,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491780041488553,
      "loss": 5.3916,
      "mean_token_accuracy": 0.15254343599081038,
      "num_tokens": 21055728.0,
      "step": 9610
    },
    {
      "entropy": 5.53176474571228,
      "epoch": 0.9542477173481541,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004917705492679529,
      "loss": 5.394,
      "mean_token_accuracy": 0.15007168799638748,
      "num_tokens": 21066057.0,
      "step": 9615
    },
    {
      "entropy": 5.5596160888671875,
      "epoch": 0.9547439460103215,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917610516719104,
      "loss": 5.399,
      "mean_token_accuracy": 0.14553048759698867,
      "num_tokens": 21075717.0,
      "step": 9620
    },
    {
      "entropy": 5.622881317138672,
      "epoch": 0.9552401746724891,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004917515487006609,
      "loss": 5.4977,
      "mean_token_accuracy": 0.14186555594205857,
      "num_tokens": 21087120.0,
      "step": 9625
    },
    {
      "entropy": 5.591256523132325,
      "epoch": 0.9557364033346566,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00049174204035444,
      "loss": 5.4933,
      "mean_token_accuracy": 0.14599344208836557,
      "num_tokens": 21100110.0,
      "step": 9630
    },
    {
      "entropy": 5.636197137832641,
      "epoch": 0.9562326319968242,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004917325266334836,
      "loss": 5.4255,
      "mean_token_accuracy": 0.1531759276986122,
      "num_tokens": 21110411.0,
      "step": 9635
    },
    {
      "entropy": 5.470142459869384,
      "epoch": 0.9567288606589917,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004917230075380276,
      "loss": 5.3263,
      "mean_token_accuracy": 0.15397564619779586,
      "num_tokens": 21120303.0,
      "step": 9640
    },
    {
      "entropy": 5.550751209259033,
      "epoch": 0.9572250893211592,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004917134830683079,
      "loss": 5.4383,
      "mean_token_accuracy": 0.1487245351076126,
      "num_tokens": 21132312.0,
      "step": 9645
    },
    {
      "entropy": 5.557334661483765,
      "epoch": 0.9577213179833267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004917039532245608,
      "loss": 5.4163,
      "mean_token_accuracy": 0.1493322432041168,
      "num_tokens": 21142472.0,
      "step": 9650
    },
    {
      "entropy": 5.52464542388916,
      "epoch": 0.9582175466454942,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004916944180070226,
      "loss": 5.4151,
      "mean_token_accuracy": 0.1486043706536293,
      "num_tokens": 21153306.0,
      "step": 9655
    },
    {
      "entropy": 5.630894470214844,
      "epoch": 0.9587137753076618,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004916848774159299,
      "loss": 5.4613,
      "mean_token_accuracy": 0.14293530061841012,
      "num_tokens": 21165606.0,
      "step": 9660
    },
    {
      "entropy": 5.568338632583618,
      "epoch": 0.9592100039698293,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000491675331451519,
      "loss": 5.398,
      "mean_token_accuracy": 0.154398550093174,
      "num_tokens": 21176734.0,
      "step": 9665
    },
    {
      "entropy": 5.501261234283447,
      "epoch": 0.9597062326319968,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004916657801140269,
      "loss": 5.3227,
      "mean_token_accuracy": 0.15781876742839812,
      "num_tokens": 21187109.0,
      "step": 9670
    },
    {
      "entropy": 5.511800527572632,
      "epoch": 0.9602024612941643,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916562234036903,
      "loss": 5.3817,
      "mean_token_accuracy": 0.15287242829799652,
      "num_tokens": 21197281.0,
      "step": 9675
    },
    {
      "entropy": 5.5300520896911625,
      "epoch": 0.9606986899563319,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916466613207462,
      "loss": 5.4282,
      "mean_token_accuracy": 0.15005420446395873,
      "num_tokens": 21208767.0,
      "step": 9680
    },
    {
      "entropy": 5.552774572372437,
      "epoch": 0.9611949186184994,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004916370938654317,
      "loss": 5.405,
      "mean_token_accuracy": 0.15134741514921188,
      "num_tokens": 21219328.0,
      "step": 9685
    },
    {
      "entropy": 5.648916387557984,
      "epoch": 0.961691147280667,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004916275210379841,
      "loss": 5.5316,
      "mean_token_accuracy": 0.14925482422113417,
      "num_tokens": 21231476.0,
      "step": 9690
    },
    {
      "entropy": 5.5897420883178714,
      "epoch": 0.9621873759428344,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004916179428386408,
      "loss": 5.4866,
      "mean_token_accuracy": 0.1526645988225937,
      "num_tokens": 21242445.0,
      "step": 9695
    },
    {
      "entropy": 5.50202898979187,
      "epoch": 0.962683604605002,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916083592676392,
      "loss": 5.2853,
      "mean_token_accuracy": 0.15403868407011032,
      "num_tokens": 21253179.0,
      "step": 9700
    },
    {
      "entropy": 5.533032608032227,
      "epoch": 0.9631798332671695,
      "grad_norm": 1.125,
      "learning_rate": 0.0004915987703252171,
      "loss": 5.325,
      "mean_token_accuracy": 0.15881736427545548,
      "num_tokens": 21263460.0,
      "step": 9705
    },
    {
      "entropy": 5.4544576644897464,
      "epoch": 0.9636760619293371,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915891760116123,
      "loss": 5.3682,
      "mean_token_accuracy": 0.14903105348348616,
      "num_tokens": 21274819.0,
      "step": 9710
    },
    {
      "entropy": 5.535036706924439,
      "epoch": 0.9641722905915046,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004915795763270626,
      "loss": 5.3527,
      "mean_token_accuracy": 0.15288055092096328,
      "num_tokens": 21286215.0,
      "step": 9715
    },
    {
      "entropy": 5.52900972366333,
      "epoch": 0.9646685192536721,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000491569971271806,
      "loss": 5.3472,
      "mean_token_accuracy": 0.15342183262109757,
      "num_tokens": 21297033.0,
      "step": 9720
    },
    {
      "entropy": 5.453045606613159,
      "epoch": 0.9651647479158396,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915603608460809,
      "loss": 5.3409,
      "mean_token_accuracy": 0.152964086830616,
      "num_tokens": 21308019.0,
      "step": 9725
    },
    {
      "entropy": 5.5228594779968265,
      "epoch": 0.9656609765780071,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004915507450501254,
      "loss": 5.4755,
      "mean_token_accuracy": 0.1421363465487957,
      "num_tokens": 21319313.0,
      "step": 9730
    },
    {
      "entropy": 5.571038627624512,
      "epoch": 0.9661572052401747,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000491541123884178,
      "loss": 5.3861,
      "mean_token_accuracy": 0.15082513093948363,
      "num_tokens": 21329950.0,
      "step": 9735
    },
    {
      "entropy": 5.5158296585083,
      "epoch": 0.9666534339023422,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004915314973484775,
      "loss": 5.3511,
      "mean_token_accuracy": 0.15325549691915513,
      "num_tokens": 21339061.0,
      "step": 9740
    },
    {
      "entropy": 5.5039702415466305,
      "epoch": 0.9671496625645097,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004915218654432624,
      "loss": 5.4407,
      "mean_token_accuracy": 0.14974859952926636,
      "num_tokens": 21350114.0,
      "step": 9745
    },
    {
      "entropy": 5.501946115493775,
      "epoch": 0.9676458912266772,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004915122281687716,
      "loss": 5.3503,
      "mean_token_accuracy": 0.1552763521671295,
      "num_tokens": 21361219.0,
      "step": 9750
    },
    {
      "entropy": 5.5238251209259035,
      "epoch": 0.9681421198888448,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004915025855252441,
      "loss": 5.3545,
      "mean_token_accuracy": 0.15624683648347854,
      "num_tokens": 21372722.0,
      "step": 9755
    },
    {
      "entropy": 5.503102207183838,
      "epoch": 0.9686383485510123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004914929375129188,
      "loss": 5.3873,
      "mean_token_accuracy": 0.15288702249526978,
      "num_tokens": 21383350.0,
      "step": 9760
    },
    {
      "entropy": 5.534746837615967,
      "epoch": 0.9691345772131799,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004914832841320353,
      "loss": 5.4282,
      "mean_token_accuracy": 0.15095237046480178,
      "num_tokens": 21394656.0,
      "step": 9765
    },
    {
      "entropy": 5.582628536224365,
      "epoch": 0.9696308058753473,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000491473625382833,
      "loss": 5.4201,
      "mean_token_accuracy": 0.14704459607601167,
      "num_tokens": 21406163.0,
      "step": 9770
    },
    {
      "entropy": 5.432381439208984,
      "epoch": 0.9701270345375149,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491463961265551,
      "loss": 5.3252,
      "mean_token_accuracy": 0.15601050108671188,
      "num_tokens": 21418703.0,
      "step": 9775
    },
    {
      "entropy": 5.45831413269043,
      "epoch": 0.9706232631996824,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004914542917804292,
      "loss": 5.337,
      "mean_token_accuracy": 0.15713040232658387,
      "num_tokens": 21428511.0,
      "step": 9780
    },
    {
      "entropy": 5.5461279392242435,
      "epoch": 0.97111949186185,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004914446169277075,
      "loss": 5.3431,
      "mean_token_accuracy": 0.15429868698120117,
      "num_tokens": 21440350.0,
      "step": 9785
    },
    {
      "entropy": 5.594333505630493,
      "epoch": 0.9716157205240175,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004914349367076258,
      "loss": 5.4696,
      "mean_token_accuracy": 0.15229581519961358,
      "num_tokens": 21451262.0,
      "step": 9790
    },
    {
      "entropy": 5.53354434967041,
      "epoch": 0.972111949186185,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914252511204238,
      "loss": 5.4177,
      "mean_token_accuracy": 0.14953596144914627,
      "num_tokens": 21462587.0,
      "step": 9795
    },
    {
      "entropy": 5.527204608917236,
      "epoch": 0.9726081778483525,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004914155601663422,
      "loss": 5.382,
      "mean_token_accuracy": 0.1509664051234722,
      "num_tokens": 21473452.0,
      "step": 9800
    },
    {
      "entropy": 5.55464997291565,
      "epoch": 0.97310440651052,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004914058638456208,
      "loss": 5.3512,
      "mean_token_accuracy": 0.15201075226068497,
      "num_tokens": 21484011.0,
      "step": 9805
    },
    {
      "entropy": 5.431203317642212,
      "epoch": 0.9736006351726876,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004913961621585003,
      "loss": 5.3319,
      "mean_token_accuracy": 0.154771988093853,
      "num_tokens": 21495959.0,
      "step": 9810
    },
    {
      "entropy": 5.5115855693817135,
      "epoch": 0.974096863834855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004913864551052214,
      "loss": 5.3721,
      "mean_token_accuracy": 0.16235917955636978,
      "num_tokens": 21507496.0,
      "step": 9815
    },
    {
      "entropy": 5.613047504425049,
      "epoch": 0.9745930924970226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004913767426860246,
      "loss": 5.4555,
      "mean_token_accuracy": 0.14365342110395432,
      "num_tokens": 21517532.0,
      "step": 9820
    },
    {
      "entropy": 5.468502807617187,
      "epoch": 0.9750893211591901,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004913670249011508,
      "loss": 5.2646,
      "mean_token_accuracy": 0.15803594142198563,
      "num_tokens": 21527969.0,
      "step": 9825
    },
    {
      "entropy": 5.478287553787231,
      "epoch": 0.9755855498213577,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004913573017508411,
      "loss": 5.3744,
      "mean_token_accuracy": 0.146577125787735,
      "num_tokens": 21538243.0,
      "step": 9830
    },
    {
      "entropy": 5.512112236022949,
      "epoch": 0.9760817784835252,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004913475732353365,
      "loss": 5.3768,
      "mean_token_accuracy": 0.156990135461092,
      "num_tokens": 21548702.0,
      "step": 9835
    },
    {
      "entropy": 5.422013282775879,
      "epoch": 0.9765780071456928,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004913378393548784,
      "loss": 5.3453,
      "mean_token_accuracy": 0.15282752811908723,
      "num_tokens": 21558583.0,
      "step": 9840
    },
    {
      "entropy": 5.529709959030152,
      "epoch": 0.9770742358078602,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004913281001097078,
      "loss": 5.3318,
      "mean_token_accuracy": 0.15674136728048324,
      "num_tokens": 21569845.0,
      "step": 9845
    },
    {
      "entropy": 5.604092121124268,
      "epoch": 0.9775704644700278,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004913183555000667,
      "loss": 5.4769,
      "mean_token_accuracy": 0.1458521731197834,
      "num_tokens": 21580540.0,
      "step": 9850
    },
    {
      "entropy": 5.530924940109253,
      "epoch": 0.9780666931321953,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004913086055261963,
      "loss": 5.3353,
      "mean_token_accuracy": 0.15436560064554214,
      "num_tokens": 21590982.0,
      "step": 9855
    },
    {
      "entropy": 5.512540817260742,
      "epoch": 0.9785629217943629,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004912988501883387,
      "loss": 5.3646,
      "mean_token_accuracy": 0.1509571775794029,
      "num_tokens": 21602817.0,
      "step": 9860
    },
    {
      "entropy": 5.542688941955566,
      "epoch": 0.9790591504565304,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004912890894867357,
      "loss": 5.4116,
      "mean_token_accuracy": 0.15492914915084838,
      "num_tokens": 21613422.0,
      "step": 9865
    },
    {
      "entropy": 5.520282173156739,
      "epoch": 0.9795553791186979,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004912793234216293,
      "loss": 5.393,
      "mean_token_accuracy": 0.1465645872056484,
      "num_tokens": 21624241.0,
      "step": 9870
    },
    {
      "entropy": 5.4686418056488035,
      "epoch": 0.9800516077808654,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004912695519932618,
      "loss": 5.388,
      "mean_token_accuracy": 0.1543208673596382,
      "num_tokens": 21636243.0,
      "step": 9875
    },
    {
      "entropy": 5.5243322372436525,
      "epoch": 0.9805478364430329,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004912597752018753,
      "loss": 5.4276,
      "mean_token_accuracy": 0.15045800805091858,
      "num_tokens": 21647918.0,
      "step": 9880
    },
    {
      "entropy": 5.579762506484985,
      "epoch": 0.9810440651052005,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004912499930477126,
      "loss": 5.3982,
      "mean_token_accuracy": 0.15023436844348909,
      "num_tokens": 21659236.0,
      "step": 9885
    },
    {
      "entropy": 5.558022212982178,
      "epoch": 0.981540293767368,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004912402055310158,
      "loss": 5.3941,
      "mean_token_accuracy": 0.15856010019779204,
      "num_tokens": 21669647.0,
      "step": 9890
    },
    {
      "entropy": 5.503682804107666,
      "epoch": 0.9820365224295355,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004912304126520279,
      "loss": 5.4421,
      "mean_token_accuracy": 0.14751182198524476,
      "num_tokens": 21680369.0,
      "step": 9895
    },
    {
      "entropy": 5.610123968124389,
      "epoch": 0.982532751091703,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004912206144109917,
      "loss": 5.4481,
      "mean_token_accuracy": 0.15194133967161177,
      "num_tokens": 21690942.0,
      "step": 9900
    },
    {
      "entropy": 5.506767797470093,
      "epoch": 0.9830289797538706,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004912108108081503,
      "loss": 5.3608,
      "mean_token_accuracy": 0.1558390349149704,
      "num_tokens": 21701945.0,
      "step": 9905
    },
    {
      "entropy": 5.537484979629516,
      "epoch": 0.9835252084160381,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912010018437465,
      "loss": 5.3901,
      "mean_token_accuracy": 0.15187972411513329,
      "num_tokens": 21712246.0,
      "step": 9910
    },
    {
      "entropy": 5.615129089355468,
      "epoch": 0.9840214370782057,
      "grad_norm": 1.0,
      "learning_rate": 0.000491191187518024,
      "loss": 5.4073,
      "mean_token_accuracy": 0.15086458772420883,
      "num_tokens": 21721660.0,
      "step": 9915
    },
    {
      "entropy": 5.471030855178833,
      "epoch": 0.9845176657403731,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004911813678312257,
      "loss": 5.2745,
      "mean_token_accuracy": 0.1562454655766487,
      "num_tokens": 21732157.0,
      "step": 9920
    },
    {
      "entropy": 5.523634719848633,
      "epoch": 0.9850138944025407,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004911715427835953,
      "loss": 5.4241,
      "mean_token_accuracy": 0.15104619264602662,
      "num_tokens": 21741727.0,
      "step": 9925
    },
    {
      "entropy": 5.522456073760987,
      "epoch": 0.9855101230647082,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004911617123753765,
      "loss": 5.3577,
      "mean_token_accuracy": 0.15722549110651016,
      "num_tokens": 21752596.0,
      "step": 9930
    },
    {
      "entropy": 5.4953498363494875,
      "epoch": 0.9860063517268758,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004911518766068131,
      "loss": 5.247,
      "mean_token_accuracy": 0.16038501113653184,
      "num_tokens": 21763978.0,
      "step": 9935
    },
    {
      "entropy": 5.484000968933105,
      "epoch": 0.9865025803890433,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004911420354781487,
      "loss": 5.411,
      "mean_token_accuracy": 0.14073246717453003,
      "num_tokens": 21774961.0,
      "step": 9940
    },
    {
      "entropy": 5.512303256988526,
      "epoch": 0.9869988090512108,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004911321889896278,
      "loss": 5.3273,
      "mean_token_accuracy": 0.15784133970737457,
      "num_tokens": 21785387.0,
      "step": 9945
    },
    {
      "entropy": 5.482459306716919,
      "epoch": 0.9874950377133783,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004911223371414943,
      "loss": 5.3815,
      "mean_token_accuracy": 0.15688561797142028,
      "num_tokens": 21795982.0,
      "step": 9950
    },
    {
      "entropy": 5.539654302597046,
      "epoch": 0.9879912663755459,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004911124799339924,
      "loss": 5.4091,
      "mean_token_accuracy": 0.15125578343868257,
      "num_tokens": 21806768.0,
      "step": 9955
    },
    {
      "entropy": 5.526731872558594,
      "epoch": 0.9884874950377134,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004911026173673667,
      "loss": 5.3836,
      "mean_token_accuracy": 0.1518492192029953,
      "num_tokens": 21818563.0,
      "step": 9960
    },
    {
      "entropy": 5.518647766113281,
      "epoch": 0.9889837236998809,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004910927494418619,
      "loss": 5.4362,
      "mean_token_accuracy": 0.15531470030546188,
      "num_tokens": 21830253.0,
      "step": 9965
    },
    {
      "entropy": 5.550831985473633,
      "epoch": 0.9894799523620484,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004910828761577224,
      "loss": 5.4372,
      "mean_token_accuracy": 0.15268764942884444,
      "num_tokens": 21842353.0,
      "step": 9970
    },
    {
      "entropy": 5.561376857757568,
      "epoch": 0.9899761810242159,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004910729975151933,
      "loss": 5.3629,
      "mean_token_accuracy": 0.1572612300515175,
      "num_tokens": 21852066.0,
      "step": 9975
    },
    {
      "entropy": 5.440600776672364,
      "epoch": 0.9904724096863835,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004910631135145194,
      "loss": 5.4017,
      "mean_token_accuracy": 0.1557166174054146,
      "num_tokens": 21862754.0,
      "step": 9980
    },
    {
      "entropy": 5.422006988525391,
      "epoch": 0.990968638348551,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004910532241559458,
      "loss": 5.3128,
      "mean_token_accuracy": 0.16325167268514634,
      "num_tokens": 21873531.0,
      "step": 9985
    },
    {
      "entropy": 5.460115575790406,
      "epoch": 0.9914648670107186,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004910433294397179,
      "loss": 5.2378,
      "mean_token_accuracy": 0.162018983066082,
      "num_tokens": 21884339.0,
      "step": 9990
    },
    {
      "entropy": 5.466616630554199,
      "epoch": 0.991961095672886,
      "grad_norm": 1.0625,
      "learning_rate": 0.000491033429366081,
      "loss": 5.3695,
      "mean_token_accuracy": 0.14877082854509355,
      "num_tokens": 21895489.0,
      "step": 9995
    },
    {
      "entropy": 5.464471864700317,
      "epoch": 0.9924573243350536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004910235239352804,
      "loss": 5.338,
      "mean_token_accuracy": 0.1583110809326172,
      "num_tokens": 21906832.0,
      "step": 10000
    },
    {
      "entropy": 5.525733613967896,
      "epoch": 0.9929535529972211,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000491013613147562,
      "loss": 5.3814,
      "mean_token_accuracy": 0.1545408546924591,
      "num_tokens": 21916979.0,
      "step": 10005
    },
    {
      "entropy": 5.528966569900513,
      "epoch": 0.9934497816593887,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004910036970031715,
      "loss": 5.2912,
      "mean_token_accuracy": 0.1568614974617958,
      "num_tokens": 21927352.0,
      "step": 10010
    },
    {
      "entropy": 5.48080267906189,
      "epoch": 0.9939460103215562,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004909937755023547,
      "loss": 5.398,
      "mean_token_accuracy": 0.14980651289224625,
      "num_tokens": 21937765.0,
      "step": 10015
    },
    {
      "entropy": 5.446048736572266,
      "epoch": 0.9944422389837237,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909838486453577,
      "loss": 5.3,
      "mean_token_accuracy": 0.15599877685308455,
      "num_tokens": 21948659.0,
      "step": 10020
    },
    {
      "entropy": 5.482259464263916,
      "epoch": 0.9949384676458912,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004909739164324267,
      "loss": 5.4098,
      "mean_token_accuracy": 0.14727606102824212,
      "num_tokens": 21959356.0,
      "step": 10025
    },
    {
      "entropy": 5.665633201599121,
      "epoch": 0.9954346963080588,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004909639788638079,
      "loss": 5.5126,
      "mean_token_accuracy": 0.1413301110267639,
      "num_tokens": 21971321.0,
      "step": 10030
    },
    {
      "entropy": 5.51676869392395,
      "epoch": 0.9959309249702263,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004909540359397477,
      "loss": 5.3872,
      "mean_token_accuracy": 0.14769014120101928,
      "num_tokens": 21981252.0,
      "step": 10035
    },
    {
      "entropy": 5.5007164001464846,
      "epoch": 0.9964271536323938,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004909440876604929,
      "loss": 5.3946,
      "mean_token_accuracy": 0.14752098619937898,
      "num_tokens": 21991820.0,
      "step": 10040
    },
    {
      "entropy": 5.527020406723023,
      "epoch": 0.9969233822945613,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049093413402629,
      "loss": 5.3403,
      "mean_token_accuracy": 0.15451508164405822,
      "num_tokens": 22003619.0,
      "step": 10045
    },
    {
      "entropy": 5.502935218811035,
      "epoch": 0.9974196109567288,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004909241750373858,
      "loss": 5.4043,
      "mean_token_accuracy": 0.1517265781760216,
      "num_tokens": 22016116.0,
      "step": 10050
    },
    {
      "entropy": 5.532568645477295,
      "epoch": 0.9979158396188964,
      "grad_norm": 1.125,
      "learning_rate": 0.0004909142106940275,
      "loss": 5.3294,
      "mean_token_accuracy": 0.15348412320017815,
      "num_tokens": 22026115.0,
      "step": 10055
    },
    {
      "entropy": 5.512277269363404,
      "epoch": 0.9984120682810639,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000490904240996462,
      "loss": 5.4013,
      "mean_token_accuracy": 0.1583564430475235,
      "num_tokens": 22038889.0,
      "step": 10060
    },
    {
      "entropy": 5.470356416702271,
      "epoch": 0.9989082969432315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908942659449364,
      "loss": 5.3683,
      "mean_token_accuracy": 0.15174410194158555,
      "num_tokens": 22050773.0,
      "step": 10065
    },
    {
      "entropy": 5.488900089263916,
      "epoch": 0.9994045256053989,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004908842855396982,
      "loss": 5.3755,
      "mean_token_accuracy": 0.14813873022794724,
      "num_tokens": 22062869.0,
      "step": 10070
    },
    {
      "entropy": 5.4824573516845705,
      "epoch": 0.9999007542675665,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004908742997809951,
      "loss": 5.3924,
      "mean_token_accuracy": 0.1515072375535965,
      "num_tokens": 22073898.0,
      "step": 10075
    },
    {
      "entropy": 5.566206359863282,
      "epoch": 1.0003969829297341,
      "grad_norm": 1.125,
      "learning_rate": 0.0004908643086690744,
      "loss": 5.2824,
      "mean_token_accuracy": 0.15128990337252618,
      "num_tokens": 22084383.0,
      "step": 10080
    },
    {
      "entropy": 5.517045164108277,
      "epoch": 1.0008932115919016,
      "grad_norm": 1.09375,
      "learning_rate": 0.000490854312204184,
      "loss": 5.3384,
      "mean_token_accuracy": 0.15505650639533997,
      "num_tokens": 22094990.0,
      "step": 10085
    },
    {
      "entropy": 5.508151578903198,
      "epoch": 1.001389440254069,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004908443103865719,
      "loss": 5.3567,
      "mean_token_accuracy": 0.15155285745859146,
      "num_tokens": 22105259.0,
      "step": 10090
    },
    {
      "entropy": 5.492518472671509,
      "epoch": 1.0018856689162365,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004908343032164859,
      "loss": 5.3243,
      "mean_token_accuracy": 0.15563040524721145,
      "num_tokens": 22116348.0,
      "step": 10095
    },
    {
      "entropy": 5.509250450134277,
      "epoch": 1.0023818975784042,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004908242906941743,
      "loss": 5.3201,
      "mean_token_accuracy": 0.1568908914923668,
      "num_tokens": 22127311.0,
      "step": 10100
    },
    {
      "entropy": 5.576716661453247,
      "epoch": 1.0028781262405717,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004908142728198853,
      "loss": 5.3531,
      "mean_token_accuracy": 0.14858218133449555,
      "num_tokens": 22137929.0,
      "step": 10105
    },
    {
      "entropy": 5.6098517894744875,
      "epoch": 1.0033743549027392,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004908042495938673,
      "loss": 5.4473,
      "mean_token_accuracy": 0.14790024757385253,
      "num_tokens": 22148202.0,
      "step": 10110
    },
    {
      "entropy": 5.461227893829346,
      "epoch": 1.0038705835649067,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004907942210163692,
      "loss": 5.2283,
      "mean_token_accuracy": 0.15380824655294417,
      "num_tokens": 22158955.0,
      "step": 10115
    },
    {
      "entropy": 5.46384973526001,
      "epoch": 1.0043668122270741,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004907841870876392,
      "loss": 5.2692,
      "mean_token_accuracy": 0.1532580226659775,
      "num_tokens": 22169569.0,
      "step": 10120
    },
    {
      "entropy": 5.482016897201538,
      "epoch": 1.0048630408892418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004907741478079263,
      "loss": 5.254,
      "mean_token_accuracy": 0.15018723905086517,
      "num_tokens": 22180171.0,
      "step": 10125
    },
    {
      "entropy": 5.4763405323028564,
      "epoch": 1.0053592695514093,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004907641031774796,
      "loss": 5.361,
      "mean_token_accuracy": 0.1540578067302704,
      "num_tokens": 22190531.0,
      "step": 10130
    },
    {
      "entropy": 5.479198741912842,
      "epoch": 1.0058554982135768,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490754053196548,
      "loss": 5.2535,
      "mean_token_accuracy": 0.15639664232730865,
      "num_tokens": 22201495.0,
      "step": 10135
    },
    {
      "entropy": 5.577369499206543,
      "epoch": 1.0063517268757443,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004907439978653809,
      "loss": 5.3222,
      "mean_token_accuracy": 0.14795395731925964,
      "num_tokens": 22212941.0,
      "step": 10140
    },
    {
      "entropy": 5.586483526229858,
      "epoch": 1.006847955537912,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004907339371842275,
      "loss": 5.3768,
      "mean_token_accuracy": 0.15148166567087173,
      "num_tokens": 22225212.0,
      "step": 10145
    },
    {
      "entropy": 5.4729358673095705,
      "epoch": 1.0073441842000794,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004907238711533373,
      "loss": 5.3264,
      "mean_token_accuracy": 0.1502268932759762,
      "num_tokens": 22236370.0,
      "step": 10150
    },
    {
      "entropy": 5.456326961517334,
      "epoch": 1.007840412862247,
      "grad_norm": 1.125,
      "learning_rate": 0.0004907137997729599,
      "loss": 5.3031,
      "mean_token_accuracy": 0.15459689497947693,
      "num_tokens": 22247641.0,
      "step": 10155
    },
    {
      "entropy": 5.485335969924927,
      "epoch": 1.0083366415244144,
      "grad_norm": 1.0625,
      "learning_rate": 0.000490703723043345,
      "loss": 5.3562,
      "mean_token_accuracy": 0.15528895556926728,
      "num_tokens": 22258599.0,
      "step": 10160
    },
    {
      "entropy": 5.484523773193359,
      "epoch": 1.008832870186582,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004906936409647428,
      "loss": 5.2824,
      "mean_token_accuracy": 0.15417744368314742,
      "num_tokens": 22270913.0,
      "step": 10165
    },
    {
      "entropy": 5.460978555679321,
      "epoch": 1.0093290988487496,
      "grad_norm": 1.203125,
      "learning_rate": 0.000490683553537403,
      "loss": 5.1868,
      "mean_token_accuracy": 0.1607227846980095,
      "num_tokens": 22281053.0,
      "step": 10170
    },
    {
      "entropy": 5.450496244430542,
      "epoch": 1.009825327510917,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004906734607615758,
      "loss": 5.1973,
      "mean_token_accuracy": 0.15885399729013444,
      "num_tokens": 22292570.0,
      "step": 10175
    },
    {
      "entropy": 5.474445009231568,
      "epoch": 1.0103215561730845,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004906633626375114,
      "loss": 5.2673,
      "mean_token_accuracy": 0.1523367166519165,
      "num_tokens": 22303548.0,
      "step": 10180
    },
    {
      "entropy": 5.516963243484497,
      "epoch": 1.010817784835252,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004906532591654604,
      "loss": 5.3145,
      "mean_token_accuracy": 0.1596198707818985,
      "num_tokens": 22314133.0,
      "step": 10185
    },
    {
      "entropy": 5.509190893173217,
      "epoch": 1.0113140134974197,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004906431503456733,
      "loss": 5.2757,
      "mean_token_accuracy": 0.1552529439330101,
      "num_tokens": 22326431.0,
      "step": 10190
    },
    {
      "entropy": 5.452233791351318,
      "epoch": 1.0118102421595871,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004906330361784007,
      "loss": 5.3215,
      "mean_token_accuracy": 0.15459493696689605,
      "num_tokens": 22338507.0,
      "step": 10195
    },
    {
      "entropy": 5.492002487182617,
      "epoch": 1.0123064708217546,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004906229166638936,
      "loss": 5.29,
      "mean_token_accuracy": 0.15530695915222167,
      "num_tokens": 22348863.0,
      "step": 10200
    },
    {
      "entropy": 5.4199137687683105,
      "epoch": 1.012802699483922,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004906127918024026,
      "loss": 5.2465,
      "mean_token_accuracy": 0.15835145115852356,
      "num_tokens": 22359000.0,
      "step": 10205
    },
    {
      "entropy": 5.53760781288147,
      "epoch": 1.0132989281460898,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004906026615941791,
      "loss": 5.3738,
      "mean_token_accuracy": 0.14833952188491822,
      "num_tokens": 22369539.0,
      "step": 10210
    },
    {
      "entropy": 5.509432744979859,
      "epoch": 1.0137951568082573,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004905925260394741,
      "loss": 5.3433,
      "mean_token_accuracy": 0.15099112391471864,
      "num_tokens": 22380700.0,
      "step": 10215
    },
    {
      "entropy": 5.455250263214111,
      "epoch": 1.0142913854704247,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490582385138539,
      "loss": 5.3165,
      "mean_token_accuracy": 0.146890389919281,
      "num_tokens": 22392151.0,
      "step": 10220
    },
    {
      "entropy": 5.4920666217803955,
      "epoch": 1.0147876141325922,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004905722388916254,
      "loss": 5.2961,
      "mean_token_accuracy": 0.1602615997195244,
      "num_tokens": 22401735.0,
      "step": 10225
    },
    {
      "entropy": 5.522230195999145,
      "epoch": 1.01528384279476,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004905620872989846,
      "loss": 5.3615,
      "mean_token_accuracy": 0.15181516110897064,
      "num_tokens": 22412549.0,
      "step": 10230
    },
    {
      "entropy": 5.532883548736573,
      "epoch": 1.0157800714569274,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004905519303608687,
      "loss": 5.3661,
      "mean_token_accuracy": 0.15341316014528275,
      "num_tokens": 22424728.0,
      "step": 10235
    },
    {
      "entropy": 5.448158216476441,
      "epoch": 1.0162763001190949,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004905417680775294,
      "loss": 5.2211,
      "mean_token_accuracy": 0.1563728019595146,
      "num_tokens": 22434197.0,
      "step": 10240
    },
    {
      "entropy": 5.54134726524353,
      "epoch": 1.0167725287812623,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004905316004492186,
      "loss": 5.4175,
      "mean_token_accuracy": 0.14646299481391906,
      "num_tokens": 22444671.0,
      "step": 10245
    },
    {
      "entropy": 5.504880762100219,
      "epoch": 1.01726875744343,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004905214274761885,
      "loss": 5.2875,
      "mean_token_accuracy": 0.15071241706609725,
      "num_tokens": 22456636.0,
      "step": 10250
    },
    {
      "entropy": 5.424612331390381,
      "epoch": 1.0177649861055975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004905112491586915,
      "loss": 5.2102,
      "mean_token_accuracy": 0.15837929844856263,
      "num_tokens": 22467011.0,
      "step": 10255
    },
    {
      "entropy": 5.468602848052979,
      "epoch": 1.018261214767765,
      "grad_norm": 1.109375,
      "learning_rate": 0.00049050106549698,
      "loss": 5.3077,
      "mean_token_accuracy": 0.1564371481537819,
      "num_tokens": 22477630.0,
      "step": 10260
    },
    {
      "entropy": 5.4724374294281,
      "epoch": 1.0187574434299325,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004904908764913063,
      "loss": 5.2562,
      "mean_token_accuracy": 0.1634463921189308,
      "num_tokens": 22488132.0,
      "step": 10265
    },
    {
      "entropy": 5.456572437286377,
      "epoch": 1.0192536720921,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004904806821419233,
      "loss": 5.3243,
      "mean_token_accuracy": 0.15164103657007216,
      "num_tokens": 22499430.0,
      "step": 10270
    },
    {
      "entropy": 5.43280520439148,
      "epoch": 1.0197499007542676,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004904704824490835,
      "loss": 5.2305,
      "mean_token_accuracy": 0.15667520463466644,
      "num_tokens": 22509728.0,
      "step": 10275
    },
    {
      "entropy": 5.473942804336548,
      "epoch": 1.020246129416435,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004904602774130402,
      "loss": 5.3493,
      "mean_token_accuracy": 0.14827937483787537,
      "num_tokens": 22520955.0,
      "step": 10280
    },
    {
      "entropy": 5.5664245128631595,
      "epoch": 1.0207423580786026,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004904500670340463,
      "loss": 5.3659,
      "mean_token_accuracy": 0.15064165666699408,
      "num_tokens": 22530440.0,
      "step": 10285
    },
    {
      "entropy": 5.488139343261719,
      "epoch": 1.02123858674077,
      "grad_norm": 1.125,
      "learning_rate": 0.000490439851312355,
      "loss": 5.3375,
      "mean_token_accuracy": 0.16039537638425827,
      "num_tokens": 22541539.0,
      "step": 10290
    },
    {
      "entropy": 5.481247282028198,
      "epoch": 1.0217348154029378,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004904296302482196,
      "loss": 5.3113,
      "mean_token_accuracy": 0.15504374355077744,
      "num_tokens": 22555717.0,
      "step": 10295
    },
    {
      "entropy": 5.572066688537598,
      "epoch": 1.0222310440651052,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004904194038418935,
      "loss": 5.4087,
      "mean_token_accuracy": 0.14975774437189102,
      "num_tokens": 22566938.0,
      "step": 10300
    },
    {
      "entropy": 5.529646492004394,
      "epoch": 1.0227272727272727,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004904091720936305,
      "loss": 5.2842,
      "mean_token_accuracy": 0.15312067568302154,
      "num_tokens": 22578725.0,
      "step": 10305
    },
    {
      "entropy": 5.4440854549407955,
      "epoch": 1.0232235013894402,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004903989350036841,
      "loss": 5.2309,
      "mean_token_accuracy": 0.15952013581991195,
      "num_tokens": 22588476.0,
      "step": 10310
    },
    {
      "entropy": 5.523025369644165,
      "epoch": 1.0237197300516079,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004903886925723083,
      "loss": 5.4108,
      "mean_token_accuracy": 0.14259994626045228,
      "num_tokens": 22600294.0,
      "step": 10315
    },
    {
      "entropy": 5.49133710861206,
      "epoch": 1.0242159587137754,
      "grad_norm": 1.109375,
      "learning_rate": 0.000490378444799757,
      "loss": 5.2986,
      "mean_token_accuracy": 0.15516656786203384,
      "num_tokens": 22610578.0,
      "step": 10320
    },
    {
      "entropy": 5.531575155258179,
      "epoch": 1.0247121873759428,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004903681916862844,
      "loss": 5.3185,
      "mean_token_accuracy": 0.15520282685756684,
      "num_tokens": 22620988.0,
      "step": 10325
    },
    {
      "entropy": 5.483933734893799,
      "epoch": 1.0252084160381103,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004903579332321447,
      "loss": 5.3033,
      "mean_token_accuracy": 0.1531194567680359,
      "num_tokens": 22632198.0,
      "step": 10330
    },
    {
      "entropy": 5.481255149841308,
      "epoch": 1.0257046447002778,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004903476694375924,
      "loss": 5.3011,
      "mean_token_accuracy": 0.15980833768844604,
      "num_tokens": 22642543.0,
      "step": 10335
    },
    {
      "entropy": 5.511505031585694,
      "epoch": 1.0262008733624455,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004903374003028819,
      "loss": 5.3095,
      "mean_token_accuracy": 0.15540917515754699,
      "num_tokens": 22655453.0,
      "step": 10340
    },
    {
      "entropy": 5.453513050079346,
      "epoch": 1.026697102024613,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004903271258282678,
      "loss": 5.2389,
      "mean_token_accuracy": 0.15846578031778336,
      "num_tokens": 22667342.0,
      "step": 10345
    },
    {
      "entropy": 5.504372501373291,
      "epoch": 1.0271933306867804,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004903168460140049,
      "loss": 5.3457,
      "mean_token_accuracy": 0.1555282548069954,
      "num_tokens": 22676890.0,
      "step": 10350
    },
    {
      "entropy": 5.4711222648620605,
      "epoch": 1.027689559348948,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004903065608603484,
      "loss": 5.3196,
      "mean_token_accuracy": 0.15337342619895936,
      "num_tokens": 22687809.0,
      "step": 10355
    },
    {
      "entropy": 5.470293235778809,
      "epoch": 1.0281857880111156,
      "grad_norm": 1.109375,
      "learning_rate": 0.000490296270367553,
      "loss": 5.3323,
      "mean_token_accuracy": 0.15351829379796983,
      "num_tokens": 22697938.0,
      "step": 10360
    },
    {
      "entropy": 5.5687767505645756,
      "epoch": 1.028682016673283,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004902859745358741,
      "loss": 5.3166,
      "mean_token_accuracy": 0.15822121798992156,
      "num_tokens": 22708303.0,
      "step": 10365
    },
    {
      "entropy": 5.480257892608643,
      "epoch": 1.0291782453354505,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004902756733655668,
      "loss": 5.2522,
      "mean_token_accuracy": 0.16192941814661027,
      "num_tokens": 22717987.0,
      "step": 10370
    },
    {
      "entropy": 5.44890022277832,
      "epoch": 1.029674473997618,
      "grad_norm": 1.0,
      "learning_rate": 0.0004902653668568868,
      "loss": 5.3225,
      "mean_token_accuracy": 0.15169527679681777,
      "num_tokens": 22729672.0,
      "step": 10375
    },
    {
      "entropy": 5.518046569824219,
      "epoch": 1.0301707026597857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004902550550100896,
      "loss": 5.2897,
      "mean_token_accuracy": 0.1602475255727768,
      "num_tokens": 22741296.0,
      "step": 10380
    },
    {
      "entropy": 5.453836154937744,
      "epoch": 1.0306669313219532,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004902447378254307,
      "loss": 5.2899,
      "mean_token_accuracy": 0.16084748953580857,
      "num_tokens": 22752337.0,
      "step": 10385
    },
    {
      "entropy": 5.53193302154541,
      "epoch": 1.0311631599841207,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004902344153031663,
      "loss": 5.3155,
      "mean_token_accuracy": 0.15546908974647522,
      "num_tokens": 22762681.0,
      "step": 10390
    },
    {
      "entropy": 5.5478943347930905,
      "epoch": 1.0316593886462881,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000490224087443552,
      "loss": 5.3538,
      "mean_token_accuracy": 0.14933211654424666,
      "num_tokens": 22774275.0,
      "step": 10395
    },
    {
      "entropy": 5.464819860458374,
      "epoch": 1.0321556173084558,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004902137542468442,
      "loss": 5.2553,
      "mean_token_accuracy": 0.1558536559343338,
      "num_tokens": 22785944.0,
      "step": 10400
    },
    {
      "entropy": 5.417013359069824,
      "epoch": 1.0326518459706233,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004902034157132991,
      "loss": 5.1636,
      "mean_token_accuracy": 0.15922075659036636,
      "num_tokens": 22796543.0,
      "step": 10405
    },
    {
      "entropy": 5.420746564865112,
      "epoch": 1.0331480746327908,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004901930718431729,
      "loss": 5.2928,
      "mean_token_accuracy": 0.15623731017112732,
      "num_tokens": 22807957.0,
      "step": 10410
    },
    {
      "entropy": 5.537975835800171,
      "epoch": 1.0336443032949583,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004901827226367223,
      "loss": 5.3901,
      "mean_token_accuracy": 0.15496519654989244,
      "num_tokens": 22818903.0,
      "step": 10415
    },
    {
      "entropy": 5.550262355804444,
      "epoch": 1.0341405319571257,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004901723680942038,
      "loss": 5.4088,
      "mean_token_accuracy": 0.15378930121660234,
      "num_tokens": 22829286.0,
      "step": 10420
    },
    {
      "entropy": 5.588795280456543,
      "epoch": 1.0346367606192934,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004901620082158742,
      "loss": 5.3203,
      "mean_token_accuracy": 0.1501934349536896,
      "num_tokens": 22839860.0,
      "step": 10425
    },
    {
      "entropy": 5.563083553314209,
      "epoch": 1.035132989281461,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004901516430019905,
      "loss": 5.3861,
      "mean_token_accuracy": 0.1440488427877426,
      "num_tokens": 22852238.0,
      "step": 10430
    },
    {
      "entropy": 5.51026759147644,
      "epoch": 1.0356292179436284,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004901412724528097,
      "loss": 5.3398,
      "mean_token_accuracy": 0.15269775912165642,
      "num_tokens": 22863013.0,
      "step": 10435
    },
    {
      "entropy": 5.439476203918457,
      "epoch": 1.0361254466057959,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901308965685889,
      "loss": 5.2029,
      "mean_token_accuracy": 0.15792632550001146,
      "num_tokens": 22874233.0,
      "step": 10440
    },
    {
      "entropy": 5.505237436294555,
      "epoch": 1.0366216752679636,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004901205153495853,
      "loss": 5.3767,
      "mean_token_accuracy": 0.15246356502175332,
      "num_tokens": 22885313.0,
      "step": 10445
    },
    {
      "entropy": 5.479188871383667,
      "epoch": 1.037117903930131,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004901101287960566,
      "loss": 5.2642,
      "mean_token_accuracy": 0.15787880420684813,
      "num_tokens": 22895365.0,
      "step": 10450
    },
    {
      "entropy": 5.42048020362854,
      "epoch": 1.0376141325922985,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900997369082602,
      "loss": 5.2609,
      "mean_token_accuracy": 0.15427735149860383,
      "num_tokens": 22906392.0,
      "step": 10455
    },
    {
      "entropy": 5.4296976089477536,
      "epoch": 1.038110361254466,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004900893396864539,
      "loss": 5.2146,
      "mean_token_accuracy": 0.16176985651254655,
      "num_tokens": 22916760.0,
      "step": 10460
    },
    {
      "entropy": 5.4844718933105465,
      "epoch": 1.0386065899166337,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004900789371308953,
      "loss": 5.2755,
      "mean_token_accuracy": 0.1553081229329109,
      "num_tokens": 22928070.0,
      "step": 10465
    },
    {
      "entropy": 5.383338832855225,
      "epoch": 1.0391028185788012,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004900685292418425,
      "loss": 5.2592,
      "mean_token_accuracy": 0.1594421610236168,
      "num_tokens": 22938815.0,
      "step": 10470
    },
    {
      "entropy": 5.492303705215454,
      "epoch": 1.0395990472409686,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004900581160195537,
      "loss": 5.3315,
      "mean_token_accuracy": 0.14876061156392098,
      "num_tokens": 22949806.0,
      "step": 10475
    },
    {
      "entropy": 5.434783554077148,
      "epoch": 1.040095275903136,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000490047697464287,
      "loss": 5.1864,
      "mean_token_accuracy": 0.15768402963876724,
      "num_tokens": 22960880.0,
      "step": 10480
    },
    {
      "entropy": 5.485425615310669,
      "epoch": 1.0405915045653038,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004900372735763009,
      "loss": 5.3281,
      "mean_token_accuracy": 0.15081650167703628,
      "num_tokens": 22972828.0,
      "step": 10485
    },
    {
      "entropy": 5.419815635681152,
      "epoch": 1.0410877332274713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004900268443558536,
      "loss": 5.223,
      "mean_token_accuracy": 0.16018084585666656,
      "num_tokens": 22981869.0,
      "step": 10490
    },
    {
      "entropy": 5.500895023345947,
      "epoch": 1.0415839618896388,
      "grad_norm": 1.203125,
      "learning_rate": 0.000490016409803204,
      "loss": 5.3328,
      "mean_token_accuracy": 0.15161504596471786,
      "num_tokens": 22993111.0,
      "step": 10495
    },
    {
      "entropy": 5.4794636249542235,
      "epoch": 1.0420801905518062,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004900059699186108,
      "loss": 5.3067,
      "mean_token_accuracy": 0.15483776479959488,
      "num_tokens": 23003115.0,
      "step": 10500
    },
    {
      "entropy": 5.569976806640625,
      "epoch": 1.0425764192139737,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004899955247023328,
      "loss": 5.3497,
      "mean_token_accuracy": 0.15440469831228257,
      "num_tokens": 23012587.0,
      "step": 10505
    },
    {
      "entropy": 5.501140880584717,
      "epoch": 1.0430726478761414,
      "grad_norm": 1.078125,
      "learning_rate": 0.000489985074154629,
      "loss": 5.3761,
      "mean_token_accuracy": 0.14949385523796083,
      "num_tokens": 23024455.0,
      "step": 10510
    },
    {
      "entropy": 5.464270210266113,
      "epoch": 1.0435688765383089,
      "grad_norm": 1.125,
      "learning_rate": 0.0004899746182757587,
      "loss": 5.2835,
      "mean_token_accuracy": 0.15245273113250732,
      "num_tokens": 23034430.0,
      "step": 10515
    },
    {
      "entropy": 5.550485849380493,
      "epoch": 1.0440651052004764,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004899641570659811,
      "loss": 5.3594,
      "mean_token_accuracy": 0.15770236030220985,
      "num_tokens": 23044811.0,
      "step": 10520
    },
    {
      "entropy": 5.521479320526123,
      "epoch": 1.0445613338626438,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004899536905255555,
      "loss": 5.3915,
      "mean_token_accuracy": 0.15099400877952576,
      "num_tokens": 23056179.0,
      "step": 10525
    },
    {
      "entropy": 5.450442981719971,
      "epoch": 1.0450575625248115,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004899432186547416,
      "loss": 5.3103,
      "mean_token_accuracy": 0.15484211593866348,
      "num_tokens": 23068331.0,
      "step": 10530
    },
    {
      "entropy": 5.47597713470459,
      "epoch": 1.045553791186979,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004899327414537991,
      "loss": 5.3017,
      "mean_token_accuracy": 0.1488495111465454,
      "num_tokens": 23079213.0,
      "step": 10535
    },
    {
      "entropy": 5.610487461090088,
      "epoch": 1.0460500198491465,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004899222589229877,
      "loss": 5.4026,
      "mean_token_accuracy": 0.1460346281528473,
      "num_tokens": 23089860.0,
      "step": 10540
    },
    {
      "entropy": 5.4342453956604,
      "epoch": 1.046546248511314,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004899117710625674,
      "loss": 5.2726,
      "mean_token_accuracy": 0.16128225922584533,
      "num_tokens": 23099431.0,
      "step": 10545
    },
    {
      "entropy": 5.472712421417237,
      "epoch": 1.0470424771734816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004899012778727983,
      "loss": 5.3441,
      "mean_token_accuracy": 0.15000689625740052,
      "num_tokens": 23110947.0,
      "step": 10550
    },
    {
      "entropy": 5.541131973266602,
      "epoch": 1.0475387058356491,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004898907793539406,
      "loss": 5.4248,
      "mean_token_accuracy": 0.15132633745670318,
      "num_tokens": 23122782.0,
      "step": 10555
    },
    {
      "entropy": 5.558660411834717,
      "epoch": 1.0480349344978166,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898802755062546,
      "loss": 5.3262,
      "mean_token_accuracy": 0.1537672460079193,
      "num_tokens": 23134254.0,
      "step": 10560
    },
    {
      "entropy": 5.41802339553833,
      "epoch": 1.048531163159984,
      "grad_norm": 1.0,
      "learning_rate": 0.0004898697663300007,
      "loss": 5.162,
      "mean_token_accuracy": 0.16230653077363968,
      "num_tokens": 23143851.0,
      "step": 10565
    },
    {
      "entropy": 5.466770982742309,
      "epoch": 1.0490273918221515,
      "grad_norm": 1.625,
      "learning_rate": 0.0004898592518254398,
      "loss": 5.3034,
      "mean_token_accuracy": 0.15558066368103027,
      "num_tokens": 23155966.0,
      "step": 10570
    },
    {
      "entropy": 5.446636486053467,
      "epoch": 1.0495236204843192,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004898487319928322,
      "loss": 5.289,
      "mean_token_accuracy": 0.14870069324970245,
      "num_tokens": 23165896.0,
      "step": 10575
    },
    {
      "entropy": 5.46178297996521,
      "epoch": 1.0500198491464867,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004898382068324392,
      "loss": 5.2766,
      "mean_token_accuracy": 0.15289286524057388,
      "num_tokens": 23177821.0,
      "step": 10580
    },
    {
      "entropy": 5.522049045562744,
      "epoch": 1.0505160778086542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004898276763445215,
      "loss": 5.3511,
      "mean_token_accuracy": 0.15181540548801423,
      "num_tokens": 23188916.0,
      "step": 10585
    },
    {
      "entropy": 5.546716737747192,
      "epoch": 1.0510123064708217,
      "grad_norm": 1.125,
      "learning_rate": 0.0004898171405293404,
      "loss": 5.3419,
      "mean_token_accuracy": 0.1544633239507675,
      "num_tokens": 23200490.0,
      "step": 10590
    },
    {
      "entropy": 5.463198900222778,
      "epoch": 1.0515085351329894,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004898065993871571,
      "loss": 5.2917,
      "mean_token_accuracy": 0.15387868136167526,
      "num_tokens": 23212991.0,
      "step": 10595
    },
    {
      "entropy": 5.4983031272888185,
      "epoch": 1.0520047637951568,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000489796052918233,
      "loss": 5.3942,
      "mean_token_accuracy": 0.14896564036607743,
      "num_tokens": 23223361.0,
      "step": 10600
    },
    {
      "entropy": 5.467923259735107,
      "epoch": 1.0525009924573243,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004897855011228295,
      "loss": 5.2993,
      "mean_token_accuracy": 0.15735386610031127,
      "num_tokens": 23233765.0,
      "step": 10605
    },
    {
      "entropy": 5.413602876663208,
      "epoch": 1.0529972211194918,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004897749440012085,
      "loss": 5.2262,
      "mean_token_accuracy": 0.16127143800258636,
      "num_tokens": 23246562.0,
      "step": 10610
    },
    {
      "entropy": 5.460925912857055,
      "epoch": 1.0534934497816595,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004897643815536316,
      "loss": 5.2699,
      "mean_token_accuracy": 0.1560393050312996,
      "num_tokens": 23256648.0,
      "step": 10615
    },
    {
      "entropy": 5.5245355606079105,
      "epoch": 1.053989678443827,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000489753813780361,
      "loss": 5.3052,
      "mean_token_accuracy": 0.14950273782014847,
      "num_tokens": 23267428.0,
      "step": 10620
    },
    {
      "entropy": 5.45337986946106,
      "epoch": 1.0544859071059944,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004897432406816585,
      "loss": 5.2404,
      "mean_token_accuracy": 0.15737607926130295,
      "num_tokens": 23278440.0,
      "step": 10625
    },
    {
      "entropy": 5.475413370132446,
      "epoch": 1.054982135768162,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897326622577863,
      "loss": 5.2896,
      "mean_token_accuracy": 0.15250853449106216,
      "num_tokens": 23290571.0,
      "step": 10630
    },
    {
      "entropy": 5.577261877059937,
      "epoch": 1.0554783644303294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004897220785090067,
      "loss": 5.38,
      "mean_token_accuracy": 0.15308707654476167,
      "num_tokens": 23301634.0,
      "step": 10635
    },
    {
      "entropy": 5.434312868118286,
      "epoch": 1.055974593092497,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004897114894355824,
      "loss": 5.2368,
      "mean_token_accuracy": 0.16050716489553452,
      "num_tokens": 23313125.0,
      "step": 10640
    },
    {
      "entropy": 5.403558254241943,
      "epoch": 1.0564708217546646,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004897008950377758,
      "loss": 5.2587,
      "mean_token_accuracy": 0.15637502074241638,
      "num_tokens": 23323610.0,
      "step": 10645
    },
    {
      "entropy": 5.382852506637573,
      "epoch": 1.056967050416832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004896902953158497,
      "loss": 5.2387,
      "mean_token_accuracy": 0.1538662537932396,
      "num_tokens": 23333652.0,
      "step": 10650
    },
    {
      "entropy": 5.450886821746826,
      "epoch": 1.0574632790789995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004896796902700667,
      "loss": 5.283,
      "mean_token_accuracy": 0.1573830544948578,
      "num_tokens": 23345954.0,
      "step": 10655
    },
    {
      "entropy": 5.4799257755279545,
      "epoch": 1.0579595077411672,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896690799006901,
      "loss": 5.2624,
      "mean_token_accuracy": 0.1561126857995987,
      "num_tokens": 23357052.0,
      "step": 10660
    },
    {
      "entropy": 5.477872514724732,
      "epoch": 1.0584557364033347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896584642079828,
      "loss": 5.2708,
      "mean_token_accuracy": 0.16045353710651397,
      "num_tokens": 23368837.0,
      "step": 10665
    },
    {
      "entropy": 5.441685199737549,
      "epoch": 1.0589519650655022,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004896478431922083,
      "loss": 5.2468,
      "mean_token_accuracy": 0.16128868162631987,
      "num_tokens": 23378608.0,
      "step": 10670
    },
    {
      "entropy": 5.4744476795196535,
      "epoch": 1.0594481937276696,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004896372168536297,
      "loss": 5.3347,
      "mean_token_accuracy": 0.1591661587357521,
      "num_tokens": 23390094.0,
      "step": 10675
    },
    {
      "entropy": 5.406071519851684,
      "epoch": 1.0599444223898373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004896265851925107,
      "loss": 5.1992,
      "mean_token_accuracy": 0.1580475613474846,
      "num_tokens": 23399991.0,
      "step": 10680
    },
    {
      "entropy": 5.4658348083496096,
      "epoch": 1.0604406510520048,
      "grad_norm": 1.0,
      "learning_rate": 0.0004896159482091148,
      "loss": 5.379,
      "mean_token_accuracy": 0.14630398601293565,
      "num_tokens": 23410705.0,
      "step": 10685
    },
    {
      "entropy": 5.468998670578003,
      "epoch": 1.0609368797141723,
      "grad_norm": 1.125,
      "learning_rate": 0.0004896053059037058,
      "loss": 5.1873,
      "mean_token_accuracy": 0.17023698836565018,
      "num_tokens": 23421939.0,
      "step": 10690
    },
    {
      "entropy": 5.457175970077515,
      "epoch": 1.0614331083763398,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004895946582765477,
      "loss": 5.3215,
      "mean_token_accuracy": 0.14924784377217293,
      "num_tokens": 23433252.0,
      "step": 10695
    },
    {
      "entropy": 5.558178853988648,
      "epoch": 1.0619293370385074,
      "grad_norm": 1.0,
      "learning_rate": 0.0004895840053279045,
      "loss": 5.3848,
      "mean_token_accuracy": 0.14604132026433944,
      "num_tokens": 23445582.0,
      "step": 10700
    },
    {
      "entropy": 5.518541860580444,
      "epoch": 1.062425565700675,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004895733470580404,
      "loss": 5.362,
      "mean_token_accuracy": 0.14813369810581206,
      "num_tokens": 23456360.0,
      "step": 10705
    },
    {
      "entropy": 5.45947036743164,
      "epoch": 1.0629217943628424,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004895626834672196,
      "loss": 5.225,
      "mean_token_accuracy": 0.15297651439905166,
      "num_tokens": 23466576.0,
      "step": 10710
    },
    {
      "entropy": 5.387494659423828,
      "epoch": 1.0634180230250099,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004895520145557066,
      "loss": 5.1809,
      "mean_token_accuracy": 0.16365393102169037,
      "num_tokens": 23476200.0,
      "step": 10715
    },
    {
      "entropy": 5.488885116577149,
      "epoch": 1.0639142516871773,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004895413403237659,
      "loss": 5.382,
      "mean_token_accuracy": 0.1556076154112816,
      "num_tokens": 23487776.0,
      "step": 10720
    },
    {
      "entropy": 5.435475635528564,
      "epoch": 1.064410480349345,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004895306607716624,
      "loss": 5.2401,
      "mean_token_accuracy": 0.16200999021530152,
      "num_tokens": 23497517.0,
      "step": 10725
    },
    {
      "entropy": 5.4859312057495115,
      "epoch": 1.0649067090115125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004895199758996607,
      "loss": 5.2669,
      "mean_token_accuracy": 0.1625832661986351,
      "num_tokens": 23507436.0,
      "step": 10730
    },
    {
      "entropy": 5.477930355072021,
      "epoch": 1.06540293767368,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004895092857080258,
      "loss": 5.2095,
      "mean_token_accuracy": 0.14888105168938637,
      "num_tokens": 23517170.0,
      "step": 10735
    },
    {
      "entropy": 5.4167698383331295,
      "epoch": 1.0658991663358475,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004894985901970229,
      "loss": 5.2061,
      "mean_token_accuracy": 0.16213526874780654,
      "num_tokens": 23527329.0,
      "step": 10740
    },
    {
      "entropy": 5.471194696426392,
      "epoch": 1.0663953949980152,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004894878893669171,
      "loss": 5.3408,
      "mean_token_accuracy": 0.1489834502339363,
      "num_tokens": 23538341.0,
      "step": 10745
    },
    {
      "entropy": 5.478268241882324,
      "epoch": 1.0668916236601826,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004894771832179739,
      "loss": 5.2421,
      "mean_token_accuracy": 0.15732278525829316,
      "num_tokens": 23549522.0,
      "step": 10750
    },
    {
      "entropy": 5.480682611465454,
      "epoch": 1.0673878523223501,
      "grad_norm": 1.125,
      "learning_rate": 0.0004894664717504587,
      "loss": 5.3016,
      "mean_token_accuracy": 0.15180493593215943,
      "num_tokens": 23559274.0,
      "step": 10755
    },
    {
      "entropy": 5.462503433227539,
      "epoch": 1.0678840809845176,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004894557549646371,
      "loss": 5.2541,
      "mean_token_accuracy": 0.16153133660554886,
      "num_tokens": 23570852.0,
      "step": 10760
    },
    {
      "entropy": 5.428463411331177,
      "epoch": 1.0683803096466853,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004894450328607749,
      "loss": 5.261,
      "mean_token_accuracy": 0.15809893161058425,
      "num_tokens": 23582918.0,
      "step": 10765
    },
    {
      "entropy": 5.405192947387695,
      "epoch": 1.0688765383088528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004894343054391379,
      "loss": 5.2942,
      "mean_token_accuracy": 0.16206492781639098,
      "num_tokens": 23594177.0,
      "step": 10770
    },
    {
      "entropy": 5.44320821762085,
      "epoch": 1.0693727669710202,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004894235726999924,
      "loss": 5.2552,
      "mean_token_accuracy": 0.1557729348540306,
      "num_tokens": 23605901.0,
      "step": 10775
    },
    {
      "entropy": 5.443066930770874,
      "epoch": 1.0698689956331877,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004894128346436042,
      "loss": 5.3,
      "mean_token_accuracy": 0.1573389694094658,
      "num_tokens": 23617354.0,
      "step": 10780
    },
    {
      "entropy": 5.4787092208862305,
      "epoch": 1.0703652242953554,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004894020912702397,
      "loss": 5.324,
      "mean_token_accuracy": 0.15245613753795623,
      "num_tokens": 23627603.0,
      "step": 10785
    },
    {
      "entropy": 5.363583040237427,
      "epoch": 1.0708614529575229,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004893913425801653,
      "loss": 5.14,
      "mean_token_accuracy": 0.16555351465940477,
      "num_tokens": 23637280.0,
      "step": 10790
    },
    {
      "entropy": 5.4262675762176515,
      "epoch": 1.0713576816196904,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004893805885736477,
      "loss": 5.3488,
      "mean_token_accuracy": 0.15274374037981034,
      "num_tokens": 23649326.0,
      "step": 10795
    },
    {
      "entropy": 5.547499084472657,
      "epoch": 1.0718539102818578,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004893698292509534,
      "loss": 5.3633,
      "mean_token_accuracy": 0.1530749499797821,
      "num_tokens": 23659561.0,
      "step": 10800
    },
    {
      "entropy": 5.493286609649658,
      "epoch": 1.0723501389440253,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004893590646123492,
      "loss": 5.2528,
      "mean_token_accuracy": 0.15883885622024535,
      "num_tokens": 23673206.0,
      "step": 10805
    },
    {
      "entropy": 5.438396310806274,
      "epoch": 1.072846367606193,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004893482946581022,
      "loss": 5.3068,
      "mean_token_accuracy": 0.15375095009803771,
      "num_tokens": 23683920.0,
      "step": 10810
    },
    {
      "entropy": 5.425176763534546,
      "epoch": 1.0733425962683605,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004893375193884793,
      "loss": 5.2276,
      "mean_token_accuracy": 0.16384204477071762,
      "num_tokens": 23693990.0,
      "step": 10815
    },
    {
      "entropy": 5.541811132431031,
      "epoch": 1.073838824930528,
      "grad_norm": 1.125,
      "learning_rate": 0.0004893267388037478,
      "loss": 5.3253,
      "mean_token_accuracy": 0.15232133120298386,
      "num_tokens": 23704870.0,
      "step": 10820
    },
    {
      "entropy": 5.443168020248413,
      "epoch": 1.0743350535926954,
      "grad_norm": 1.21875,
      "learning_rate": 0.000489315952904175,
      "loss": 5.3189,
      "mean_token_accuracy": 0.15552222728729248,
      "num_tokens": 23715159.0,
      "step": 10825
    },
    {
      "entropy": 5.4529327869415285,
      "epoch": 1.0748312822548631,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004893051616900283,
      "loss": 5.237,
      "mean_token_accuracy": 0.15299863517284393,
      "num_tokens": 23725252.0,
      "step": 10830
    },
    {
      "entropy": 5.41984920501709,
      "epoch": 1.0753275109170306,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004892943651615756,
      "loss": 5.2676,
      "mean_token_accuracy": 0.161262047290802,
      "num_tokens": 23736472.0,
      "step": 10835
    },
    {
      "entropy": 5.453280067443847,
      "epoch": 1.075823739579198,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004892835633190843,
      "loss": 5.1748,
      "mean_token_accuracy": 0.1539055213332176,
      "num_tokens": 23747565.0,
      "step": 10840
    },
    {
      "entropy": 5.460154247283936,
      "epoch": 1.0763199682413656,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892727561628223,
      "loss": 5.2733,
      "mean_token_accuracy": 0.1554302155971527,
      "num_tokens": 23758201.0,
      "step": 10845
    },
    {
      "entropy": 5.48637580871582,
      "epoch": 1.0768161969035333,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004892619436930577,
      "loss": 5.3215,
      "mean_token_accuracy": 0.1610202595591545,
      "num_tokens": 23769087.0,
      "step": 10850
    },
    {
      "entropy": 5.461966180801392,
      "epoch": 1.0773124255657007,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004892511259100585,
      "loss": 5.325,
      "mean_token_accuracy": 0.15464484244585036,
      "num_tokens": 23780598.0,
      "step": 10855
    },
    {
      "entropy": 5.455869579315186,
      "epoch": 1.0778086542278682,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004892403028140932,
      "loss": 5.248,
      "mean_token_accuracy": 0.1552821636199951,
      "num_tokens": 23791449.0,
      "step": 10860
    },
    {
      "entropy": 5.455142402648926,
      "epoch": 1.0783048828900357,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048922947440543,
      "loss": 5.2345,
      "mean_token_accuracy": 0.15825797766447067,
      "num_tokens": 23803071.0,
      "step": 10865
    },
    {
      "entropy": 5.528193044662475,
      "epoch": 1.0788011115522032,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004892186406843374,
      "loss": 5.3066,
      "mean_token_accuracy": 0.15140582099556923,
      "num_tokens": 23813372.0,
      "step": 10870
    },
    {
      "entropy": 5.451062297821045,
      "epoch": 1.0792973402143708,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004892078016510841,
      "loss": 5.2872,
      "mean_token_accuracy": 0.1561243087053299,
      "num_tokens": 23824979.0,
      "step": 10875
    },
    {
      "entropy": 5.450916814804077,
      "epoch": 1.0797935688765383,
      "grad_norm": 1.0625,
      "learning_rate": 0.000489196957305939,
      "loss": 5.2756,
      "mean_token_accuracy": 0.15553804039955138,
      "num_tokens": 23836335.0,
      "step": 10880
    },
    {
      "entropy": 5.462556266784668,
      "epoch": 1.0802897975387058,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004891861076491709,
      "loss": 5.2489,
      "mean_token_accuracy": 0.1570493221282959,
      "num_tokens": 23848147.0,
      "step": 10885
    },
    {
      "entropy": 5.400393867492676,
      "epoch": 1.0807860262008733,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004891752526810488,
      "loss": 5.1968,
      "mean_token_accuracy": 0.16255719363689422,
      "num_tokens": 23859502.0,
      "step": 10890
    },
    {
      "entropy": 5.474165725708008,
      "epoch": 1.081282254863041,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000489164392401842,
      "loss": 5.365,
      "mean_token_accuracy": 0.1561533987522125,
      "num_tokens": 23870433.0,
      "step": 10895
    },
    {
      "entropy": 5.491074895858764,
      "epoch": 1.0817784835252084,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004891535268118196,
      "loss": 5.1902,
      "mean_token_accuracy": 0.1606265053153038,
      "num_tokens": 23879581.0,
      "step": 10900
    },
    {
      "entropy": 5.36847677230835,
      "epoch": 1.082274712187376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004891426559112514,
      "loss": 5.2435,
      "mean_token_accuracy": 0.1624028652906418,
      "num_tokens": 23890889.0,
      "step": 10905
    },
    {
      "entropy": 5.449931955337524,
      "epoch": 1.0827709408495434,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004891317797004066,
      "loss": 5.23,
      "mean_token_accuracy": 0.16299335807561874,
      "num_tokens": 23901577.0,
      "step": 10910
    },
    {
      "entropy": 5.418058633804321,
      "epoch": 1.083267169511711,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004891208981795552,
      "loss": 5.3161,
      "mean_token_accuracy": 0.15882405042648315,
      "num_tokens": 23912698.0,
      "step": 10915
    },
    {
      "entropy": 5.38012204170227,
      "epoch": 1.0837633981738786,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004891100113489669,
      "loss": 5.1951,
      "mean_token_accuracy": 0.16206218302249908,
      "num_tokens": 23923056.0,
      "step": 10920
    },
    {
      "entropy": 5.47346863746643,
      "epoch": 1.084259626836046,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004890991192089116,
      "loss": 5.3346,
      "mean_token_accuracy": 0.15577610433101655,
      "num_tokens": 23933998.0,
      "step": 10925
    },
    {
      "entropy": 5.524325323104859,
      "epoch": 1.0847558554982135,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004890882217596595,
      "loss": 5.2224,
      "mean_token_accuracy": 0.16234321594238282,
      "num_tokens": 23943980.0,
      "step": 10930
    },
    {
      "entropy": 5.451502561569214,
      "epoch": 1.085252084160381,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004890773190014809,
      "loss": 5.2997,
      "mean_token_accuracy": 0.15383434295654297,
      "num_tokens": 23955646.0,
      "step": 10935
    },
    {
      "entropy": 5.399806308746338,
      "epoch": 1.0857483128225487,
      "grad_norm": 1.046875,
      "learning_rate": 0.000489066410934646,
      "loss": 5.2639,
      "mean_token_accuracy": 0.1613013505935669,
      "num_tokens": 23966381.0,
      "step": 10940
    },
    {
      "entropy": 5.545250701904297,
      "epoch": 1.0862445414847162,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004890554975594255,
      "loss": 5.3428,
      "mean_token_accuracy": 0.14589644968509674,
      "num_tokens": 23978669.0,
      "step": 10945
    },
    {
      "entropy": 5.607036828994751,
      "epoch": 1.0867407701468836,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004890445788760898,
      "loss": 5.4318,
      "mean_token_accuracy": 0.14692081958055497,
      "num_tokens": 23989467.0,
      "step": 10950
    },
    {
      "entropy": 5.395125246047973,
      "epoch": 1.0872369988090511,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004890336548849098,
      "loss": 5.2108,
      "mean_token_accuracy": 0.1551661491394043,
      "num_tokens": 24000797.0,
      "step": 10955
    },
    {
      "entropy": 5.4902256488800045,
      "epoch": 1.0877332274712188,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004890227255861564,
      "loss": 5.3674,
      "mean_token_accuracy": 0.15386535674333573,
      "num_tokens": 24013182.0,
      "step": 10960
    },
    {
      "entropy": 5.495345497131348,
      "epoch": 1.0882294561333863,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004890117909801005,
      "loss": 5.2489,
      "mean_token_accuracy": 0.15925555527210236,
      "num_tokens": 24024005.0,
      "step": 10965
    },
    {
      "entropy": 5.436870002746582,
      "epoch": 1.0887256847955538,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004890008510670134,
      "loss": 5.2511,
      "mean_token_accuracy": 0.1584889307618141,
      "num_tokens": 24034524.0,
      "step": 10970
    },
    {
      "entropy": 5.453549861907959,
      "epoch": 1.0892219134577212,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004889899058471663,
      "loss": 5.3038,
      "mean_token_accuracy": 0.16058337539434434,
      "num_tokens": 24045944.0,
      "step": 10975
    },
    {
      "entropy": 5.543638610839844,
      "epoch": 1.089718142119889,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004889789553208306,
      "loss": 5.3755,
      "mean_token_accuracy": 0.15898314416408538,
      "num_tokens": 24058612.0,
      "step": 10980
    },
    {
      "entropy": 5.502096033096313,
      "epoch": 1.0902143707820564,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000488967999488278,
      "loss": 5.239,
      "mean_token_accuracy": 0.1605779632925987,
      "num_tokens": 24068912.0,
      "step": 10985
    },
    {
      "entropy": 5.390856504440308,
      "epoch": 1.0907105994442239,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048895703834978,
      "loss": 5.2362,
      "mean_token_accuracy": 0.16214291155338287,
      "num_tokens": 24079879.0,
      "step": 10990
    },
    {
      "entropy": 5.460503387451172,
      "epoch": 1.0912068281063914,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004889460719056085,
      "loss": 5.266,
      "mean_token_accuracy": 0.15560845881700516,
      "num_tokens": 24091268.0,
      "step": 10995
    },
    {
      "entropy": 5.419476890563965,
      "epoch": 1.091703056768559,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004889351001560354,
      "loss": 5.2868,
      "mean_token_accuracy": 0.15701774507761002,
      "num_tokens": 24101428.0,
      "step": 11000
    },
    {
      "entropy": 5.432912302017212,
      "epoch": 1.0921992854307265,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889241231013329,
      "loss": 5.2584,
      "mean_token_accuracy": 0.16208694875240326,
      "num_tokens": 24112920.0,
      "step": 11005
    },
    {
      "entropy": 5.413339424133301,
      "epoch": 1.092695514092894,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000488913140741773,
      "loss": 5.2035,
      "mean_token_accuracy": 0.157074174284935,
      "num_tokens": 24125731.0,
      "step": 11010
    },
    {
      "entropy": 5.376967191696167,
      "epoch": 1.0931917427550615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004889021530776282,
      "loss": 5.2047,
      "mean_token_accuracy": 0.1598724752664566,
      "num_tokens": 24136606.0,
      "step": 11015
    },
    {
      "entropy": 5.475081586837769,
      "epoch": 1.093687971417229,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004888911601091709,
      "loss": 5.2165,
      "mean_token_accuracy": 0.16434352844953537,
      "num_tokens": 24146783.0,
      "step": 11020
    },
    {
      "entropy": 5.439157724380493,
      "epoch": 1.0941842000793967,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004888801618366737,
      "loss": 5.3209,
      "mean_token_accuracy": 0.15623154491186142,
      "num_tokens": 24157412.0,
      "step": 11025
    },
    {
      "entropy": 5.462990617752075,
      "epoch": 1.0946804287415641,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004888691582604095,
      "loss": 5.3143,
      "mean_token_accuracy": 0.15274647772312164,
      "num_tokens": 24168077.0,
      "step": 11030
    },
    {
      "entropy": 5.504316806793213,
      "epoch": 1.0951766574037316,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000488858149380651,
      "loss": 5.3604,
      "mean_token_accuracy": 0.154996919631958,
      "num_tokens": 24178913.0,
      "step": 11035
    },
    {
      "entropy": 5.534307146072388,
      "epoch": 1.095672886065899,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000488847135197671,
      "loss": 5.404,
      "mean_token_accuracy": 0.14685525894165039,
      "num_tokens": 24190356.0,
      "step": 11040
    },
    {
      "entropy": 5.527395009994507,
      "epoch": 1.0961691147280668,
      "grad_norm": 1.078125,
      "learning_rate": 0.000488836115711743,
      "loss": 5.3543,
      "mean_token_accuracy": 0.15240919068455697,
      "num_tokens": 24199772.0,
      "step": 11045
    },
    {
      "entropy": 5.514452075958252,
      "epoch": 1.0966653433902342,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004888250909231401,
      "loss": 5.3009,
      "mean_token_accuracy": 0.153182964771986,
      "num_tokens": 24211807.0,
      "step": 11050
    },
    {
      "entropy": 5.4453723430633545,
      "epoch": 1.0971615720524017,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004888140608321357,
      "loss": 5.2519,
      "mean_token_accuracy": 0.16168430745601653,
      "num_tokens": 24223156.0,
      "step": 11055
    },
    {
      "entropy": 5.504981660842896,
      "epoch": 1.0976578007145692,
      "grad_norm": 1.125,
      "learning_rate": 0.0004888030254390033,
      "loss": 5.3067,
      "mean_token_accuracy": 0.1493518441915512,
      "num_tokens": 24234352.0,
      "step": 11060
    },
    {
      "entropy": 5.470440864562988,
      "epoch": 1.098154029376737,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004887919847440166,
      "loss": 5.2327,
      "mean_token_accuracy": 0.16139766424894333,
      "num_tokens": 24244137.0,
      "step": 11065
    },
    {
      "entropy": 5.380899667739868,
      "epoch": 1.0986502580389044,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004887809387474494,
      "loss": 5.2373,
      "mean_token_accuracy": 0.15720967203378677,
      "num_tokens": 24253250.0,
      "step": 11070
    },
    {
      "entropy": 5.490319633483887,
      "epoch": 1.0991464867010718,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004887698874495755,
      "loss": 5.3579,
      "mean_token_accuracy": 0.1561625674366951,
      "num_tokens": 24264551.0,
      "step": 11075
    },
    {
      "entropy": 5.4714947700500485,
      "epoch": 1.0996427153632393,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000488758830850669,
      "loss": 5.1778,
      "mean_token_accuracy": 0.1629674255847931,
      "num_tokens": 24275522.0,
      "step": 11080
    },
    {
      "entropy": 5.4344175338745115,
      "epoch": 1.100138944025407,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004887477689510042,
      "loss": 5.3339,
      "mean_token_accuracy": 0.155118890106678,
      "num_tokens": 24286333.0,
      "step": 11085
    },
    {
      "entropy": 5.411440372467041,
      "epoch": 1.1006351726875745,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004887367017508553,
      "loss": 5.3211,
      "mean_token_accuracy": 0.15520284026861192,
      "num_tokens": 24297627.0,
      "step": 11090
    },
    {
      "entropy": 5.4355206966400145,
      "epoch": 1.101131401349742,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004887256292504967,
      "loss": 5.2731,
      "mean_token_accuracy": 0.1553366869688034,
      "num_tokens": 24308246.0,
      "step": 11095
    },
    {
      "entropy": 5.485727071762085,
      "epoch": 1.1016276300119094,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000488714551450203,
      "loss": 5.3317,
      "mean_token_accuracy": 0.14475269615650177,
      "num_tokens": 24317718.0,
      "step": 11100
    },
    {
      "entropy": 5.495961380004883,
      "epoch": 1.102123858674077,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000488703468350249,
      "loss": 5.2468,
      "mean_token_accuracy": 0.15890163332223892,
      "num_tokens": 24329242.0,
      "step": 11105
    },
    {
      "entropy": 5.531225633621216,
      "epoch": 1.1026200873362446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004886923799509095,
      "loss": 5.4268,
      "mean_token_accuracy": 0.14818619042634965,
      "num_tokens": 24340146.0,
      "step": 11110
    },
    {
      "entropy": 5.514152956008911,
      "epoch": 1.103116315998412,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004886812862524595,
      "loss": 5.3058,
      "mean_token_accuracy": 0.1534282833337784,
      "num_tokens": 24349370.0,
      "step": 11115
    },
    {
      "entropy": 5.548317384719849,
      "epoch": 1.1036125446605796,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004886701872551739,
      "loss": 5.2789,
      "mean_token_accuracy": 0.15132023617625237,
      "num_tokens": 24359770.0,
      "step": 11120
    },
    {
      "entropy": 5.448024415969849,
      "epoch": 1.104108773322747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004886590829593281,
      "loss": 5.3084,
      "mean_token_accuracy": 0.15640848577022554,
      "num_tokens": 24370974.0,
      "step": 11125
    },
    {
      "entropy": 5.453985643386841,
      "epoch": 1.1046050019849147,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004886479733651974,
      "loss": 5.3177,
      "mean_token_accuracy": 0.15365736186504364,
      "num_tokens": 24381771.0,
      "step": 11130
    },
    {
      "entropy": 5.4032796859741214,
      "epoch": 1.1051012306470822,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004886368584730575,
      "loss": 5.2727,
      "mean_token_accuracy": 0.15119307339191437,
      "num_tokens": 24393067.0,
      "step": 11135
    },
    {
      "entropy": 5.477257966995239,
      "epoch": 1.1055974593092497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004886257382831837,
      "loss": 5.3445,
      "mean_token_accuracy": 0.15073836371302604,
      "num_tokens": 24402866.0,
      "step": 11140
    },
    {
      "entropy": 5.541209888458252,
      "epoch": 1.1060936879714172,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000488614612795852,
      "loss": 5.2808,
      "mean_token_accuracy": 0.1595071643590927,
      "num_tokens": 24412956.0,
      "step": 11145
    },
    {
      "entropy": 5.5288714408874515,
      "epoch": 1.1065899166335849,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886034820113382,
      "loss": 5.3312,
      "mean_token_accuracy": 0.15079039782285691,
      "num_tokens": 24424650.0,
      "step": 11150
    },
    {
      "entropy": 5.490571355819702,
      "epoch": 1.1070861452957523,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004885923459299182,
      "loss": 5.3517,
      "mean_token_accuracy": 0.15200869292020797,
      "num_tokens": 24436541.0,
      "step": 11155
    },
    {
      "entropy": 5.510240888595581,
      "epoch": 1.1075823739579198,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004885812045518685,
      "loss": 5.3339,
      "mean_token_accuracy": 0.15058897137641908,
      "num_tokens": 24447588.0,
      "step": 11160
    },
    {
      "entropy": 5.4990355491638185,
      "epoch": 1.1080786026200873,
      "grad_norm": 1.0625,
      "learning_rate": 0.000488570057877465,
      "loss": 5.243,
      "mean_token_accuracy": 0.1611657977104187,
      "num_tokens": 24458854.0,
      "step": 11165
    },
    {
      "entropy": 5.456546497344971,
      "epoch": 1.108574831282255,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004885589059069842,
      "loss": 5.3179,
      "mean_token_accuracy": 0.1550692707300186,
      "num_tokens": 24469427.0,
      "step": 11170
    },
    {
      "entropy": 5.426618814468384,
      "epoch": 1.1090710599444225,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004885477486407028,
      "loss": 5.3116,
      "mean_token_accuracy": 0.15141913294792175,
      "num_tokens": 24479352.0,
      "step": 11175
    },
    {
      "entropy": 5.579872083663941,
      "epoch": 1.10956728860659,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004885365860788974,
      "loss": 5.307,
      "mean_token_accuracy": 0.15684168934822082,
      "num_tokens": 24489580.0,
      "step": 11180
    },
    {
      "entropy": 5.449553394317627,
      "epoch": 1.1100635172687574,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004885254182218448,
      "loss": 5.2589,
      "mean_token_accuracy": 0.15555014014244078,
      "num_tokens": 24501141.0,
      "step": 11185
    },
    {
      "entropy": 5.403925180435181,
      "epoch": 1.1105597459309249,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004885142450698219,
      "loss": 5.2687,
      "mean_token_accuracy": 0.1562810182571411,
      "num_tokens": 24512079.0,
      "step": 11190
    },
    {
      "entropy": 5.481047916412353,
      "epoch": 1.1110559745930926,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004885030666231057,
      "loss": 5.3136,
      "mean_token_accuracy": 0.15418094992637635,
      "num_tokens": 24521839.0,
      "step": 11195
    },
    {
      "entropy": 5.525297927856445,
      "epoch": 1.11155220325526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004884918828819738,
      "loss": 5.3483,
      "mean_token_accuracy": 0.15666994899511338,
      "num_tokens": 24532704.0,
      "step": 11200
    },
    {
      "entropy": 5.504165267944336,
      "epoch": 1.1120484319174275,
      "grad_norm": 1.0625,
      "learning_rate": 0.000488480693846703,
      "loss": 5.2636,
      "mean_token_accuracy": 0.159215197712183,
      "num_tokens": 24543790.0,
      "step": 11205
    },
    {
      "entropy": 5.445603132247925,
      "epoch": 1.112544660579595,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004884694995175711,
      "loss": 5.3145,
      "mean_token_accuracy": 0.15664611160755157,
      "num_tokens": 24554765.0,
      "step": 11210
    },
    {
      "entropy": 5.58459153175354,
      "epoch": 1.1130408892417627,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004884582998948555,
      "loss": 5.3975,
      "mean_token_accuracy": 0.14638945534825326,
      "num_tokens": 24565888.0,
      "step": 11215
    },
    {
      "entropy": 5.456152296066284,
      "epoch": 1.1135371179039302,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004884470949788342,
      "loss": 5.255,
      "mean_token_accuracy": 0.15497650504112243,
      "num_tokens": 24575482.0,
      "step": 11220
    },
    {
      "entropy": 5.506486511230468,
      "epoch": 1.1140333465660976,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004884358847697847,
      "loss": 5.3778,
      "mean_token_accuracy": 0.14256357103586198,
      "num_tokens": 24588249.0,
      "step": 11225
    },
    {
      "entropy": 5.5314994812011715,
      "epoch": 1.1145295752282651,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004884246692679852,
      "loss": 5.319,
      "mean_token_accuracy": 0.1584802120923996,
      "num_tokens": 24599546.0,
      "step": 11230
    },
    {
      "entropy": 5.454435682296753,
      "epoch": 1.1150258038904326,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004884134484737139,
      "loss": 5.2535,
      "mean_token_accuracy": 0.15498768538236618,
      "num_tokens": 24611777.0,
      "step": 11235
    },
    {
      "entropy": 5.431316471099853,
      "epoch": 1.1155220325526003,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004884022223872489,
      "loss": 5.241,
      "mean_token_accuracy": 0.15740324705839157,
      "num_tokens": 24623675.0,
      "step": 11240
    },
    {
      "entropy": 5.475382232666016,
      "epoch": 1.1160182612147678,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004883909910088686,
      "loss": 5.2565,
      "mean_token_accuracy": 0.15649086236953735,
      "num_tokens": 24633929.0,
      "step": 11245
    },
    {
      "entropy": 5.439922714233399,
      "epoch": 1.1165144898769352,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004883797543388515,
      "loss": 5.2249,
      "mean_token_accuracy": 0.16053610816597938,
      "num_tokens": 24642800.0,
      "step": 11250
    },
    {
      "entropy": 5.4203695297241214,
      "epoch": 1.1170107185391027,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004883685123774764,
      "loss": 5.2288,
      "mean_token_accuracy": 0.15934738367795945,
      "num_tokens": 24653780.0,
      "step": 11255
    },
    {
      "entropy": 5.434363126754761,
      "epoch": 1.1175069472012704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004883572651250218,
      "loss": 5.2456,
      "mean_token_accuracy": 0.15415403470396996,
      "num_tokens": 24663912.0,
      "step": 11260
    },
    {
      "entropy": 5.4572563648223875,
      "epoch": 1.118003175863438,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004883460125817669,
      "loss": 5.3486,
      "mean_token_accuracy": 0.15579213052988053,
      "num_tokens": 24675222.0,
      "step": 11265
    },
    {
      "entropy": 5.530616760253906,
      "epoch": 1.1184994045256054,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004883347547479905,
      "loss": 5.3558,
      "mean_token_accuracy": 0.15403239876031877,
      "num_tokens": 24686697.0,
      "step": 11270
    },
    {
      "entropy": 5.5334470748901365,
      "epoch": 1.1189956331877728,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000488323491623972,
      "loss": 5.2733,
      "mean_token_accuracy": 0.1599091187119484,
      "num_tokens": 24697067.0,
      "step": 11275
    },
    {
      "entropy": 5.500115251541137,
      "epoch": 1.1194918618499405,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004883122232099905,
      "loss": 5.2786,
      "mean_token_accuracy": 0.1548775777220726,
      "num_tokens": 24707989.0,
      "step": 11280
    },
    {
      "entropy": 5.392176485061645,
      "epoch": 1.119988090512108,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004883009495063256,
      "loss": 5.2226,
      "mean_token_accuracy": 0.16512216925621032,
      "num_tokens": 24718826.0,
      "step": 11285
    },
    {
      "entropy": 5.4927661418914795,
      "epoch": 1.1204843191742755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004882896705132566,
      "loss": 5.3178,
      "mean_token_accuracy": 0.15257630199193956,
      "num_tokens": 24729913.0,
      "step": 11290
    },
    {
      "entropy": 5.423000240325928,
      "epoch": 1.120980547836443,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048827838623106355,
      "loss": 5.1806,
      "mean_token_accuracy": 0.16208620071411134,
      "num_tokens": 24741621.0,
      "step": 11295
    },
    {
      "entropy": 5.434277105331421,
      "epoch": 1.1214767764986107,
      "grad_norm": 1.046875,
      "learning_rate": 0.000488267096660026,
      "loss": 5.2452,
      "mean_token_accuracy": 0.15617795139551163,
      "num_tokens": 24753427.0,
      "step": 11300
    },
    {
      "entropy": 5.462497711181641,
      "epoch": 1.1219730051607781,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048825580180042406,
      "loss": 5.3062,
      "mean_token_accuracy": 0.15538909286260605,
      "num_tokens": 24763580.0,
      "step": 11305
    },
    {
      "entropy": 5.464938831329346,
      "epoch": 1.1224692338229456,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004882445016525377,
      "loss": 5.2776,
      "mean_token_accuracy": 0.15929068103432656,
      "num_tokens": 24773955.0,
      "step": 11310
    },
    {
      "entropy": 5.4187171936035154,
      "epoch": 1.122965462485113,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048823319621664723,
      "loss": 5.1935,
      "mean_token_accuracy": 0.1642131805419922,
      "num_tokens": 24784039.0,
      "step": 11315
    },
    {
      "entropy": 5.520082330703735,
      "epoch": 1.1234616911472806,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048822188549303296,
      "loss": 5.301,
      "mean_token_accuracy": 0.1521339997649193,
      "num_tokens": 24794630.0,
      "step": 11320
    },
    {
      "entropy": 5.492739534378051,
      "epoch": 1.1239579198094483,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004882105694819754,
      "loss": 5.3355,
      "mean_token_accuracy": 0.15180297791957856,
      "num_tokens": 24805625.0,
      "step": 11325
    },
    {
      "entropy": 5.512796068191529,
      "epoch": 1.1244541484716157,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004881992481837551,
      "loss": 5.2797,
      "mean_token_accuracy": 0.160688117146492,
      "num_tokens": 24817356.0,
      "step": 11330
    },
    {
      "entropy": 5.447163486480713,
      "epoch": 1.1249503771337832,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004881879215986529,
      "loss": 5.2474,
      "mean_token_accuracy": 0.146769430488348,
      "num_tokens": 24828349.0,
      "step": 11335
    },
    {
      "entropy": 5.464562845230103,
      "epoch": 1.1254466057959507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048817658972694967,
      "loss": 5.2898,
      "mean_token_accuracy": 0.1592294231057167,
      "num_tokens": 24840138.0,
      "step": 11340
    },
    {
      "entropy": 5.482720279693604,
      "epoch": 1.1259428344581184,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004881652525689263,
      "loss": 5.2811,
      "mean_token_accuracy": 0.15433214753866195,
      "num_tokens": 24851570.0,
      "step": 11345
    },
    {
      "entropy": 5.3687985897064205,
      "epoch": 1.1264390631202859,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000488153910124864,
      "loss": 5.1901,
      "mean_token_accuracy": 0.16005050241947175,
      "num_tokens": 24861396.0,
      "step": 11350
    },
    {
      "entropy": 5.460654973983765,
      "epoch": 1.1269352917824533,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000488142562395044,
      "loss": 5.2985,
      "mean_token_accuracy": 0.14988509118556975,
      "num_tokens": 24873519.0,
      "step": 11355
    },
    {
      "entropy": 5.515922689437867,
      "epoch": 1.1274315204446208,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048813120937974774,
      "loss": 5.3301,
      "mean_token_accuracy": 0.15362645983695983,
      "num_tokens": 24886856.0,
      "step": 11360
    },
    {
      "entropy": 5.43823766708374,
      "epoch": 1.1279277491067885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048811985107925684,
      "loss": 5.2727,
      "mean_token_accuracy": 0.1524615228176117,
      "num_tokens": 24897292.0,
      "step": 11365
    },
    {
      "entropy": 5.450088167190552,
      "epoch": 1.128423977768956,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048810848749385286,
      "loss": 5.2759,
      "mean_token_accuracy": 0.15722230076789856,
      "num_tokens": 24907757.0,
      "step": 11370
    },
    {
      "entropy": 5.414643096923828,
      "epoch": 1.1289202064311235,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048809711862381753,
      "loss": 5.2547,
      "mean_token_accuracy": 0.1595899485051632,
      "num_tokens": 24918071.0,
      "step": 11375
    },
    {
      "entropy": 5.44346432685852,
      "epoch": 1.129416435093291,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048808574446943274,
      "loss": 5.246,
      "mean_token_accuracy": 0.15908097922801973,
      "num_tokens": 24927967.0,
      "step": 11380
    },
    {
      "entropy": 5.483906316757202,
      "epoch": 1.1299126637554586,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004880743650309808,
      "loss": 5.3125,
      "mean_token_accuracy": 0.16103851795196533,
      "num_tokens": 24939170.0,
      "step": 11385
    },
    {
      "entropy": 5.452262258529663,
      "epoch": 1.130408892417626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048806298030874373,
      "loss": 5.3009,
      "mean_token_accuracy": 0.15707661658525468,
      "num_tokens": 24950937.0,
      "step": 11390
    },
    {
      "entropy": 5.487660980224609,
      "epoch": 1.1309051210797936,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048805159030300383,
      "loss": 5.2867,
      "mean_token_accuracy": 0.1550215482711792,
      "num_tokens": 24962813.0,
      "step": 11395
    },
    {
      "entropy": 5.435828781127929,
      "epoch": 1.131401349741961,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048804019501404357,
      "loss": 5.1876,
      "mean_token_accuracy": 0.15997025221586228,
      "num_tokens": 24974192.0,
      "step": 11400
    },
    {
      "entropy": 5.42454891204834,
      "epoch": 1.1318975784041285,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048802879444214553,
      "loss": 5.1675,
      "mean_token_accuracy": 0.1638774111866951,
      "num_tokens": 24985353.0,
      "step": 11405
    },
    {
      "entropy": 5.443878030776977,
      "epoch": 1.1323938070662962,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048801738858759243,
      "loss": 5.282,
      "mean_token_accuracy": 0.1630396142601967,
      "num_tokens": 24997259.0,
      "step": 11410
    },
    {
      "entropy": 5.44436354637146,
      "epoch": 1.1328900357284637,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004880059774506671,
      "loss": 5.2639,
      "mean_token_accuracy": 0.1533152535557747,
      "num_tokens": 25008222.0,
      "step": 11415
    },
    {
      "entropy": 5.465674543380738,
      "epoch": 1.1333862643906312,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004879945610316525,
      "loss": 5.3071,
      "mean_token_accuracy": 0.15470327734947203,
      "num_tokens": 25020267.0,
      "step": 11420
    },
    {
      "entropy": 5.4833605766296385,
      "epoch": 1.1338824930527986,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048798313933083175,
      "loss": 5.28,
      "mean_token_accuracy": 0.15688425451517105,
      "num_tokens": 25031212.0,
      "step": 11425
    },
    {
      "entropy": 5.453396368026733,
      "epoch": 1.1343787217149663,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004879717123484882,
      "loss": 5.2632,
      "mean_token_accuracy": 0.15119045674800874,
      "num_tokens": 25042204.0,
      "step": 11430
    },
    {
      "entropy": 5.487443208694458,
      "epoch": 1.1348749503771338,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000487960280084905,
      "loss": 5.2772,
      "mean_token_accuracy": 0.16374703198671342,
      "num_tokens": 25052010.0,
      "step": 11435
    },
    {
      "entropy": 5.492530965805054,
      "epoch": 1.1353711790393013,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004879488425403657,
      "loss": 5.3353,
      "mean_token_accuracy": 0.1498436562716961,
      "num_tokens": 25062209.0,
      "step": 11440
    },
    {
      "entropy": 5.426764869689942,
      "epoch": 1.1358674077014688,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048793739971515407,
      "loss": 5.3035,
      "mean_token_accuracy": 0.1568688541650772,
      "num_tokens": 25072758.0,
      "step": 11445
    },
    {
      "entropy": 5.4608433723449705,
      "epoch": 1.1363636363636362,
      "grad_norm": 1.125,
      "learning_rate": 0.00048792595160955374,
      "loss": 5.3644,
      "mean_token_accuracy": 0.15430160611867905,
      "num_tokens": 25084318.0,
      "step": 11450
    },
    {
      "entropy": 5.546369934082032,
      "epoch": 1.136859865025804,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048791449822384856,
      "loss": 5.2565,
      "mean_token_accuracy": 0.1596905842423439,
      "num_tokens": 25095715.0,
      "step": 11455
    },
    {
      "entropy": 5.524097633361817,
      "epoch": 1.1373560936879714,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004879030395583226,
      "loss": 5.2803,
      "mean_token_accuracy": 0.15604273974895477,
      "num_tokens": 25106163.0,
      "step": 11460
    },
    {
      "entropy": 5.485153532028198,
      "epoch": 1.137852322350139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004878915756132601,
      "loss": 5.3381,
      "mean_token_accuracy": 0.15305267423391342,
      "num_tokens": 25118745.0,
      "step": 11465
    },
    {
      "entropy": 5.463880395889282,
      "epoch": 1.1383485510123066,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048788010638894523,
      "loss": 5.3424,
      "mean_token_accuracy": 0.14730097204446793,
      "num_tokens": 25130066.0,
      "step": 11470
    },
    {
      "entropy": 5.432196998596192,
      "epoch": 1.138844779674474,
      "grad_norm": 1.125,
      "learning_rate": 0.0004878686318856624,
      "loss": 5.1562,
      "mean_token_accuracy": 0.16133119463920592,
      "num_tokens": 25139895.0,
      "step": 11475
    },
    {
      "entropy": 5.5185362815856935,
      "epoch": 1.1393410083366415,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048785715210369635,
      "loss": 5.3953,
      "mean_token_accuracy": 0.14914272204041482,
      "num_tokens": 25150644.0,
      "step": 11480
    },
    {
      "entropy": 5.487786722183228,
      "epoch": 1.139837236998809,
      "grad_norm": 1.0,
      "learning_rate": 0.0004878456670433314,
      "loss": 5.3111,
      "mean_token_accuracy": 0.15341423004865645,
      "num_tokens": 25163038.0,
      "step": 11485
    },
    {
      "entropy": 5.463690519332886,
      "epoch": 1.1403334656609765,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004878341767048527,
      "loss": 5.2288,
      "mean_token_accuracy": 0.1608066141605377,
      "num_tokens": 25174154.0,
      "step": 11490
    },
    {
      "entropy": 5.525653123855591,
      "epoch": 1.1408296943231442,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048782268108854494,
      "loss": 5.3382,
      "mean_token_accuracy": 0.14976068288087846,
      "num_tokens": 25185369.0,
      "step": 11495
    },
    {
      "entropy": 5.4214763164520265,
      "epoch": 1.1413259229853117,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004878111801946933,
      "loss": 5.3506,
      "mean_token_accuracy": 0.15307861864566802,
      "num_tokens": 25197875.0,
      "step": 11500
    },
    {
      "entropy": 5.409812641143799,
      "epoch": 1.1418221516474791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048779967402358296,
      "loss": 5.2224,
      "mean_token_accuracy": 0.16769037544727325,
      "num_tokens": 25208362.0,
      "step": 11505
    },
    {
      "entropy": 5.55489411354065,
      "epoch": 1.1423183803096466,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004877881625754993,
      "loss": 5.4137,
      "mean_token_accuracy": 0.14632707163691522,
      "num_tokens": 25220441.0,
      "step": 11510
    },
    {
      "entropy": 5.390773296356201,
      "epoch": 1.1428146089718143,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048777664585072765,
      "loss": 5.1931,
      "mean_token_accuracy": 0.16045051217079162,
      "num_tokens": 25231134.0,
      "step": 11515
    },
    {
      "entropy": 5.506291484832763,
      "epoch": 1.1433108376339818,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004877651238495537,
      "loss": 5.44,
      "mean_token_accuracy": 0.15050924941897392,
      "num_tokens": 25242327.0,
      "step": 11520
    },
    {
      "entropy": 5.5009935855865475,
      "epoch": 1.1438070662961493,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048775359657226316,
      "loss": 5.2736,
      "mean_token_accuracy": 0.15368279367685317,
      "num_tokens": 25253571.0,
      "step": 11525
    },
    {
      "entropy": 5.42882137298584,
      "epoch": 1.1443032949583167,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004877420640191418,
      "loss": 5.2436,
      "mean_token_accuracy": 0.15605375170707703,
      "num_tokens": 25264573.0,
      "step": 11530
    },
    {
      "entropy": 5.479844808578491,
      "epoch": 1.1447995236204842,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004877305261904758,
      "loss": 5.318,
      "mean_token_accuracy": 0.1551748126745224,
      "num_tokens": 25275197.0,
      "step": 11535
    },
    {
      "entropy": 5.4426939487457275,
      "epoch": 1.145295752282652,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048771898308655107,
      "loss": 5.1929,
      "mean_token_accuracy": 0.163699235022068,
      "num_tokens": 25286564.0,
      "step": 11540
    },
    {
      "entropy": 5.472001934051514,
      "epoch": 1.1457919809448194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004877074347076539,
      "loss": 5.2407,
      "mean_token_accuracy": 0.15860285013914108,
      "num_tokens": 25297874.0,
      "step": 11545
    },
    {
      "entropy": 5.487885618209839,
      "epoch": 1.1462882096069869,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048769588105407075,
      "loss": 5.3151,
      "mean_token_accuracy": 0.15399954319000245,
      "num_tokens": 25308047.0,
      "step": 11550
    },
    {
      "entropy": 5.5249897003173825,
      "epoch": 1.1467844382691545,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048768432212608807,
      "loss": 5.352,
      "mean_token_accuracy": 0.14692576974630356,
      "num_tokens": 25318527.0,
      "step": 11555
    },
    {
      "entropy": 5.502056503295899,
      "epoch": 1.147280666931322,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004876727579239925,
      "loss": 5.2724,
      "mean_token_accuracy": 0.16056878119707108,
      "num_tokens": 25328242.0,
      "step": 11560
    },
    {
      "entropy": 5.4387226581573485,
      "epoch": 1.1477768955934895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004876611884480708,
      "loss": 5.247,
      "mean_token_accuracy": 0.15914848148822786,
      "num_tokens": 25338214.0,
      "step": 11565
    },
    {
      "entropy": 5.4934248447418215,
      "epoch": 1.148273124255657,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048764961369860977,
      "loss": 5.3058,
      "mean_token_accuracy": 0.155112823843956,
      "num_tokens": 25349762.0,
      "step": 11570
    },
    {
      "entropy": 5.4696643352508545,
      "epoch": 1.1487693529178244,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048763803367589663,
      "loss": 5.3017,
      "mean_token_accuracy": 0.15766253024339677,
      "num_tokens": 25361241.0,
      "step": 11575
    },
    {
      "entropy": 5.451131200790405,
      "epoch": 1.1492655815799921,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048762644838021843,
      "loss": 5.3222,
      "mean_token_accuracy": 0.15278395190834998,
      "num_tokens": 25373846.0,
      "step": 11580
    },
    {
      "entropy": 5.422038650512695,
      "epoch": 1.1497618102421596,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048761485781186243,
      "loss": 5.2466,
      "mean_token_accuracy": 0.15599873214960097,
      "num_tokens": 25385331.0,
      "step": 11585
    },
    {
      "entropy": 5.447986316680908,
      "epoch": 1.150258038904327,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004876032619711162,
      "loss": 5.2737,
      "mean_token_accuracy": 0.16089896559715272,
      "num_tokens": 25395140.0,
      "step": 11590
    },
    {
      "entropy": 5.526353788375855,
      "epoch": 1.1507542675664946,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048759166085826714,
      "loss": 5.3287,
      "mean_token_accuracy": 0.15200670212507247,
      "num_tokens": 25405455.0,
      "step": 11595
    },
    {
      "entropy": 5.478128814697266,
      "epoch": 1.1512504962286623,
      "grad_norm": 1.40625,
      "learning_rate": 0.000487580054473603,
      "loss": 5.3317,
      "mean_token_accuracy": 0.15748160630464553,
      "num_tokens": 25417000.0,
      "step": 11600
    },
    {
      "entropy": 5.423494148254394,
      "epoch": 1.1517467248908297,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048756844281741156,
      "loss": 5.2373,
      "mean_token_accuracy": 0.161028254032135,
      "num_tokens": 25427808.0,
      "step": 11605
    },
    {
      "entropy": 5.473474359512329,
      "epoch": 1.1522429535529972,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048755682588998085,
      "loss": 5.27,
      "mean_token_accuracy": 0.15297403335571289,
      "num_tokens": 25438302.0,
      "step": 11610
    },
    {
      "entropy": 5.37503228187561,
      "epoch": 1.1527391822151647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004875452036915987,
      "loss": 5.1566,
      "mean_token_accuracy": 0.16137976348400115,
      "num_tokens": 25449335.0,
      "step": 11615
    },
    {
      "entropy": 5.527887487411499,
      "epoch": 1.1532354108773322,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004875335762225536,
      "loss": 5.3629,
      "mean_token_accuracy": 0.1491509534418583,
      "num_tokens": 25461173.0,
      "step": 11620
    },
    {
      "entropy": 5.535806131362915,
      "epoch": 1.1537316395394999,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048752194348313383,
      "loss": 5.2472,
      "mean_token_accuracy": 0.1580447621643543,
      "num_tokens": 25472538.0,
      "step": 11625
    },
    {
      "entropy": 5.414586925506592,
      "epoch": 1.1542278682016673,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004875103054736278,
      "loss": 5.2899,
      "mean_token_accuracy": 0.15897171944379807,
      "num_tokens": 25483386.0,
      "step": 11630
    },
    {
      "entropy": 5.487336826324463,
      "epoch": 1.1547240968638348,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004874986621943241,
      "loss": 5.3199,
      "mean_token_accuracy": 0.15200241655111313,
      "num_tokens": 25495505.0,
      "step": 11635
    },
    {
      "entropy": 5.469439840316772,
      "epoch": 1.1552203255260023,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048748701364551147,
      "loss": 5.2921,
      "mean_token_accuracy": 0.15704389959573745,
      "num_tokens": 25506144.0,
      "step": 11640
    },
    {
      "entropy": 5.463355636596679,
      "epoch": 1.15571655418817,
      "grad_norm": 1.125,
      "learning_rate": 0.0004874753598274787,
      "loss": 5.2802,
      "mean_token_accuracy": 0.15217718109488487,
      "num_tokens": 25516185.0,
      "step": 11645
    },
    {
      "entropy": 5.553554248809815,
      "epoch": 1.1562127828503375,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004874637007405149,
      "loss": 5.3634,
      "mean_token_accuracy": 0.15233521461486815,
      "num_tokens": 25526657.0,
      "step": 11650
    },
    {
      "entropy": 5.436895513534546,
      "epoch": 1.156709011512505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048745203638490917,
      "loss": 5.2288,
      "mean_token_accuracy": 0.1625049576163292,
      "num_tokens": 25537101.0,
      "step": 11655
    },
    {
      "entropy": 5.396178722381592,
      "epoch": 1.1572052401746724,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048744036676095067,
      "loss": 5.3135,
      "mean_token_accuracy": 0.15204431414604186,
      "num_tokens": 25547672.0,
      "step": 11660
    },
    {
      "entropy": 5.572168445587158,
      "epoch": 1.15770146883684,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004874286918689289,
      "loss": 5.3647,
      "mean_token_accuracy": 0.15444307029247284,
      "num_tokens": 25558408.0,
      "step": 11665
    },
    {
      "entropy": 5.537053108215332,
      "epoch": 1.1581976974990076,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048741701170913326,
      "loss": 5.2757,
      "mean_token_accuracy": 0.15206874310970306,
      "num_tokens": 25569823.0,
      "step": 11670
    },
    {
      "entropy": 5.447519302368164,
      "epoch": 1.158693926161175,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048740532628185353,
      "loss": 5.2715,
      "mean_token_accuracy": 0.1583130106329918,
      "num_tokens": 25579705.0,
      "step": 11675
    },
    {
      "entropy": 5.446307182312012,
      "epoch": 1.1591901548233425,
      "grad_norm": 1.125,
      "learning_rate": 0.0004873936355873793,
      "loss": 5.3459,
      "mean_token_accuracy": 0.14970098733901976,
      "num_tokens": 25590573.0,
      "step": 11680
    },
    {
      "entropy": 5.460212087631225,
      "epoch": 1.1596863834855102,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004873819396260007,
      "loss": 5.2488,
      "mean_token_accuracy": 0.1574046179652214,
      "num_tokens": 25601418.0,
      "step": 11685
    },
    {
      "entropy": 5.502991390228272,
      "epoch": 1.1601826121476777,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048737023839800754,
      "loss": 5.3356,
      "mean_token_accuracy": 0.15483759641647338,
      "num_tokens": 25612409.0,
      "step": 11690
    },
    {
      "entropy": 5.443919229507446,
      "epoch": 1.1606788408098452,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004873585319036901,
      "loss": 5.235,
      "mean_token_accuracy": 0.16078144013881684,
      "num_tokens": 25623547.0,
      "step": 11695
    },
    {
      "entropy": 5.476826858520508,
      "epoch": 1.1611750694720127,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004873468201433387,
      "loss": 5.4024,
      "mean_token_accuracy": 0.14993560314178467,
      "num_tokens": 25636537.0,
      "step": 11700
    },
    {
      "entropy": 5.510700273513794,
      "epoch": 1.1616712981341801,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004873351031172437,
      "loss": 5.2631,
      "mean_token_accuracy": 0.1606941193342209,
      "num_tokens": 25646734.0,
      "step": 11705
    },
    {
      "entropy": 5.5217937469482425,
      "epoch": 1.1621675267963478,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048732338082569564,
      "loss": 5.3611,
      "mean_token_accuracy": 0.15862078219652176,
      "num_tokens": 25658058.0,
      "step": 11710
    },
    {
      "entropy": 5.438635969161988,
      "epoch": 1.1626637554585153,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004873116532689853,
      "loss": 5.1834,
      "mean_token_accuracy": 0.16246994733810424,
      "num_tokens": 25669630.0,
      "step": 11715
    },
    {
      "entropy": 5.368061685562134,
      "epoch": 1.1631599841206828,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004872999204474035,
      "loss": 5.2035,
      "mean_token_accuracy": 0.15695789009332656,
      "num_tokens": 25681928.0,
      "step": 11720
    },
    {
      "entropy": 5.376949691772461,
      "epoch": 1.1636562127828503,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000487288182361241,
      "loss": 5.2062,
      "mean_token_accuracy": 0.16101142168045043,
      "num_tokens": 25690653.0,
      "step": 11725
    },
    {
      "entropy": 5.535763072967529,
      "epoch": 1.164152441445018,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004872764390107891,
      "loss": 5.3203,
      "mean_token_accuracy": 0.14774771630764008,
      "num_tokens": 25701201.0,
      "step": 11730
    },
    {
      "entropy": 5.4236232280731205,
      "epoch": 1.1646486701071854,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048726469039633895,
      "loss": 5.2217,
      "mean_token_accuracy": 0.1604251116514206,
      "num_tokens": 25710927.0,
      "step": 11735
    },
    {
      "entropy": 5.445154428482056,
      "epoch": 1.165144898769353,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048725293651818183,
      "loss": 5.2836,
      "mean_token_accuracy": 0.1559119626879692,
      "num_tokens": 25722136.0,
      "step": 11740
    },
    {
      "entropy": 5.427000331878662,
      "epoch": 1.1656411274315204,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004872411773766093,
      "loss": 5.1561,
      "mean_token_accuracy": 0.167077474296093,
      "num_tokens": 25733386.0,
      "step": 11745
    },
    {
      "entropy": 5.395485830307007,
      "epoch": 1.1661373560936878,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048722941297191284,
      "loss": 5.2531,
      "mean_token_accuracy": 0.1571579933166504,
      "num_tokens": 25742801.0,
      "step": 11750
    },
    {
      "entropy": 5.514714527130127,
      "epoch": 1.1666335847558555,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004872176433043843,
      "loss": 5.3579,
      "mean_token_accuracy": 0.15434617176651955,
      "num_tokens": 25755594.0,
      "step": 11755
    },
    {
      "entropy": 5.495013809204101,
      "epoch": 1.167129813418023,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004872058683743155,
      "loss": 5.2789,
      "mean_token_accuracy": 0.15668174028396606,
      "num_tokens": 25766271.0,
      "step": 11760
    },
    {
      "entropy": 5.489176511764526,
      "epoch": 1.1676260420801905,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048719408818199836,
      "loss": 5.3363,
      "mean_token_accuracy": 0.1541507363319397,
      "num_tokens": 25777338.0,
      "step": 11765
    },
    {
      "entropy": 5.484441328048706,
      "epoch": 1.1681222707423582,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048718230272772517,
      "loss": 5.2065,
      "mean_token_accuracy": 0.16040641069412231,
      "num_tokens": 25787968.0,
      "step": 11770
    },
    {
      "entropy": 5.461015701293945,
      "epoch": 1.1686184994045257,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048717051201178804,
      "loss": 5.2937,
      "mean_token_accuracy": 0.1613932505249977,
      "num_tokens": 25798274.0,
      "step": 11775
    },
    {
      "entropy": 5.369739437103272,
      "epoch": 1.1691147280666931,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004871587160344794,
      "loss": 5.2767,
      "mean_token_accuracy": 0.15359314531087875,
      "num_tokens": 25809140.0,
      "step": 11780
    },
    {
      "entropy": 5.465685081481934,
      "epoch": 1.1696109567288606,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004871469147960917,
      "loss": 5.3155,
      "mean_token_accuracy": 0.15784543603658677,
      "num_tokens": 25821502.0,
      "step": 11785
    },
    {
      "entropy": 5.485592031478882,
      "epoch": 1.170107185391028,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004871351082969177,
      "loss": 5.1733,
      "mean_token_accuracy": 0.1617198333144188,
      "num_tokens": 25833231.0,
      "step": 11790
    },
    {
      "entropy": 5.52383074760437,
      "epoch": 1.1706034140531958,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004871232965372501,
      "loss": 5.3626,
      "mean_token_accuracy": 0.15265980064868928,
      "num_tokens": 25844867.0,
      "step": 11795
    },
    {
      "entropy": 5.470095491409301,
      "epoch": 1.1710996427153633,
      "grad_norm": 1.25,
      "learning_rate": 0.0004871114795173819,
      "loss": 5.3209,
      "mean_token_accuracy": 0.1609694853425026,
      "num_tokens": 25855959.0,
      "step": 11800
    },
    {
      "entropy": 5.472194004058838,
      "epoch": 1.1715958713775307,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004870996572376061,
      "loss": 5.4074,
      "mean_token_accuracy": 0.1549374654889107,
      "num_tokens": 25866926.0,
      "step": 11805
    },
    {
      "entropy": 5.57244725227356,
      "epoch": 1.1720921000396982,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004870878296982157,
      "loss": 5.3589,
      "mean_token_accuracy": 0.15105366110801696,
      "num_tokens": 25877699.0,
      "step": 11810
    },
    {
      "entropy": 5.560077714920044,
      "epoch": 1.172588328701866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004870759968995042,
      "loss": 5.2794,
      "mean_token_accuracy": 0.16240156888961793,
      "num_tokens": 25888640.0,
      "step": 11815
    },
    {
      "entropy": 5.446789073944092,
      "epoch": 1.1730845573640334,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048706415884176493,
      "loss": 5.2802,
      "mean_token_accuracy": 0.15544371157884598,
      "num_tokens": 25900651.0,
      "step": 11820
    },
    {
      "entropy": 5.4609541416168215,
      "epoch": 1.1735807860262009,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004870523155252915,
      "loss": 5.2562,
      "mean_token_accuracy": 0.1606263667345047,
      "num_tokens": 25911775.0,
      "step": 11825
    },
    {
      "entropy": 5.454714822769165,
      "epoch": 1.1740770146883683,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004870404669503776,
      "loss": 5.2263,
      "mean_token_accuracy": 0.1595461130142212,
      "num_tokens": 25922126.0,
      "step": 11830
    },
    {
      "entropy": 5.472031116485596,
      "epoch": 1.1745732433505358,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000487028613117317,
      "loss": 5.2827,
      "mean_token_accuracy": 0.1567227929830551,
      "num_tokens": 25932945.0,
      "step": 11835
    },
    {
      "entropy": 5.489677000045776,
      "epoch": 1.1750694720127035,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004870167540264037,
      "loss": 5.3367,
      "mean_token_accuracy": 0.15337247252464295,
      "num_tokens": 25944277.0,
      "step": 11840
    },
    {
      "entropy": 5.437878227233886,
      "epoch": 1.175565700674871,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004870048896779317,
      "loss": 5.1901,
      "mean_token_accuracy": 0.16377597004175187,
      "num_tokens": 25956838.0,
      "step": 11845
    },
    {
      "entropy": 5.4647480010986325,
      "epoch": 1.1760619293370385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048699302007219533,
      "loss": 5.2576,
      "mean_token_accuracy": 0.15625229477882385,
      "num_tokens": 25966639.0,
      "step": 11850
    },
    {
      "entropy": 5.447669744491577,
      "epoch": 1.1765581579992062,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004869811452094889,
      "loss": 5.3012,
      "mean_token_accuracy": 0.16445562094449998,
      "num_tokens": 25978209.0,
      "step": 11855
    },
    {
      "entropy": 5.396578168869018,
      "epoch": 1.1770543866613736,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048696926509010673,
      "loss": 5.239,
      "mean_token_accuracy": 0.16259317398071288,
      "num_tokens": 25989306.0,
      "step": 11860
    },
    {
      "entropy": 5.432253026962281,
      "epoch": 1.177550615323541,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004869573797143437,
      "loss": 5.3353,
      "mean_token_accuracy": 0.1516199916601181,
      "num_tokens": 26000235.0,
      "step": 11865
    },
    {
      "entropy": 5.4656126499176025,
      "epoch": 1.1780468439857086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004869454890824942,
      "loss": 5.2328,
      "mean_token_accuracy": 0.16042409241199493,
      "num_tokens": 26010616.0,
      "step": 11870
    },
    {
      "entropy": 5.469156980514526,
      "epoch": 1.178543072647876,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048693359319485337,
      "loss": 5.3181,
      "mean_token_accuracy": 0.15194744765758514,
      "num_tokens": 26021303.0,
      "step": 11875
    },
    {
      "entropy": 5.457164525985718,
      "epoch": 1.1790393013100438,
      "grad_norm": 1.0,
      "learning_rate": 0.0004869216920517161,
      "loss": 5.245,
      "mean_token_accuracy": 0.15838841199874878,
      "num_tokens": 26032386.0,
      "step": 11880
    },
    {
      "entropy": 5.380097198486328,
      "epoch": 1.1795355299722112,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048690978565337753,
      "loss": 5.1986,
      "mean_token_accuracy": 0.15893911719322204,
      "num_tokens": 26042889.0,
      "step": 11885
    },
    {
      "entropy": 5.483877944946289,
      "epoch": 1.1800317586343787,
      "grad_norm": 1.125,
      "learning_rate": 0.00048689787400013287,
      "loss": 5.3213,
      "mean_token_accuracy": 0.15259282886981965,
      "num_tokens": 26054299.0,
      "step": 11890
    },
    {
      "entropy": 5.52237343788147,
      "epoch": 1.1805279872965462,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004868859570922775,
      "loss": 5.3439,
      "mean_token_accuracy": 0.1520078107714653,
      "num_tokens": 26064587.0,
      "step": 11895
    },
    {
      "entropy": 5.506268882751465,
      "epoch": 1.1810242159587139,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004868740349301071,
      "loss": 5.2694,
      "mean_token_accuracy": 0.16637126058340074,
      "num_tokens": 26076495.0,
      "step": 11900
    },
    {
      "entropy": 5.45013198852539,
      "epoch": 1.1815204446208813,
      "grad_norm": 1.140625,
      "learning_rate": 0.000486862107513917,
      "loss": 5.2769,
      "mean_token_accuracy": 0.1554826468229294,
      "num_tokens": 26086496.0,
      "step": 11905
    },
    {
      "entropy": 5.419335412979126,
      "epoch": 1.1820166732830488,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004868501748440033,
      "loss": 5.2661,
      "mean_token_accuracy": 0.15864377617835998,
      "num_tokens": 26097263.0,
      "step": 11910
    },
    {
      "entropy": 5.488437509536743,
      "epoch": 1.1825129019452163,
      "grad_norm": 1.125,
      "learning_rate": 0.0004868382369206617,
      "loss": 5.2648,
      "mean_token_accuracy": 0.15624529719352723,
      "num_tokens": 26108264.0,
      "step": 11915
    },
    {
      "entropy": 5.481942796707154,
      "epoch": 1.1830091306073838,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004868262937441883,
      "loss": 5.3273,
      "mean_token_accuracy": 0.16052356511354446,
      "num_tokens": 26119730.0,
      "step": 11920
    },
    {
      "entropy": 5.435774183273315,
      "epoch": 1.1835053592695515,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048681434531487925,
      "loss": 5.22,
      "mean_token_accuracy": 0.1556362360715866,
      "num_tokens": 26131055.0,
      "step": 11925
    },
    {
      "entropy": 5.362377500534057,
      "epoch": 1.184001587931719,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048680239163303085,
      "loss": 5.2349,
      "mean_token_accuracy": 0.1595527246594429,
      "num_tokens": 26141577.0,
      "step": 11930
    },
    {
      "entropy": 5.4207220554351805,
      "epoch": 1.1844978165938864,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004867904326989396,
      "loss": 5.2734,
      "mean_token_accuracy": 0.15625436902046203,
      "num_tokens": 26152733.0,
      "step": 11935
    },
    {
      "entropy": 5.4987890243530275,
      "epoch": 1.184994045256054,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004867784685129019,
      "loss": 5.326,
      "mean_token_accuracy": 0.15520620197057725,
      "num_tokens": 26164763.0,
      "step": 11940
    },
    {
      "entropy": 5.42806224822998,
      "epoch": 1.1854902739182216,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048676649907521455,
      "loss": 5.1822,
      "mean_token_accuracy": 0.15977985560894012,
      "num_tokens": 26174391.0,
      "step": 11945
    },
    {
      "entropy": 5.354955196380615,
      "epoch": 1.185986502580389,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048675452438617434,
      "loss": 5.1671,
      "mean_token_accuracy": 0.16593013256788253,
      "num_tokens": 26184772.0,
      "step": 11950
    },
    {
      "entropy": 5.527537059783936,
      "epoch": 1.1864827312425565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004867425444460782,
      "loss": 5.3377,
      "mean_token_accuracy": 0.1457195945084095,
      "num_tokens": 26195411.0,
      "step": 11955
    },
    {
      "entropy": 5.39585657119751,
      "epoch": 1.186978959904724,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004867305592552233,
      "loss": 5.2503,
      "mean_token_accuracy": 0.1577274650335312,
      "num_tokens": 26206719.0,
      "step": 11960
    },
    {
      "entropy": 5.418163919448853,
      "epoch": 1.1874751885668917,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004867185688139067,
      "loss": 5.1819,
      "mean_token_accuracy": 0.163595113158226,
      "num_tokens": 26217469.0,
      "step": 11965
    },
    {
      "entropy": 5.357501745223999,
      "epoch": 1.1879714172290592,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004867065731224258,
      "loss": 5.2161,
      "mean_token_accuracy": 0.15580180138349534,
      "num_tokens": 26228556.0,
      "step": 11970
    },
    {
      "entropy": 5.43242621421814,
      "epoch": 1.1884676458912267,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004866945721810781,
      "loss": 5.3254,
      "mean_token_accuracy": 0.1595786213874817,
      "num_tokens": 26239891.0,
      "step": 11975
    },
    {
      "entropy": 5.487291240692139,
      "epoch": 1.1889638745533941,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004866825659901612,
      "loss": 5.2814,
      "mean_token_accuracy": 0.16263574957847596,
      "num_tokens": 26251793.0,
      "step": 11980
    },
    {
      "entropy": 5.377985858917237,
      "epoch": 1.1894601032155618,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048667055454997276,
      "loss": 5.165,
      "mean_token_accuracy": 0.16263482421636583,
      "num_tokens": 26262367.0,
      "step": 11985
    },
    {
      "entropy": 5.38570761680603,
      "epoch": 1.1899563318777293,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048665853786081063,
      "loss": 5.2562,
      "mean_token_accuracy": 0.1547637924551964,
      "num_tokens": 26272681.0,
      "step": 11990
    },
    {
      "entropy": 5.454520893096924,
      "epoch": 1.1904525605398968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048664651592297294,
      "loss": 5.3224,
      "mean_token_accuracy": 0.15762927532196044,
      "num_tokens": 26281194.0,
      "step": 11995
    },
    {
      "entropy": 5.481055879592896,
      "epoch": 1.1909487892020643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048663448873675766,
      "loss": 5.216,
      "mean_token_accuracy": 0.1549502730369568,
      "num_tokens": 26292682.0,
      "step": 12000
    },
    {
      "epoch": 1.1909487892020643,
      "eval_entropy": 5.36984923617181,
      "eval_loss": 5.418365955352783,
      "eval_mean_token_accuracy": 0.15612898885152293,
      "eval_num_tokens": 26292682.0,
      "eval_runtime": 26.7352,
      "eval_samples_per_second": 1259.352,
      "eval_steps_per_second": 157.433,
      "step": 12000
    },
    {
      "entropy": 5.467216491699219,
      "epoch": 1.1914450178642317,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004866224563024631,
      "loss": 5.3171,
      "mean_token_accuracy": 0.15042169094085694,
      "num_tokens": 26305500.0,
      "step": 12005
    },
    {
      "entropy": 5.459606790542603,
      "epoch": 1.1919412465263994,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004866104186203877,
      "loss": 5.2342,
      "mean_token_accuracy": 0.15368821769952773,
      "num_tokens": 26317207.0,
      "step": 12010
    },
    {
      "entropy": 5.450319528579712,
      "epoch": 1.192437475188567,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048659837569082977,
      "loss": 5.2367,
      "mean_token_accuracy": 0.16350379288196565,
      "num_tokens": 26328728.0,
      "step": 12015
    },
    {
      "entropy": 5.494319152832031,
      "epoch": 1.1929337038507344,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004865863275140882,
      "loss": 5.2824,
      "mean_token_accuracy": 0.1590137556195259,
      "num_tokens": 26338622.0,
      "step": 12020
    },
    {
      "entropy": 5.434467363357544,
      "epoch": 1.1934299325129019,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004865742740904615,
      "loss": 5.275,
      "mean_token_accuracy": 0.15567538738250733,
      "num_tokens": 26348807.0,
      "step": 12025
    },
    {
      "entropy": 5.371427583694458,
      "epoch": 1.1939261611750696,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004865622154202488,
      "loss": 5.2191,
      "mean_token_accuracy": 0.15983818769454955,
      "num_tokens": 26358481.0,
      "step": 12030
    },
    {
      "entropy": 5.467825984954834,
      "epoch": 1.194422389837237,
      "grad_norm": 1.0,
      "learning_rate": 0.00048655015150374904,
      "loss": 5.309,
      "mean_token_accuracy": 0.15003999173641205,
      "num_tokens": 26370519.0,
      "step": 12035
    },
    {
      "entropy": 5.535327005386352,
      "epoch": 1.1949186184994045,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004865380823412613,
      "loss": 5.3006,
      "mean_token_accuracy": 0.16214030832052231,
      "num_tokens": 26380988.0,
      "step": 12040
    },
    {
      "entropy": 5.396744632720948,
      "epoch": 1.195414847161572,
      "grad_norm": 1.046875,
      "learning_rate": 0.000486526007933085,
      "loss": 5.1947,
      "mean_token_accuracy": 0.1666797786951065,
      "num_tokens": 26390308.0,
      "step": 12045
    },
    {
      "entropy": 5.425592088699341,
      "epoch": 1.1959110758237395,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048651392827951944,
      "loss": 5.3025,
      "mean_token_accuracy": 0.14808757305145265,
      "num_tokens": 26401145.0,
      "step": 12050
    },
    {
      "entropy": 5.518236780166626,
      "epoch": 1.1964073044859072,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048650184338086436,
      "loss": 5.2444,
      "mean_token_accuracy": 0.15794756561517714,
      "num_tokens": 26411930.0,
      "step": 12055
    },
    {
      "entropy": 5.4554167747497555,
      "epoch": 1.1969035331480746,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048648975323741916,
      "loss": 5.1862,
      "mean_token_accuracy": 0.15503725558519363,
      "num_tokens": 26422320.0,
      "step": 12060
    },
    {
      "entropy": 5.511041784286499,
      "epoch": 1.197399761810242,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048647765784948395,
      "loss": 5.3152,
      "mean_token_accuracy": 0.15139880031347275,
      "num_tokens": 26433760.0,
      "step": 12065
    },
    {
      "entropy": 5.508217906951904,
      "epoch": 1.1978959904724098,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004864655572173584,
      "loss": 5.2314,
      "mean_token_accuracy": 0.16251991987228392,
      "num_tokens": 26444256.0,
      "step": 12070
    },
    {
      "entropy": 5.435006618499756,
      "epoch": 1.1983922191345773,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048645345134134274,
      "loss": 5.3236,
      "mean_token_accuracy": 0.15879910439252853,
      "num_tokens": 26456433.0,
      "step": 12075
    },
    {
      "entropy": 5.4997878074646,
      "epoch": 1.1988884477967447,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048644134022173707,
      "loss": 5.2844,
      "mean_token_accuracy": 0.15812809616327286,
      "num_tokens": 26466909.0,
      "step": 12080
    },
    {
      "entropy": 5.490894174575805,
      "epoch": 1.1993846764589122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004864292238588419,
      "loss": 5.283,
      "mean_token_accuracy": 0.162313973903656,
      "num_tokens": 26478785.0,
      "step": 12085
    },
    {
      "entropy": 5.403077507019043,
      "epoch": 1.1998809051210797,
      "grad_norm": 1.125,
      "learning_rate": 0.00048641710225295746,
      "loss": 5.3103,
      "mean_token_accuracy": 0.16022782772779465,
      "num_tokens": 26489596.0,
      "step": 12090
    },
    {
      "entropy": 5.456048679351807,
      "epoch": 1.2003771337832474,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048640497540438455,
      "loss": 5.2357,
      "mean_token_accuracy": 0.16017713844776155,
      "num_tokens": 26500376.0,
      "step": 12095
    },
    {
      "entropy": 5.415544557571411,
      "epoch": 1.2008733624454149,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004863928433134237,
      "loss": 5.1638,
      "mean_token_accuracy": 0.16080620735883713,
      "num_tokens": 26510339.0,
      "step": 12100
    },
    {
      "entropy": 5.467069101333618,
      "epoch": 1.2013695911075823,
      "grad_norm": 1.0,
      "learning_rate": 0.00048638070598037585,
      "loss": 5.2643,
      "mean_token_accuracy": 0.15873608887195587,
      "num_tokens": 26521359.0,
      "step": 12105
    },
    {
      "entropy": 5.527863597869873,
      "epoch": 1.2018658197697498,
      "grad_norm": 1.15625,
      "learning_rate": 0.000486368563405542,
      "loss": 5.3448,
      "mean_token_accuracy": 0.1558574303984642,
      "num_tokens": 26532716.0,
      "step": 12110
    },
    {
      "entropy": 5.426868629455567,
      "epoch": 1.2023620484319175,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004863564155892232,
      "loss": 5.185,
      "mean_token_accuracy": 0.15701333284378052,
      "num_tokens": 26542624.0,
      "step": 12115
    },
    {
      "entropy": 5.463325691223145,
      "epoch": 1.202858277094085,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004863442625317208,
      "loss": 5.3031,
      "mean_token_accuracy": 0.16040219217538834,
      "num_tokens": 26552907.0,
      "step": 12120
    },
    {
      "entropy": 5.390181636810302,
      "epoch": 1.2033545057562525,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048633210423333607,
      "loss": 5.2353,
      "mean_token_accuracy": 0.15812619924545288,
      "num_tokens": 26564871.0,
      "step": 12125
    },
    {
      "entropy": 5.404935359954834,
      "epoch": 1.20385073441842,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004863199406943705,
      "loss": 5.23,
      "mean_token_accuracy": 0.16342412680387497,
      "num_tokens": 26576254.0,
      "step": 12130
    },
    {
      "entropy": 5.491464710235595,
      "epoch": 1.2043469630805874,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048630777191512584,
      "loss": 5.2788,
      "mean_token_accuracy": 0.16016832888126373,
      "num_tokens": 26587518.0,
      "step": 12135
    },
    {
      "entropy": 5.407528972625732,
      "epoch": 1.2048431917427551,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048629559789590366,
      "loss": 5.1804,
      "mean_token_accuracy": 0.16357915550470353,
      "num_tokens": 26597485.0,
      "step": 12140
    },
    {
      "entropy": 5.414377117156983,
      "epoch": 1.2053394204049226,
      "grad_norm": 1.078125,
      "learning_rate": 0.000486283418637006,
      "loss": 5.2302,
      "mean_token_accuracy": 0.16146788001060486,
      "num_tokens": 26607696.0,
      "step": 12145
    },
    {
      "entropy": 5.502345371246338,
      "epoch": 1.20583564906709,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004862712341387348,
      "loss": 5.2614,
      "mean_token_accuracy": 0.15649989247322083,
      "num_tokens": 26618974.0,
      "step": 12150
    },
    {
      "entropy": 5.457761478424072,
      "epoch": 1.2063318777292578,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004862590444013923,
      "loss": 5.3045,
      "mean_token_accuracy": 0.15584758222103118,
      "num_tokens": 26629949.0,
      "step": 12155
    },
    {
      "entropy": 5.45476336479187,
      "epoch": 1.2068281063914252,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004862468494252806,
      "loss": 5.2479,
      "mean_token_accuracy": 0.1577911466360092,
      "num_tokens": 26641460.0,
      "step": 12160
    },
    {
      "entropy": 5.506238079071045,
      "epoch": 1.2073243350535927,
      "grad_norm": 1.125,
      "learning_rate": 0.0004862346492107024,
      "loss": 5.3383,
      "mean_token_accuracy": 0.15595106780529022,
      "num_tokens": 26652553.0,
      "step": 12165
    },
    {
      "entropy": 5.432679224014282,
      "epoch": 1.2078205637157602,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048622244375796,
      "loss": 5.2556,
      "mean_token_accuracy": 0.1539401277899742,
      "num_tokens": 26662479.0,
      "step": 12170
    },
    {
      "entropy": 5.477621555328369,
      "epoch": 1.2083167923779277,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004862102330673561,
      "loss": 5.2814,
      "mean_token_accuracy": 0.15400275811553002,
      "num_tokens": 26673326.0,
      "step": 12175
    },
    {
      "entropy": 5.499663639068603,
      "epoch": 1.2088130210400954,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048619801713919357,
      "loss": 5.3433,
      "mean_token_accuracy": 0.1530168578028679,
      "num_tokens": 26684189.0,
      "step": 12180
    },
    {
      "entropy": 5.521960258483887,
      "epoch": 1.2093092497022628,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004861857959737753,
      "loss": 5.3357,
      "mean_token_accuracy": 0.16050476878881453,
      "num_tokens": 26695399.0,
      "step": 12185
    },
    {
      "entropy": 5.5194926261901855,
      "epoch": 1.2098054783644303,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004861735695714044,
      "loss": 5.2915,
      "mean_token_accuracy": 0.15331920832395554,
      "num_tokens": 26705993.0,
      "step": 12190
    },
    {
      "entropy": 5.472894716262817,
      "epoch": 1.2103017070265978,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048616133793238394,
      "loss": 5.2612,
      "mean_token_accuracy": 0.15859028249979018,
      "num_tokens": 26716120.0,
      "step": 12195
    },
    {
      "entropy": 5.465536975860596,
      "epoch": 1.2107979356887655,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048614910105701737,
      "loss": 5.3795,
      "mean_token_accuracy": 0.15031011998653412,
      "num_tokens": 26727736.0,
      "step": 12200
    },
    {
      "entropy": 5.501601505279541,
      "epoch": 1.211294164350933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004861368589456081,
      "loss": 5.2392,
      "mean_token_accuracy": 0.1577613100409508,
      "num_tokens": 26737825.0,
      "step": 12205
    },
    {
      "entropy": 5.53739013671875,
      "epoch": 1.2117903930131004,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004861246115984596,
      "loss": 5.3262,
      "mean_token_accuracy": 0.15262310951948166,
      "num_tokens": 26748035.0,
      "step": 12210
    },
    {
      "entropy": 5.488145256042481,
      "epoch": 1.212286621675268,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048611235901587575,
      "loss": 5.2308,
      "mean_token_accuracy": 0.16155769675970078,
      "num_tokens": 26759899.0,
      "step": 12215
    },
    {
      "entropy": 5.3996055126190186,
      "epoch": 1.2127828503374354,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004861001011981602,
      "loss": 5.1845,
      "mean_token_accuracy": 0.16370288729667665,
      "num_tokens": 26771562.0,
      "step": 12220
    },
    {
      "entropy": 5.477347469329834,
      "epoch": 1.213279078999603,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004860878381456171,
      "loss": 5.3089,
      "mean_token_accuracy": 0.16055908054113388,
      "num_tokens": 26783002.0,
      "step": 12225
    },
    {
      "entropy": 5.469759464263916,
      "epoch": 1.2137753076617706,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048607556985855047,
      "loss": 5.3182,
      "mean_token_accuracy": 0.1572047784924507,
      "num_tokens": 26794136.0,
      "step": 12230
    },
    {
      "entropy": 5.410419607162476,
      "epoch": 1.214271536323938,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048606329633726447,
      "loss": 5.2739,
      "mean_token_accuracy": 0.15901596695184708,
      "num_tokens": 26804776.0,
      "step": 12235
    },
    {
      "entropy": 5.435255479812622,
      "epoch": 1.2147677649861055,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004860510175820636,
      "loss": 5.1976,
      "mean_token_accuracy": 0.16329383105039597,
      "num_tokens": 26816696.0,
      "step": 12240
    },
    {
      "entropy": 5.438877010345459,
      "epoch": 1.2152639936482732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004860387335932522,
      "loss": 5.2413,
      "mean_token_accuracy": 0.16022688150405884,
      "num_tokens": 26827789.0,
      "step": 12245
    },
    {
      "entropy": 5.451368761062622,
      "epoch": 1.2157602223104407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004860264443711351,
      "loss": 5.1949,
      "mean_token_accuracy": 0.15956586897373198,
      "num_tokens": 26838763.0,
      "step": 12250
    },
    {
      "entropy": 5.365965032577515,
      "epoch": 1.2162564509726081,
      "grad_norm": 1.125,
      "learning_rate": 0.0004860141499160167,
      "loss": 5.235,
      "mean_token_accuracy": 0.15303678512573243,
      "num_tokens": 26849304.0,
      "step": 12255
    },
    {
      "entropy": 5.481586599349976,
      "epoch": 1.2167526796347756,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004860018502282022,
      "loss": 5.3047,
      "mean_token_accuracy": 0.15839675515890123,
      "num_tokens": 26858777.0,
      "step": 12260
    },
    {
      "entropy": 5.532920598983765,
      "epoch": 1.2172489082969433,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048598954530799644,
      "loss": 5.2854,
      "mean_token_accuracy": 0.1543389528989792,
      "num_tokens": 26870286.0,
      "step": 12265
    },
    {
      "entropy": 5.509565591812134,
      "epoch": 1.2177451369591108,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048597723515570465,
      "loss": 5.2573,
      "mean_token_accuracy": 0.1629301980137825,
      "num_tokens": 26880485.0,
      "step": 12270
    },
    {
      "entropy": 5.368143510818482,
      "epoch": 1.2182413656212783,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000485964919771632,
      "loss": 5.161,
      "mean_token_accuracy": 0.16418050825595856,
      "num_tokens": 26893556.0,
      "step": 12275
    },
    {
      "entropy": 5.3909220695495605,
      "epoch": 1.2187375942834457,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048595259915608396,
      "loss": 5.2411,
      "mean_token_accuracy": 0.15505783557891845,
      "num_tokens": 26905699.0,
      "step": 12280
    },
    {
      "entropy": 5.470983552932739,
      "epoch": 1.2192338229456134,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048594027330936594,
      "loss": 5.2627,
      "mean_token_accuracy": 0.15373864024877548,
      "num_tokens": 26917188.0,
      "step": 12285
    },
    {
      "entropy": 5.460317707061767,
      "epoch": 1.219730051607781,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048592794223178374,
      "loss": 5.2396,
      "mean_token_accuracy": 0.15770459324121475,
      "num_tokens": 26927905.0,
      "step": 12290
    },
    {
      "entropy": 5.3755223751068115,
      "epoch": 1.2202262802699484,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048591560592364314,
      "loss": 5.2271,
      "mean_token_accuracy": 0.15978424102067948,
      "num_tokens": 26938665.0,
      "step": 12295
    },
    {
      "entropy": 5.46402497291565,
      "epoch": 1.2207225089321159,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004859032643852499,
      "loss": 5.3289,
      "mean_token_accuracy": 0.15560518950223923,
      "num_tokens": 26950516.0,
      "step": 12300
    },
    {
      "entropy": 5.531369543075561,
      "epoch": 1.2212187375942833,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004858909176169102,
      "loss": 5.3122,
      "mean_token_accuracy": 0.16247982680797576,
      "num_tokens": 26960018.0,
      "step": 12305
    },
    {
      "entropy": 5.434589862823486,
      "epoch": 1.221714966256451,
      "grad_norm": 1.125,
      "learning_rate": 0.00048587856561893014,
      "loss": 5.2893,
      "mean_token_accuracy": 0.1528005138039589,
      "num_tokens": 26969825.0,
      "step": 12310
    },
    {
      "entropy": 5.5181420803070065,
      "epoch": 1.2222111949186185,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004858662083916161,
      "loss": 5.3757,
      "mean_token_accuracy": 0.1497234098613262,
      "num_tokens": 26979873.0,
      "step": 12315
    },
    {
      "entropy": 5.553140926361084,
      "epoch": 1.222707423580786,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004858538459352744,
      "loss": 5.3502,
      "mean_token_accuracy": 0.15362712144851684,
      "num_tokens": 26990577.0,
      "step": 12320
    },
    {
      "entropy": 5.425366497039795,
      "epoch": 1.2232036522429535,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004858414782502117,
      "loss": 5.1813,
      "mean_token_accuracy": 0.16415750235319138,
      "num_tokens": 27001598.0,
      "step": 12325
    },
    {
      "entropy": 5.371681499481201,
      "epoch": 1.2236998809051212,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048582910533673474,
      "loss": 5.1977,
      "mean_token_accuracy": 0.16368423998355866,
      "num_tokens": 27012535.0,
      "step": 12330
    },
    {
      "entropy": 5.455473613739014,
      "epoch": 1.2241961095672886,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004858167271951502,
      "loss": 5.2587,
      "mean_token_accuracy": 0.15064289569854736,
      "num_tokens": 27021867.0,
      "step": 12335
    },
    {
      "entropy": 5.394765901565552,
      "epoch": 1.2246923382294561,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004858043438257651,
      "loss": 5.2301,
      "mean_token_accuracy": 0.16016123741865157,
      "num_tokens": 27032540.0,
      "step": 12340
    },
    {
      "entropy": 5.375723838806152,
      "epoch": 1.2251885668916236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004857919552288865,
      "loss": 5.2194,
      "mean_token_accuracy": 0.15854422748088837,
      "num_tokens": 27044008.0,
      "step": 12345
    },
    {
      "entropy": 5.471513986587524,
      "epoch": 1.225684795553791,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004857795614048217,
      "loss": 5.2181,
      "mean_token_accuracy": 0.16486064195632935,
      "num_tokens": 27053646.0,
      "step": 12350
    },
    {
      "entropy": 5.518558692932129,
      "epoch": 1.2261810242159588,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048576716235387786,
      "loss": 5.3532,
      "mean_token_accuracy": 0.15143815353512763,
      "num_tokens": 27065834.0,
      "step": 12355
    },
    {
      "entropy": 5.429163646697998,
      "epoch": 1.2266772528781262,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048575475807636267,
      "loss": 5.294,
      "mean_token_accuracy": 0.15594286620616912,
      "num_tokens": 27077544.0,
      "step": 12360
    },
    {
      "entropy": 5.500197219848633,
      "epoch": 1.2271734815402937,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048574234857258356,
      "loss": 5.3423,
      "mean_token_accuracy": 0.15699008107185364,
      "num_tokens": 27088294.0,
      "step": 12365
    },
    {
      "entropy": 5.49582257270813,
      "epoch": 1.2276697102024614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048572993384284833,
      "loss": 5.2787,
      "mean_token_accuracy": 0.15313283652067183,
      "num_tokens": 27098848.0,
      "step": 12370
    },
    {
      "entropy": 5.379373931884766,
      "epoch": 1.2281659388646289,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004857175138874648,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1576991483569145,
      "num_tokens": 27112319.0,
      "step": 12375
    },
    {
      "entropy": 5.442618560791016,
      "epoch": 1.2286621675267964,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048570508870674097,
      "loss": 5.2885,
      "mean_token_accuracy": 0.16073673069477082,
      "num_tokens": 27121840.0,
      "step": 12380
    },
    {
      "entropy": 5.449471855163575,
      "epoch": 1.2291583961889638,
      "grad_norm": 1.0,
      "learning_rate": 0.00048569265830098493,
      "loss": 5.2994,
      "mean_token_accuracy": 0.1580878347158432,
      "num_tokens": 27133619.0,
      "step": 12385
    },
    {
      "entropy": 5.514698266983032,
      "epoch": 1.2296546248511313,
      "grad_norm": 1.125,
      "learning_rate": 0.000485680222670505,
      "loss": 5.2806,
      "mean_token_accuracy": 0.1572817638516426,
      "num_tokens": 27144161.0,
      "step": 12390
    },
    {
      "entropy": 5.443016242980957,
      "epoch": 1.230150853513299,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004856677818156096,
      "loss": 5.1954,
      "mean_token_accuracy": 0.1621087223291397,
      "num_tokens": 27154906.0,
      "step": 12395
    },
    {
      "entropy": 5.3978368759155275,
      "epoch": 1.2306470821754665,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048565533573660707,
      "loss": 5.2309,
      "mean_token_accuracy": 0.16087841242551804,
      "num_tokens": 27165746.0,
      "step": 12400
    },
    {
      "entropy": 5.390279006958008,
      "epoch": 1.231143310837634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004856428844338062,
      "loss": 5.2189,
      "mean_token_accuracy": 0.16318794041872026,
      "num_tokens": 27176108.0,
      "step": 12405
    },
    {
      "entropy": 5.440843152999878,
      "epoch": 1.2316395394998014,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004856304279075156,
      "loss": 5.3029,
      "mean_token_accuracy": 0.15514287799596788,
      "num_tokens": 27186721.0,
      "step": 12410
    },
    {
      "entropy": 5.47254638671875,
      "epoch": 1.2321357681619691,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004856179661580444,
      "loss": 5.1781,
      "mean_token_accuracy": 0.16377765834331512,
      "num_tokens": 27196120.0,
      "step": 12415
    },
    {
      "entropy": 5.470973491668701,
      "epoch": 1.2326319968241366,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004856054991857013,
      "loss": 5.2969,
      "mean_token_accuracy": 0.15238352119922638,
      "num_tokens": 27207431.0,
      "step": 12420
    },
    {
      "entropy": 5.408423566818238,
      "epoch": 1.233128225486304,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048559302699079576,
      "loss": 5.189,
      "mean_token_accuracy": 0.16476137489080428,
      "num_tokens": 27218489.0,
      "step": 12425
    },
    {
      "entropy": 5.482289457321167,
      "epoch": 1.2336244541484715,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004855805495736369,
      "loss": 5.3126,
      "mean_token_accuracy": 0.15745091587305068,
      "num_tokens": 27230849.0,
      "step": 12430
    },
    {
      "entropy": 5.47808518409729,
      "epoch": 1.234120682810639,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004855680669345342,
      "loss": 5.275,
      "mean_token_accuracy": 0.15710455626249314,
      "num_tokens": 27242838.0,
      "step": 12435
    },
    {
      "entropy": 5.458873510360718,
      "epoch": 1.2346169114728067,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048555557907379714,
      "loss": 5.2973,
      "mean_token_accuracy": 0.1645626664161682,
      "num_tokens": 27253945.0,
      "step": 12440
    },
    {
      "entropy": 5.473442220687867,
      "epoch": 1.2351131401349742,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048554308599173544,
      "loss": 5.3559,
      "mean_token_accuracy": 0.14874770417809485,
      "num_tokens": 27263999.0,
      "step": 12445
    },
    {
      "entropy": 5.4941641330719,
      "epoch": 1.2356093687971417,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004855305876886589,
      "loss": 5.2042,
      "mean_token_accuracy": 0.16152806282043458,
      "num_tokens": 27274677.0,
      "step": 12450
    },
    {
      "entropy": 5.4457518577575685,
      "epoch": 1.2361055974593094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048551808416487743,
      "loss": 5.2796,
      "mean_token_accuracy": 0.15692635476589203,
      "num_tokens": 27286512.0,
      "step": 12455
    },
    {
      "entropy": 5.401173067092896,
      "epoch": 1.2366018261214768,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004855055754207011,
      "loss": 5.2702,
      "mean_token_accuracy": 0.1593891829252243,
      "num_tokens": 27297620.0,
      "step": 12460
    },
    {
      "entropy": 5.477277612686157,
      "epoch": 1.2370980547836443,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004854930614564401,
      "loss": 5.286,
      "mean_token_accuracy": 0.15676653534173965,
      "num_tokens": 27309963.0,
      "step": 12465
    },
    {
      "entropy": 5.577436017990112,
      "epoch": 1.2375942834458118,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004854805422724047,
      "loss": 5.3836,
      "mean_token_accuracy": 0.15321325063705443,
      "num_tokens": 27320740.0,
      "step": 12470
    },
    {
      "entropy": 5.479734563827515,
      "epoch": 1.2380905121079793,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004854680178689055,
      "loss": 5.2101,
      "mean_token_accuracy": 0.16389924585819243,
      "num_tokens": 27331543.0,
      "step": 12475
    },
    {
      "entropy": 5.338040208816528,
      "epoch": 1.238586740770147,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048545548824625287,
      "loss": 5.1843,
      "mean_token_accuracy": 0.16034260243177414,
      "num_tokens": 27343222.0,
      "step": 12480
    },
    {
      "entropy": 5.419043350219726,
      "epoch": 1.2390829694323144,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004854429534047577,
      "loss": 5.2539,
      "mean_token_accuracy": 0.15834279954433442,
      "num_tokens": 27354694.0,
      "step": 12485
    },
    {
      "entropy": 5.446694993972779,
      "epoch": 1.239579198094482,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004854304133447307,
      "loss": 5.2319,
      "mean_token_accuracy": 0.15416265428066253,
      "num_tokens": 27365462.0,
      "step": 12490
    },
    {
      "entropy": 5.377581787109375,
      "epoch": 1.2400754267566494,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048541786806648293,
      "loss": 5.1437,
      "mean_token_accuracy": 0.165875668823719,
      "num_tokens": 27377548.0,
      "step": 12495
    },
    {
      "entropy": 5.46746039390564,
      "epoch": 1.240571655418817,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004854053175703254,
      "loss": 5.2827,
      "mean_token_accuracy": 0.15562227591872216,
      "num_tokens": 27388221.0,
      "step": 12500
    },
    {
      "entropy": 5.479973983764649,
      "epoch": 1.2410678840809846,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004853927618565694,
      "loss": 5.2663,
      "mean_token_accuracy": 0.16169948875904083,
      "num_tokens": 27400497.0,
      "step": 12505
    },
    {
      "entropy": 5.415345335006714,
      "epoch": 1.241564112743152,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048538020092552626,
      "loss": 5.1717,
      "mean_token_accuracy": 0.1623190850019455,
      "num_tokens": 27410700.0,
      "step": 12510
    },
    {
      "entropy": 5.467405557632446,
      "epoch": 1.2420603414053195,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048536763477750747,
      "loss": 5.3033,
      "mean_token_accuracy": 0.1585919052362442,
      "num_tokens": 27421286.0,
      "step": 12515
    },
    {
      "entropy": 5.530794715881347,
      "epoch": 1.242556570067487,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004853550634128246,
      "loss": 5.3127,
      "mean_token_accuracy": 0.1558150351047516,
      "num_tokens": 27433922.0,
      "step": 12520
    },
    {
      "entropy": 5.455029392242432,
      "epoch": 1.2430527987296547,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004853424868317895,
      "loss": 5.2483,
      "mean_token_accuracy": 0.15644557029008865,
      "num_tokens": 27444684.0,
      "step": 12525
    },
    {
      "entropy": 5.414221382141113,
      "epoch": 1.2435490273918222,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004853299050347139,
      "loss": 5.2273,
      "mean_token_accuracy": 0.1570575624704361,
      "num_tokens": 27455494.0,
      "step": 12530
    },
    {
      "entropy": 5.447025012969971,
      "epoch": 1.2440452560539896,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048531731802190994,
      "loss": 5.2644,
      "mean_token_accuracy": 0.1550910845398903,
      "num_tokens": 27464927.0,
      "step": 12535
    },
    {
      "entropy": 5.493202114105225,
      "epoch": 1.244541484716157,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004853047257936897,
      "loss": 5.2782,
      "mean_token_accuracy": 0.15360616892576218,
      "num_tokens": 27476228.0,
      "step": 12540
    },
    {
      "entropy": 5.498160648345947,
      "epoch": 1.2450377133783248,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004852921283503653,
      "loss": 5.2763,
      "mean_token_accuracy": 0.15904087871313094,
      "num_tokens": 27488835.0,
      "step": 12545
    },
    {
      "entropy": 5.423143196105957,
      "epoch": 1.2455339420404923,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004852795256922494,
      "loss": 5.2444,
      "mean_token_accuracy": 0.15751414149999618,
      "num_tokens": 27501481.0,
      "step": 12550
    },
    {
      "entropy": 5.433859539031983,
      "epoch": 1.2460301707026598,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048526691781965425,
      "loss": 5.2522,
      "mean_token_accuracy": 0.15632500350475312,
      "num_tokens": 27512234.0,
      "step": 12555
    },
    {
      "entropy": 5.524566173553467,
      "epoch": 1.2465263993648272,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048525430473289274,
      "loss": 5.3173,
      "mean_token_accuracy": 0.14935122281312943,
      "num_tokens": 27523756.0,
      "step": 12560
    },
    {
      "entropy": 5.510654544830322,
      "epoch": 1.247022628026995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048524168643227747,
      "loss": 5.3131,
      "mean_token_accuracy": 0.15719311088323593,
      "num_tokens": 27534579.0,
      "step": 12565
    },
    {
      "entropy": 5.433606863021851,
      "epoch": 1.2475188566891624,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048522906291812145,
      "loss": 5.3472,
      "mean_token_accuracy": 0.15292004644870758,
      "num_tokens": 27546420.0,
      "step": 12570
    },
    {
      "entropy": 5.452826833724975,
      "epoch": 1.2480150853513299,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048521643419073763,
      "loss": 5.2306,
      "mean_token_accuracy": 0.15502957105636597,
      "num_tokens": 27557701.0,
      "step": 12575
    },
    {
      "entropy": 5.430633211135865,
      "epoch": 1.2485113140134974,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004852038002504392,
      "loss": 5.2351,
      "mean_token_accuracy": 0.1686185970902443,
      "num_tokens": 27568592.0,
      "step": 12580
    },
    {
      "entropy": 5.474829244613647,
      "epoch": 1.249007542675665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048519116109753955,
      "loss": 5.2824,
      "mean_token_accuracy": 0.16372230350971223,
      "num_tokens": 27579286.0,
      "step": 12585
    },
    {
      "entropy": 5.431408071517945,
      "epoch": 1.2495037713378325,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048517851673235196,
      "loss": 5.2247,
      "mean_token_accuracy": 0.15881977826356888,
      "num_tokens": 27590031.0,
      "step": 12590
    },
    {
      "entropy": 5.454576015472412,
      "epoch": 1.25,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004851658671551901,
      "loss": 5.2668,
      "mean_token_accuracy": 0.15683535784482955,
      "num_tokens": 27600999.0,
      "step": 12595
    },
    {
      "entropy": 5.428625106811523,
      "epoch": 1.2504962286621675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048515321236636754,
      "loss": 5.232,
      "mean_token_accuracy": 0.15394672006368637,
      "num_tokens": 27613394.0,
      "step": 12600
    },
    {
      "entropy": 5.427497720718383,
      "epoch": 1.250992457324335,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004851405523661982,
      "loss": 5.2156,
      "mean_token_accuracy": 0.16744298934936525,
      "num_tokens": 27624819.0,
      "step": 12605
    },
    {
      "entropy": 5.411412620544434,
      "epoch": 1.2514886859865026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048512788715499597,
      "loss": 5.2426,
      "mean_token_accuracy": 0.1642121136188507,
      "num_tokens": 27635993.0,
      "step": 12610
    },
    {
      "entropy": 5.453460597991944,
      "epoch": 1.2519849146486701,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004851152167330749,
      "loss": 5.3191,
      "mean_token_accuracy": 0.15603729486465454,
      "num_tokens": 27647298.0,
      "step": 12615
    },
    {
      "entropy": 5.52367582321167,
      "epoch": 1.2524811433108376,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048510254110074923,
      "loss": 5.2943,
      "mean_token_accuracy": 0.1561756029725075,
      "num_tokens": 27659381.0,
      "step": 12620
    },
    {
      "entropy": 5.458435106277466,
      "epoch": 1.2529773719730053,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004850898602583333,
      "loss": 5.2899,
      "mean_token_accuracy": 0.15891631841659545,
      "num_tokens": 27671122.0,
      "step": 12625
    },
    {
      "entropy": 5.4852752685546875,
      "epoch": 1.2534736006351728,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048507717420614153,
      "loss": 5.2645,
      "mean_token_accuracy": 0.16209448724985123,
      "num_tokens": 27683258.0,
      "step": 12630
    },
    {
      "entropy": 5.345997619628906,
      "epoch": 1.2539698292973402,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850644829444884,
      "loss": 5.1341,
      "mean_token_accuracy": 0.16174716949462892,
      "num_tokens": 27693116.0,
      "step": 12635
    },
    {
      "entropy": 5.363073873519897,
      "epoch": 1.2544660579595077,
      "grad_norm": 1.0,
      "learning_rate": 0.00048505178647368887,
      "loss": 5.2326,
      "mean_token_accuracy": 0.16226858794689178,
      "num_tokens": 27704859.0,
      "step": 12640
    },
    {
      "entropy": 5.401530504226685,
      "epoch": 1.2549622866216752,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850390847940576,
      "loss": 5.1732,
      "mean_token_accuracy": 0.1622050777077675,
      "num_tokens": 27715854.0,
      "step": 12645
    },
    {
      "entropy": 5.424926805496216,
      "epoch": 1.2554585152838427,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004850263779059097,
      "loss": 5.2275,
      "mean_token_accuracy": 0.1570133313536644,
      "num_tokens": 27726398.0,
      "step": 12650
    },
    {
      "entropy": 5.452268362045288,
      "epoch": 1.2559547439460104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004850136658095601,
      "loss": 5.2428,
      "mean_token_accuracy": 0.16335160434246063,
      "num_tokens": 27737337.0,
      "step": 12655
    },
    {
      "entropy": 5.438182544708252,
      "epoch": 1.2564509726081778,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048500094850532417,
      "loss": 5.2664,
      "mean_token_accuracy": 0.16045714169740677,
      "num_tokens": 27748593.0,
      "step": 12660
    },
    {
      "entropy": 5.485013818740844,
      "epoch": 1.2569472012703453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004849882259935172,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1532875694334507,
      "num_tokens": 27760908.0,
      "step": 12665
    },
    {
      "entropy": 5.463221406936645,
      "epoch": 1.257443429932513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004849754982744548,
      "loss": 5.3331,
      "mean_token_accuracy": 0.1553870901465416,
      "num_tokens": 27772104.0,
      "step": 12670
    },
    {
      "entropy": 5.413131809234619,
      "epoch": 1.2579396585946805,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004849627653484524,
      "loss": 5.2439,
      "mean_token_accuracy": 0.15981703847646714,
      "num_tokens": 27781654.0,
      "step": 12675
    },
    {
      "entropy": 5.4714501857757565,
      "epoch": 1.258435887256848,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004849500272158259,
      "loss": 5.2522,
      "mean_token_accuracy": 0.16087721139192582,
      "num_tokens": 27793026.0,
      "step": 12680
    },
    {
      "entropy": 5.468347978591919,
      "epoch": 1.2589321159190154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004849372838768911,
      "loss": 5.2191,
      "mean_token_accuracy": 0.1600153088569641,
      "num_tokens": 27803447.0,
      "step": 12685
    },
    {
      "entropy": 5.485860109329224,
      "epoch": 1.259428344581183,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000484924535331964,
      "loss": 5.229,
      "mean_token_accuracy": 0.15862605720758438,
      "num_tokens": 27814632.0,
      "step": 12690
    },
    {
      "entropy": 5.473284387588501,
      "epoch": 1.2599245732433506,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048491178158136086,
      "loss": 5.3207,
      "mean_token_accuracy": 0.15263260900974274,
      "num_tokens": 27826582.0,
      "step": 12695
    },
    {
      "entropy": 5.442102718353271,
      "epoch": 1.260420801905518,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004848990226253978,
      "loss": 5.1727,
      "mean_token_accuracy": 0.15927495211362838,
      "num_tokens": 27838065.0,
      "step": 12700
    },
    {
      "entropy": 5.373810768127441,
      "epoch": 1.2609170305676856,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048488625846439135,
      "loss": 5.2259,
      "mean_token_accuracy": 0.15756784081459047,
      "num_tokens": 27848182.0,
      "step": 12705
    },
    {
      "entropy": 5.497313022613525,
      "epoch": 1.261413259229853,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004848734890986578,
      "loss": 5.2506,
      "mean_token_accuracy": 0.15465472787618637,
      "num_tokens": 27858524.0,
      "step": 12710
    },
    {
      "entropy": 5.425972604751587,
      "epoch": 1.2619094878920207,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004848607145285141,
      "loss": 5.1337,
      "mean_token_accuracy": 0.16513666808605193,
      "num_tokens": 27868588.0,
      "step": 12715
    },
    {
      "entropy": 5.351064777374267,
      "epoch": 1.2624057165541882,
      "grad_norm": 1.125,
      "learning_rate": 0.0004848479347542768,
      "loss": 5.1665,
      "mean_token_accuracy": 0.1588658720254898,
      "num_tokens": 27878829.0,
      "step": 12720
    },
    {
      "entropy": 5.4185553073883055,
      "epoch": 1.2629019452163557,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000484835149776263,
      "loss": 5.2421,
      "mean_token_accuracy": 0.16729347854852678,
      "num_tokens": 27889872.0,
      "step": 12725
    },
    {
      "entropy": 5.475226163864136,
      "epoch": 1.2633981738785232,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048482235959478953,
      "loss": 5.3095,
      "mean_token_accuracy": 0.1644190326333046,
      "num_tokens": 27900567.0,
      "step": 12730
    },
    {
      "entropy": 5.479246377944946,
      "epoch": 1.2638944025406906,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004848095642101737,
      "loss": 5.1883,
      "mean_token_accuracy": 0.15778561532497407,
      "num_tokens": 27911594.0,
      "step": 12735
    },
    {
      "entropy": 5.490329551696777,
      "epoch": 1.2643906312028583,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048479676362273277,
      "loss": 5.2982,
      "mean_token_accuracy": 0.15147269666194915,
      "num_tokens": 27923694.0,
      "step": 12740
    },
    {
      "entropy": 5.458056497573852,
      "epoch": 1.2648868598650258,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048478395783278425,
      "loss": 5.3645,
      "mean_token_accuracy": 0.14628463387489318,
      "num_tokens": 27934168.0,
      "step": 12745
    },
    {
      "entropy": 5.501045370101929,
      "epoch": 1.2653830885271933,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004847711468406455,
      "loss": 5.2538,
      "mean_token_accuracy": 0.15901567190885543,
      "num_tokens": 27944401.0,
      "step": 12750
    },
    {
      "entropy": 5.394528722763061,
      "epoch": 1.265879317189361,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048475833064663443,
      "loss": 5.1484,
      "mean_token_accuracy": 0.16940377056598663,
      "num_tokens": 27956219.0,
      "step": 12755
    },
    {
      "entropy": 5.442562913894653,
      "epoch": 1.2663755458515285,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004847455092510687,
      "loss": 5.2646,
      "mean_token_accuracy": 0.15838275849819183,
      "num_tokens": 27967457.0,
      "step": 12760
    },
    {
      "entropy": 5.391359663009643,
      "epoch": 1.266871774513696,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004847326826542663,
      "loss": 5.2741,
      "mean_token_accuracy": 0.15450644344091416,
      "num_tokens": 27978287.0,
      "step": 12765
    },
    {
      "entropy": 5.478402137756348,
      "epoch": 1.2673680031758634,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048471985085654537,
      "loss": 5.2841,
      "mean_token_accuracy": 0.16297248303890227,
      "num_tokens": 27989703.0,
      "step": 12770
    },
    {
      "entropy": 5.439664506912232,
      "epoch": 1.2678642318380309,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048470701385822394,
      "loss": 5.2401,
      "mean_token_accuracy": 0.15903167128562928,
      "num_tokens": 28002071.0,
      "step": 12775
    },
    {
      "entropy": 5.476836776733398,
      "epoch": 1.2683604605001986,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048469417165962047,
      "loss": 5.3045,
      "mean_token_accuracy": 0.15857402980327606,
      "num_tokens": 28014517.0,
      "step": 12780
    },
    {
      "entropy": 5.476937770843506,
      "epoch": 1.268856689162366,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004846813242610535,
      "loss": 5.2646,
      "mean_token_accuracy": 0.16445594578981398,
      "num_tokens": 28025135.0,
      "step": 12785
    },
    {
      "entropy": 5.424613046646118,
      "epoch": 1.2693529178245335,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004846684716628414,
      "loss": 5.2031,
      "mean_token_accuracy": 0.16359377205371856,
      "num_tokens": 28036105.0,
      "step": 12790
    },
    {
      "entropy": 5.427709150314331,
      "epoch": 1.269849146486701,
      "grad_norm": 1.125,
      "learning_rate": 0.000484655613865303,
      "loss": 5.2402,
      "mean_token_accuracy": 0.16433196216821672,
      "num_tokens": 28048825.0,
      "step": 12795
    },
    {
      "entropy": 5.482144069671631,
      "epoch": 1.2703453751488687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048464275086875717,
      "loss": 5.1955,
      "mean_token_accuracy": 0.15961596816778184,
      "num_tokens": 28060408.0,
      "step": 12800
    },
    {
      "entropy": 5.440124464035034,
      "epoch": 1.2708416038110362,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004846298826735229,
      "loss": 5.2729,
      "mean_token_accuracy": 0.15589027851819992,
      "num_tokens": 28072179.0,
      "step": 12805
    },
    {
      "entropy": 5.417941570281982,
      "epoch": 1.2713378324732036,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004846170092799192,
      "loss": 5.2602,
      "mean_token_accuracy": 0.15393998324871064,
      "num_tokens": 28083825.0,
      "step": 12810
    },
    {
      "entropy": 5.631198501586914,
      "epoch": 1.2718340611353711,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048460413068826536,
      "loss": 5.3975,
      "mean_token_accuracy": 0.15384954810142518,
      "num_tokens": 28095136.0,
      "step": 12815
    },
    {
      "entropy": 5.4774641513824465,
      "epoch": 1.2723302897975386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004845912468988807,
      "loss": 5.2111,
      "mean_token_accuracy": 0.159006205201149,
      "num_tokens": 28104763.0,
      "step": 12820
    },
    {
      "entropy": 5.391424465179443,
      "epoch": 1.2728265184597063,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004845783579120849,
      "loss": 5.2655,
      "mean_token_accuracy": 0.15990522652864456,
      "num_tokens": 28115771.0,
      "step": 12825
    },
    {
      "entropy": 5.363130331039429,
      "epoch": 1.2733227471218738,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004845654637281973,
      "loss": 5.1943,
      "mean_token_accuracy": 0.16067807078361512,
      "num_tokens": 28128039.0,
      "step": 12830
    },
    {
      "entropy": 5.406911563873291,
      "epoch": 1.2738189757840412,
      "grad_norm": 1.125,
      "learning_rate": 0.00048455256434753774,
      "loss": 5.211,
      "mean_token_accuracy": 0.16531694084405898,
      "num_tokens": 28139163.0,
      "step": 12835
    },
    {
      "entropy": 5.425388240814209,
      "epoch": 1.274315204446209,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004845396597704262,
      "loss": 5.255,
      "mean_token_accuracy": 0.15361825972795487,
      "num_tokens": 28150625.0,
      "step": 12840
    },
    {
      "entropy": 5.49505729675293,
      "epoch": 1.2748114331083764,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048452674999718257,
      "loss": 5.3386,
      "mean_token_accuracy": 0.1585584133863449,
      "num_tokens": 28163072.0,
      "step": 12845
    },
    {
      "entropy": 5.5166792392730715,
      "epoch": 1.2753076617705439,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048451383502812705,
      "loss": 5.354,
      "mean_token_accuracy": 0.15685149729251863,
      "num_tokens": 28174192.0,
      "step": 12850
    },
    {
      "entropy": 5.39316840171814,
      "epoch": 1.2758038904327114,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048450091486357987,
      "loss": 5.2044,
      "mean_token_accuracy": 0.16577997505664827,
      "num_tokens": 28184477.0,
      "step": 12855
    },
    {
      "entropy": 5.347987508773803,
      "epoch": 1.2763001190948788,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048448798950386147,
      "loss": 5.1248,
      "mean_token_accuracy": 0.16144870817661286,
      "num_tokens": 28194552.0,
      "step": 12860
    },
    {
      "entropy": 5.431000471115112,
      "epoch": 1.2767963477570463,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048447505894929233,
      "loss": 5.2997,
      "mean_token_accuracy": 0.15895115584135056,
      "num_tokens": 28205712.0,
      "step": 12865
    },
    {
      "entropy": 5.476150751113892,
      "epoch": 1.277292576419214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048446212320019306,
      "loss": 5.2231,
      "mean_token_accuracy": 0.15961794853210448,
      "num_tokens": 28217246.0,
      "step": 12870
    },
    {
      "entropy": 5.450198030471801,
      "epoch": 1.2777888050813815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004844491822568846,
      "loss": 5.2389,
      "mean_token_accuracy": 0.15589951127767562,
      "num_tokens": 28227933.0,
      "step": 12875
    },
    {
      "entropy": 5.459042358398437,
      "epoch": 1.278285033743549,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048443623611968764,
      "loss": 5.2695,
      "mean_token_accuracy": 0.16296703591942788,
      "num_tokens": 28239692.0,
      "step": 12880
    },
    {
      "entropy": 5.414179563522339,
      "epoch": 1.2787812624057167,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048442328478892334,
      "loss": 5.2367,
      "mean_token_accuracy": 0.16240196526050568,
      "num_tokens": 28249815.0,
      "step": 12885
    },
    {
      "entropy": 5.475827407836914,
      "epoch": 1.2792774910678841,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004844103282649128,
      "loss": 5.2561,
      "mean_token_accuracy": 0.1568821534514427,
      "num_tokens": 28260385.0,
      "step": 12890
    },
    {
      "entropy": 5.430259895324707,
      "epoch": 1.2797737197300516,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004843973665479774,
      "loss": 5.2741,
      "mean_token_accuracy": 0.14875324964523315,
      "num_tokens": 28271858.0,
      "step": 12895
    },
    {
      "entropy": 5.45551643371582,
      "epoch": 1.280269948392219,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048438439963843853,
      "loss": 5.2657,
      "mean_token_accuracy": 0.1545419618487358,
      "num_tokens": 28283988.0,
      "step": 12900
    },
    {
      "entropy": 5.447976779937744,
      "epoch": 1.2807661770543866,
      "grad_norm": 1.25,
      "learning_rate": 0.00048437142753661775,
      "loss": 5.2955,
      "mean_token_accuracy": 0.15742011964321137,
      "num_tokens": 28293281.0,
      "step": 12905
    },
    {
      "entropy": 5.513677358627319,
      "epoch": 1.2812624057165543,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004843584502428367,
      "loss": 5.2451,
      "mean_token_accuracy": 0.15367969125509262,
      "num_tokens": 28304468.0,
      "step": 12910
    },
    {
      "entropy": 5.363796043395996,
      "epoch": 1.2817586343787217,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004843454677574173,
      "loss": 5.2067,
      "mean_token_accuracy": 0.16433411389589309,
      "num_tokens": 28315864.0,
      "step": 12915
    },
    {
      "entropy": 5.415114498138427,
      "epoch": 1.2822548630408892,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048433248008068123,
      "loss": 5.2602,
      "mean_token_accuracy": 0.1626923307776451,
      "num_tokens": 28327212.0,
      "step": 12920
    },
    {
      "entropy": 5.473800706863403,
      "epoch": 1.282751091703057,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048431948721295085,
      "loss": 5.2714,
      "mean_token_accuracy": 0.16290073543787004,
      "num_tokens": 28339256.0,
      "step": 12925
    },
    {
      "entropy": 5.447014856338501,
      "epoch": 1.2832473203652244,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004843064891545482,
      "loss": 5.2978,
      "mean_token_accuracy": 0.15564928948879242,
      "num_tokens": 28350196.0,
      "step": 12930
    },
    {
      "entropy": 5.439802265167236,
      "epoch": 1.2837435490273919,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004842934859057956,
      "loss": 5.2397,
      "mean_token_accuracy": 0.15595787987113,
      "num_tokens": 28361241.0,
      "step": 12935
    },
    {
      "entropy": 5.5439653396606445,
      "epoch": 1.2842397776895593,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004842804774670156,
      "loss": 5.3798,
      "mean_token_accuracy": 0.15411950945854186,
      "num_tokens": 28373369.0,
      "step": 12940
    },
    {
      "entropy": 5.481730556488037,
      "epoch": 1.2847360063517268,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048426746383853066,
      "loss": 5.19,
      "mean_token_accuracy": 0.16079164296388626,
      "num_tokens": 28384970.0,
      "step": 12945
    },
    {
      "entropy": 5.425660753250122,
      "epoch": 1.2852322350138943,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048425444502066354,
      "loss": 5.2157,
      "mean_token_accuracy": 0.16058826595544815,
      "num_tokens": 28396469.0,
      "step": 12950
    },
    {
      "entropy": 5.393751335144043,
      "epoch": 1.285728463676062,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048424142101373706,
      "loss": 5.2744,
      "mean_token_accuracy": 0.16253739297389985,
      "num_tokens": 28406975.0,
      "step": 12955
    },
    {
      "entropy": 5.4042277336120605,
      "epoch": 1.2862246923382294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048422839181807426,
      "loss": 5.1502,
      "mean_token_accuracy": 0.1732693150639534,
      "num_tokens": 28418285.0,
      "step": 12960
    },
    {
      "entropy": 5.397644281387329,
      "epoch": 1.286720921000397,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004842153574339982,
      "loss": 5.3001,
      "mean_token_accuracy": 0.15233973711729049,
      "num_tokens": 28429769.0,
      "step": 12965
    },
    {
      "entropy": 5.372210311889648,
      "epoch": 1.2872171496625646,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048420231786183216,
      "loss": 5.2496,
      "mean_token_accuracy": 0.16142528802156447,
      "num_tokens": 28441188.0,
      "step": 12970
    },
    {
      "entropy": 5.564180707931518,
      "epoch": 1.287713378324732,
      "grad_norm": 1.125,
      "learning_rate": 0.0004841892731018994,
      "loss": 5.3126,
      "mean_token_accuracy": 0.15863290429115295,
      "num_tokens": 28453504.0,
      "step": 12975
    },
    {
      "entropy": 5.472735595703125,
      "epoch": 1.2882096069868996,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004841762231545233,
      "loss": 5.2928,
      "mean_token_accuracy": 0.16464899927377702,
      "num_tokens": 28465930.0,
      "step": 12980
    },
    {
      "entropy": 5.423328161239624,
      "epoch": 1.288705835649067,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048416316802002776,
      "loss": 5.2271,
      "mean_token_accuracy": 0.1633104532957077,
      "num_tokens": 28477803.0,
      "step": 12985
    },
    {
      "entropy": 5.496878576278687,
      "epoch": 1.2892020643112345,
      "grad_norm": 1.125,
      "learning_rate": 0.00048415010769873625,
      "loss": 5.3079,
      "mean_token_accuracy": 0.15376178473234176,
      "num_tokens": 28488556.0,
      "step": 12990
    },
    {
      "entropy": 5.463488435745239,
      "epoch": 1.2896982929734022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004841370421909728,
      "loss": 5.2008,
      "mean_token_accuracy": 0.15984019190073012,
      "num_tokens": 28499275.0,
      "step": 12995
    },
    {
      "entropy": 5.434684228897095,
      "epoch": 1.2901945216355697,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048412397149706133,
      "loss": 5.2308,
      "mean_token_accuracy": 0.16234491169452667,
      "num_tokens": 28511327.0,
      "step": 13000
    },
    {
      "entropy": 5.428905010223389,
      "epoch": 1.2906907502977372,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000484110895617326,
      "loss": 5.2737,
      "mean_token_accuracy": 0.1637100651860237,
      "num_tokens": 28523937.0,
      "step": 13005
    },
    {
      "entropy": 5.355237007141113,
      "epoch": 1.2911869789599046,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048409781455209104,
      "loss": 5.1894,
      "mean_token_accuracy": 0.15871017873287202,
      "num_tokens": 28534028.0,
      "step": 13010
    },
    {
      "entropy": 5.475120115280151,
      "epoch": 1.2916832076220723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004840847283016809,
      "loss": 5.2675,
      "mean_token_accuracy": 0.14666867703199388,
      "num_tokens": 28545480.0,
      "step": 13015
    },
    {
      "entropy": 5.446512889862061,
      "epoch": 1.2921794362842398,
      "grad_norm": 1.125,
      "learning_rate": 0.00048407163686642,
      "loss": 5.2266,
      "mean_token_accuracy": 0.1667512908577919,
      "num_tokens": 28555401.0,
      "step": 13020
    },
    {
      "entropy": 5.454007244110107,
      "epoch": 1.2926756649464073,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000484058540246633,
      "loss": 5.2795,
      "mean_token_accuracy": 0.16046481654047967,
      "num_tokens": 28565497.0,
      "step": 13025
    },
    {
      "entropy": 5.437551355361938,
      "epoch": 1.2931718936085748,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004840454384426447,
      "loss": 5.3755,
      "mean_token_accuracy": 0.15050119608640672,
      "num_tokens": 28576186.0,
      "step": 13030
    },
    {
      "entropy": 5.517925643920899,
      "epoch": 1.2936681222707422,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004840323314547799,
      "loss": 5.2336,
      "mean_token_accuracy": 0.1635855406522751,
      "num_tokens": 28586202.0,
      "step": 13035
    },
    {
      "entropy": 5.437497520446778,
      "epoch": 1.29416435093291,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048401921928336385,
      "loss": 5.2194,
      "mean_token_accuracy": 0.1597439870238304,
      "num_tokens": 28596486.0,
      "step": 13040
    },
    {
      "entropy": 5.479133129119873,
      "epoch": 1.2946605795950774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004840061019287215,
      "loss": 5.2462,
      "mean_token_accuracy": 0.16039675027132033,
      "num_tokens": 28607934.0,
      "step": 13045
    },
    {
      "entropy": 5.511542654037475,
      "epoch": 1.2951568082572449,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048399297939117817,
      "loss": 5.2805,
      "mean_token_accuracy": 0.15433167964220046,
      "num_tokens": 28617673.0,
      "step": 13050
    },
    {
      "entropy": 5.5139344215393065,
      "epoch": 1.2956530369194126,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004839798516710593,
      "loss": 5.2733,
      "mean_token_accuracy": 0.15749592632055281,
      "num_tokens": 28629811.0,
      "step": 13055
    },
    {
      "entropy": 5.475889873504639,
      "epoch": 1.29614926558158,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004839667187686904,
      "loss": 5.3559,
      "mean_token_accuracy": 0.15619957000017165,
      "num_tokens": 28640371.0,
      "step": 13060
    },
    {
      "entropy": 5.474207782745362,
      "epoch": 1.2966454942437475,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004839535806843972,
      "loss": 5.275,
      "mean_token_accuracy": 0.16233165711164474,
      "num_tokens": 28651904.0,
      "step": 13065
    },
    {
      "entropy": 5.412385177612305,
      "epoch": 1.297141722905915,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004839404374185054,
      "loss": 5.185,
      "mean_token_accuracy": 0.16208725273609162,
      "num_tokens": 28662241.0,
      "step": 13070
    },
    {
      "entropy": 5.39664216041565,
      "epoch": 1.2976379515680825,
      "grad_norm": 1.078125,
      "learning_rate": 0.000483927288971341,
      "loss": 5.1843,
      "mean_token_accuracy": 0.16416138410568237,
      "num_tokens": 28672791.0,
      "step": 13075
    },
    {
      "entropy": 5.48431167602539,
      "epoch": 1.2981341802302502,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004839141353432301,
      "loss": 5.306,
      "mean_token_accuracy": 0.15669557601213455,
      "num_tokens": 28684218.0,
      "step": 13080
    },
    {
      "entropy": 5.45332612991333,
      "epoch": 1.2986304088924177,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004839009765344987,
      "loss": 5.2456,
      "mean_token_accuracy": 0.15901148617267608,
      "num_tokens": 28694690.0,
      "step": 13085
    },
    {
      "entropy": 5.4030426979064945,
      "epoch": 1.2991266375545851,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048388781254547324,
      "loss": 5.2422,
      "mean_token_accuracy": 0.16391725838184357,
      "num_tokens": 28705890.0,
      "step": 13090
    },
    {
      "entropy": 5.457072830200195,
      "epoch": 1.2996228662167526,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048387464337648026,
      "loss": 5.264,
      "mean_token_accuracy": 0.15748645216226578,
      "num_tokens": 28716935.0,
      "step": 13095
    },
    {
      "entropy": 5.485350322723389,
      "epoch": 1.3001190948789203,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048386146902784606,
      "loss": 5.2788,
      "mean_token_accuracy": 0.15885828882455827,
      "num_tokens": 28728053.0,
      "step": 13100
    },
    {
      "entropy": 5.4243358135223385,
      "epoch": 1.3006153235410878,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048384828949989757,
      "loss": 5.2184,
      "mean_token_accuracy": 0.15988664329051971,
      "num_tokens": 28738940.0,
      "step": 13105
    },
    {
      "entropy": 5.445251941680908,
      "epoch": 1.3011115522032553,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048383510479296153,
      "loss": 5.2712,
      "mean_token_accuracy": 0.1605836808681488,
      "num_tokens": 28750259.0,
      "step": 13110
    },
    {
      "entropy": 5.480368328094483,
      "epoch": 1.3016077808654227,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004838219149073649,
      "loss": 5.2293,
      "mean_token_accuracy": 0.16371279954910278,
      "num_tokens": 28760119.0,
      "step": 13115
    },
    {
      "entropy": 5.495299625396728,
      "epoch": 1.3021040095275902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838087198434347,
      "loss": 5.2718,
      "mean_token_accuracy": 0.15402703285217284,
      "num_tokens": 28772022.0,
      "step": 13120
    },
    {
      "entropy": 5.483900785446167,
      "epoch": 1.302600238189758,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004837955196014982,
      "loss": 5.2384,
      "mean_token_accuracy": 0.15144469439983368,
      "num_tokens": 28782935.0,
      "step": 13125
    },
    {
      "entropy": 5.493810749053955,
      "epoch": 1.3030964668519254,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048378231418188275,
      "loss": 5.3357,
      "mean_token_accuracy": 0.1554546281695366,
      "num_tokens": 28793604.0,
      "step": 13130
    },
    {
      "entropy": 5.474476480484009,
      "epoch": 1.3035926955140928,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048376910358491576,
      "loss": 5.2941,
      "mean_token_accuracy": 0.15278302729129792,
      "num_tokens": 28803494.0,
      "step": 13135
    },
    {
      "entropy": 5.439769506454468,
      "epoch": 1.3040889241762605,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004837558878109249,
      "loss": 5.2651,
      "mean_token_accuracy": 0.15758197754621506,
      "num_tokens": 28816232.0,
      "step": 13140
    },
    {
      "entropy": 5.446805477142334,
      "epoch": 1.304585152838428,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004837426668602378,
      "loss": 5.2186,
      "mean_token_accuracy": 0.1628431960940361,
      "num_tokens": 28827056.0,
      "step": 13145
    },
    {
      "entropy": 5.402556037902832,
      "epoch": 1.3050813815005955,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004837294407331825,
      "loss": 5.2154,
      "mean_token_accuracy": 0.15868862271308898,
      "num_tokens": 28837217.0,
      "step": 13150
    },
    {
      "entropy": 5.341824913024903,
      "epoch": 1.305577610162763,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004837162094300867,
      "loss": 5.1095,
      "mean_token_accuracy": 0.16336964070796967,
      "num_tokens": 28846838.0,
      "step": 13155
    },
    {
      "entropy": 5.415543889999389,
      "epoch": 1.3060738388249304,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004837029729512787,
      "loss": 5.1688,
      "mean_token_accuracy": 0.16921347081661225,
      "num_tokens": 28856269.0,
      "step": 13160
    },
    {
      "entropy": 5.44448413848877,
      "epoch": 1.306570067487098,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048368973129708667,
      "loss": 5.2123,
      "mean_token_accuracy": 0.16404034346342086,
      "num_tokens": 28867142.0,
      "step": 13165
    },
    {
      "entropy": 5.4194519996643065,
      "epoch": 1.3070662961492656,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048367648446783894,
      "loss": 5.1812,
      "mean_token_accuracy": 0.16141366064548493,
      "num_tokens": 28877346.0,
      "step": 13170
    },
    {
      "entropy": 5.391416358947754,
      "epoch": 1.307562524811433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004836632324638642,
      "loss": 5.1768,
      "mean_token_accuracy": 0.16250848323106765,
      "num_tokens": 28888425.0,
      "step": 13175
    },
    {
      "entropy": 5.39745020866394,
      "epoch": 1.3080587534736006,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048364997528549077,
      "loss": 5.2144,
      "mean_token_accuracy": 0.16203493177890776,
      "num_tokens": 28899575.0,
      "step": 13180
    },
    {
      "entropy": 5.404217338562011,
      "epoch": 1.3085549821357683,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048363671293304763,
      "loss": 5.2087,
      "mean_token_accuracy": 0.1592944845557213,
      "num_tokens": 28910769.0,
      "step": 13185
    },
    {
      "entropy": 5.434482288360596,
      "epoch": 1.3090512107979357,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004836234454068636,
      "loss": 5.2196,
      "mean_token_accuracy": 0.16287839263677598,
      "num_tokens": 28921224.0,
      "step": 13190
    },
    {
      "entropy": 5.397690677642823,
      "epoch": 1.3095474394601032,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048361017270726764,
      "loss": 5.2179,
      "mean_token_accuracy": 0.15796797573566437,
      "num_tokens": 28932585.0,
      "step": 13195
    },
    {
      "entropy": 5.436833620071411,
      "epoch": 1.3100436681222707,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004835968948345889,
      "loss": 5.2587,
      "mean_token_accuracy": 0.15844152718782425,
      "num_tokens": 28943471.0,
      "step": 13200
    },
    {
      "entropy": 5.418719005584717,
      "epoch": 1.3105398967844382,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048358361178915666,
      "loss": 5.1759,
      "mean_token_accuracy": 0.17389153838157653,
      "num_tokens": 28956336.0,
      "step": 13205
    },
    {
      "entropy": 5.5087456703186035,
      "epoch": 1.3110361254466059,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004835703235713003,
      "loss": 5.2817,
      "mean_token_accuracy": 0.15867420583963393,
      "num_tokens": 28966689.0,
      "step": 13210
    },
    {
      "entropy": 5.463102722167969,
      "epoch": 1.3115323541087733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048355703018134935,
      "loss": 5.4108,
      "mean_token_accuracy": 0.16246972233057022,
      "num_tokens": 28978548.0,
      "step": 13215
    },
    {
      "entropy": 5.526914739608765,
      "epoch": 1.3120285827709408,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004835437316196335,
      "loss": 5.3097,
      "mean_token_accuracy": 0.16153138130903244,
      "num_tokens": 28988899.0,
      "step": 13220
    },
    {
      "entropy": 5.42088885307312,
      "epoch": 1.3125248114331085,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004835304278864824,
      "loss": 5.2294,
      "mean_token_accuracy": 0.16416267901659012,
      "num_tokens": 28999666.0,
      "step": 13225
    },
    {
      "entropy": 5.462984180450439,
      "epoch": 1.313021040095276,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004835171189822261,
      "loss": 5.2352,
      "mean_token_accuracy": 0.1573748394846916,
      "num_tokens": 29010185.0,
      "step": 13230
    },
    {
      "entropy": 5.443790102005005,
      "epoch": 1.3135172687574435,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004835038049071945,
      "loss": 5.2518,
      "mean_token_accuracy": 0.16044870018959045,
      "num_tokens": 29020761.0,
      "step": 13235
    },
    {
      "entropy": 5.517003726959229,
      "epoch": 1.314013497419611,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004834904856617178,
      "loss": 5.3953,
      "mean_token_accuracy": 0.15485914349555968,
      "num_tokens": 29031844.0,
      "step": 13240
    },
    {
      "entropy": 5.552556276321411,
      "epoch": 1.3145097260817784,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048347716124612643,
      "loss": 5.3451,
      "mean_token_accuracy": 0.1536606267094612,
      "num_tokens": 29043101.0,
      "step": 13245
    },
    {
      "entropy": 5.517656326293945,
      "epoch": 1.3150059547439459,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004834638316607506,
      "loss": 5.2726,
      "mean_token_accuracy": 0.16224616169929504,
      "num_tokens": 29054016.0,
      "step": 13250
    },
    {
      "entropy": 5.386099910736084,
      "epoch": 1.3155021834061136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048345049690592105,
      "loss": 5.2617,
      "mean_token_accuracy": 0.1594995751976967,
      "num_tokens": 29065127.0,
      "step": 13255
    },
    {
      "entropy": 5.480104112625122,
      "epoch": 1.315998412068281,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048343715698196814,
      "loss": 5.2178,
      "mean_token_accuracy": 0.15768117904663087,
      "num_tokens": 29075130.0,
      "step": 13260
    },
    {
      "entropy": 5.525743722915649,
      "epoch": 1.3164946407304485,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004834238118892231,
      "loss": 5.3428,
      "mean_token_accuracy": 0.15055608823895456,
      "num_tokens": 29086405.0,
      "step": 13265
    },
    {
      "entropy": 5.508456659317017,
      "epoch": 1.3169908693926162,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004834104616280165,
      "loss": 5.305,
      "mean_token_accuracy": 0.15388960167765617,
      "num_tokens": 29096945.0,
      "step": 13270
    },
    {
      "entropy": 5.513326787948609,
      "epoch": 1.3174870980547837,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004833971061986796,
      "loss": 5.3594,
      "mean_token_accuracy": 0.16100183725357056,
      "num_tokens": 29108275.0,
      "step": 13275
    },
    {
      "entropy": 5.473592615127563,
      "epoch": 1.3179833267169512,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048338374560154344,
      "loss": 5.2018,
      "mean_token_accuracy": 0.1585369810461998,
      "num_tokens": 29118832.0,
      "step": 13280
    },
    {
      "entropy": 5.422845888137817,
      "epoch": 1.3184795553791187,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004833703798369395,
      "loss": 5.2259,
      "mean_token_accuracy": 0.15805260837078094,
      "num_tokens": 29130720.0,
      "step": 13285
    },
    {
      "entropy": 5.472645711898804,
      "epoch": 1.3189757840412861,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004833570089051992,
      "loss": 5.2109,
      "mean_token_accuracy": 0.1614932969212532,
      "num_tokens": 29140441.0,
      "step": 13290
    },
    {
      "entropy": 5.4839965343475345,
      "epoch": 1.3194720127034538,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048334363280665393,
      "loss": 5.2765,
      "mean_token_accuracy": 0.15560702234506607,
      "num_tokens": 29151544.0,
      "step": 13295
    },
    {
      "entropy": 5.496813678741455,
      "epoch": 1.3199682413656213,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004833302515416355,
      "loss": 5.3392,
      "mean_token_accuracy": 0.1576893538236618,
      "num_tokens": 29165475.0,
      "step": 13300
    },
    {
      "entropy": 5.419232559204102,
      "epoch": 1.3204644700277888,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048331686511047584,
      "loss": 5.1866,
      "mean_token_accuracy": 0.16232384145259857,
      "num_tokens": 29177139.0,
      "step": 13305
    },
    {
      "entropy": 5.4873758316040036,
      "epoch": 1.3209606986899565,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048330347351350676,
      "loss": 5.2603,
      "mean_token_accuracy": 0.16144658923149108,
      "num_tokens": 29187812.0,
      "step": 13310
    },
    {
      "entropy": 5.374246120452881,
      "epoch": 1.321456927352124,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048329007675106036,
      "loss": 5.134,
      "mean_token_accuracy": 0.1635843724012375,
      "num_tokens": 29197932.0,
      "step": 13315
    },
    {
      "entropy": 5.393495321273804,
      "epoch": 1.3219531560142914,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048327667482346895,
      "loss": 5.1641,
      "mean_token_accuracy": 0.1646159902215004,
      "num_tokens": 29209529.0,
      "step": 13320
    },
    {
      "entropy": 5.437481212615967,
      "epoch": 1.322449384676459,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004832632677310647,
      "loss": 5.2287,
      "mean_token_accuracy": 0.16262610107660294,
      "num_tokens": 29219027.0,
      "step": 13325
    },
    {
      "entropy": 5.430460739135742,
      "epoch": 1.3229456133386264,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048324985547418026,
      "loss": 5.2683,
      "mean_token_accuracy": 0.15471749305725097,
      "num_tokens": 29229647.0,
      "step": 13330
    },
    {
      "entropy": 5.503086090087891,
      "epoch": 1.3234418420007938,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004832364380531482,
      "loss": 5.3008,
      "mean_token_accuracy": 0.16816799491643905,
      "num_tokens": 29240263.0,
      "step": 13335
    },
    {
      "entropy": 5.507397079467774,
      "epoch": 1.3239380706629615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048322301546830103,
      "loss": 5.3063,
      "mean_token_accuracy": 0.15139021575450898,
      "num_tokens": 29251115.0,
      "step": 13340
    },
    {
      "entropy": 5.426276111602784,
      "epoch": 1.324434299325129,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004832095877199718,
      "loss": 5.2769,
      "mean_token_accuracy": 0.15310217440128326,
      "num_tokens": 29261888.0,
      "step": 13345
    },
    {
      "entropy": 5.386424016952515,
      "epoch": 1.3249305279872965,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048319615480849347,
      "loss": 5.162,
      "mean_token_accuracy": 0.16827259361743926,
      "num_tokens": 29271948.0,
      "step": 13350
    },
    {
      "entropy": 5.467912197113037,
      "epoch": 1.3254267566494642,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048318271673419914,
      "loss": 5.2269,
      "mean_token_accuracy": 0.1585451304912567,
      "num_tokens": 29282449.0,
      "step": 13355
    },
    {
      "entropy": 5.451964282989502,
      "epoch": 1.3259229853116317,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048316927349742207,
      "loss": 5.2465,
      "mean_token_accuracy": 0.1593168169260025,
      "num_tokens": 29294280.0,
      "step": 13360
    },
    {
      "entropy": 5.45820198059082,
      "epoch": 1.3264192139737991,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004831558250984955,
      "loss": 5.2387,
      "mean_token_accuracy": 0.1614450052380562,
      "num_tokens": 29305143.0,
      "step": 13365
    },
    {
      "entropy": 5.43507981300354,
      "epoch": 1.3269154426359666,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048314237153775303,
      "loss": 5.1494,
      "mean_token_accuracy": 0.16275084018707275,
      "num_tokens": 29314697.0,
      "step": 13370
    },
    {
      "entropy": 5.3810076236724855,
      "epoch": 1.327411671298134,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004831289128155283,
      "loss": 5.1841,
      "mean_token_accuracy": 0.16088908016681672,
      "num_tokens": 29325257.0,
      "step": 13375
    },
    {
      "entropy": 5.443231010437012,
      "epoch": 1.3279078999603018,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048311544893215496,
      "loss": 5.2829,
      "mean_token_accuracy": 0.15415675044059754,
      "num_tokens": 29335449.0,
      "step": 13380
    },
    {
      "entropy": 5.409874534606933,
      "epoch": 1.3284041286224693,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000483101979887967,
      "loss": 5.2761,
      "mean_token_accuracy": 0.15508345365524293,
      "num_tokens": 29346898.0,
      "step": 13385
    },
    {
      "entropy": 5.533391523361206,
      "epoch": 1.3289003572846367,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048308850568329825,
      "loss": 5.2486,
      "mean_token_accuracy": 0.1561342492699623,
      "num_tokens": 29356118.0,
      "step": 13390
    },
    {
      "entropy": 5.44662561416626,
      "epoch": 1.3293965859468042,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048307502631848306,
      "loss": 5.3085,
      "mean_token_accuracy": 0.15349682420492172,
      "num_tokens": 29367597.0,
      "step": 13395
    },
    {
      "entropy": 5.4303826808929445,
      "epoch": 1.329892814608972,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048306154179385546,
      "loss": 5.1839,
      "mean_token_accuracy": 0.16108398735523224,
      "num_tokens": 29377829.0,
      "step": 13400
    },
    {
      "entropy": 5.435200357437134,
      "epoch": 1.3303890432711394,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004830480521097501,
      "loss": 5.1443,
      "mean_token_accuracy": 0.16647381484508514,
      "num_tokens": 29387945.0,
      "step": 13405
    },
    {
      "entropy": 5.350883197784424,
      "epoch": 1.3308852719333069,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004830345572665012,
      "loss": 5.1278,
      "mean_token_accuracy": 0.16601821184158325,
      "num_tokens": 29398269.0,
      "step": 13410
    },
    {
      "entropy": 5.464433670043945,
      "epoch": 1.3313815005954743,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048302105726444366,
      "loss": 5.3021,
      "mean_token_accuracy": 0.1527259513735771,
      "num_tokens": 29410946.0,
      "step": 13415
    },
    {
      "entropy": 5.431884860992431,
      "epoch": 1.3318777292576418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048300755210391205,
      "loss": 5.1375,
      "mean_token_accuracy": 0.16259538531303405,
      "num_tokens": 29421770.0,
      "step": 13420
    },
    {
      "entropy": 5.464356756210327,
      "epoch": 1.3323739579198095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004829940417852415,
      "loss": 5.304,
      "mean_token_accuracy": 0.1601225182414055,
      "num_tokens": 29433717.0,
      "step": 13425
    },
    {
      "entropy": 5.343073511123658,
      "epoch": 1.332870186581977,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004829805263087668,
      "loss": 5.137,
      "mean_token_accuracy": 0.16888256669044494,
      "num_tokens": 29444382.0,
      "step": 13430
    },
    {
      "entropy": 5.4375629901885985,
      "epoch": 1.3333664152441445,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048296700567482324,
      "loss": 5.2669,
      "mean_token_accuracy": 0.1564920425415039,
      "num_tokens": 29454680.0,
      "step": 13435
    },
    {
      "entropy": 5.5113794803619385,
      "epoch": 1.3338626439063122,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004829534798837461,
      "loss": 5.2806,
      "mean_token_accuracy": 0.1601371943950653,
      "num_tokens": 29465065.0,
      "step": 13440
    },
    {
      "entropy": 5.474467086791992,
      "epoch": 1.3343588725684796,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004829399489358707,
      "loss": 5.2094,
      "mean_token_accuracy": 0.1600073143839836,
      "num_tokens": 29475704.0,
      "step": 13445
    },
    {
      "entropy": 5.407029533386231,
      "epoch": 1.334855101230647,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004829264128315326,
      "loss": 5.2439,
      "mean_token_accuracy": 0.1586716890335083,
      "num_tokens": 29486418.0,
      "step": 13450
    },
    {
      "entropy": 5.511453151702881,
      "epoch": 1.3353513298928146,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004829128715710676,
      "loss": 5.3336,
      "mean_token_accuracy": 0.15129120871424676,
      "num_tokens": 29498939.0,
      "step": 13455
    },
    {
      "entropy": 5.426885366439819,
      "epoch": 1.335847558554982,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004828993251548114,
      "loss": 5.1485,
      "mean_token_accuracy": 0.16758292466402053,
      "num_tokens": 29508525.0,
      "step": 13460
    },
    {
      "entropy": 5.425980806350708,
      "epoch": 1.3363437872171495,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004828857735830999,
      "loss": 5.2782,
      "mean_token_accuracy": 0.16180317848920822,
      "num_tokens": 29519962.0,
      "step": 13465
    },
    {
      "entropy": 5.441163492202759,
      "epoch": 1.3368400158793172,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004828722168562691,
      "loss": 5.2355,
      "mean_token_accuracy": 0.1576477900147438,
      "num_tokens": 29529586.0,
      "step": 13470
    },
    {
      "entropy": 5.494522666931152,
      "epoch": 1.3373362445414847,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004828586549746553,
      "loss": 5.2983,
      "mean_token_accuracy": 0.15155503302812576,
      "num_tokens": 29540670.0,
      "step": 13475
    },
    {
      "entropy": 5.450711822509765,
      "epoch": 1.3378324732036522,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004828450879385947,
      "loss": 5.2803,
      "mean_token_accuracy": 0.15215998515486717,
      "num_tokens": 29551484.0,
      "step": 13480
    },
    {
      "entropy": 5.483923244476318,
      "epoch": 1.3383287018658199,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004828315157484239,
      "loss": 5.2163,
      "mean_token_accuracy": 0.15837326049804687,
      "num_tokens": 29561596.0,
      "step": 13485
    },
    {
      "entropy": 5.483451509475708,
      "epoch": 1.3388249305279873,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004828179384044793,
      "loss": 5.224,
      "mean_token_accuracy": 0.1604718819260597,
      "num_tokens": 29572837.0,
      "step": 13490
    },
    {
      "entropy": 5.40222487449646,
      "epoch": 1.3393211591901548,
      "grad_norm": 1.125,
      "learning_rate": 0.00048280435590709756,
      "loss": 5.1965,
      "mean_token_accuracy": 0.15962852239608766,
      "num_tokens": 29584095.0,
      "step": 13495
    },
    {
      "entropy": 5.430434370040894,
      "epoch": 1.3398173878523223,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004827907682566157,
      "loss": 5.242,
      "mean_token_accuracy": 0.155342960357666,
      "num_tokens": 29594903.0,
      "step": 13500
    },
    {
      "entropy": 5.441515970230102,
      "epoch": 1.3403136165144898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004827771754533704,
      "loss": 5.2032,
      "mean_token_accuracy": 0.1633856937289238,
      "num_tokens": 29605199.0,
      "step": 13505
    },
    {
      "entropy": 5.499683475494384,
      "epoch": 1.3408098451766575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048276357749769895,
      "loss": 5.3698,
      "mean_token_accuracy": 0.15231950134038924,
      "num_tokens": 29617535.0,
      "step": 13510
    },
    {
      "entropy": 5.402033281326294,
      "epoch": 1.341306073838825,
      "grad_norm": 1.125,
      "learning_rate": 0.00048274997438993845,
      "loss": 5.1805,
      "mean_token_accuracy": 0.16218443512916564,
      "num_tokens": 29628227.0,
      "step": 13515
    },
    {
      "entropy": 5.440347623825073,
      "epoch": 1.3418023025009924,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048273636613042626,
      "loss": 5.3096,
      "mean_token_accuracy": 0.15728188157081605,
      "num_tokens": 29639112.0,
      "step": 13520
    },
    {
      "entropy": 5.4324558734893795,
      "epoch": 1.3422985311631601,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004827227527194998,
      "loss": 5.1871,
      "mean_token_accuracy": 0.16021134406328202,
      "num_tokens": 29650768.0,
      "step": 13525
    },
    {
      "entropy": 5.476371479034424,
      "epoch": 1.3427947598253276,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004827091341574967,
      "loss": 5.3054,
      "mean_token_accuracy": 0.16274742782115936,
      "num_tokens": 29662388.0,
      "step": 13530
    },
    {
      "entropy": 5.4228832721710205,
      "epoch": 1.343290988487495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004826955104447546,
      "loss": 5.1887,
      "mean_token_accuracy": 0.1663537934422493,
      "num_tokens": 29672231.0,
      "step": 13535
    },
    {
      "entropy": 5.396336030960083,
      "epoch": 1.3437872171496625,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004826818815816114,
      "loss": 5.2363,
      "mean_token_accuracy": 0.16658409386873246,
      "num_tokens": 29683019.0,
      "step": 13540
    },
    {
      "entropy": 5.42900652885437,
      "epoch": 1.34428344581183,
      "grad_norm": 1.125,
      "learning_rate": 0.00048266824756840513,
      "loss": 5.2532,
      "mean_token_accuracy": 0.16128573268651963,
      "num_tokens": 29694233.0,
      "step": 13545
    },
    {
      "entropy": 5.480251789093018,
      "epoch": 1.3447796744739975,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004826546084054737,
      "loss": 5.2799,
      "mean_token_accuracy": 0.15953433960676194,
      "num_tokens": 29704023.0,
      "step": 13550
    },
    {
      "entropy": 5.551623106002808,
      "epoch": 1.3452759031361652,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004826409640931555,
      "loss": 5.3629,
      "mean_token_accuracy": 0.15119170248508454,
      "num_tokens": 29715407.0,
      "step": 13555
    },
    {
      "entropy": 5.430735683441162,
      "epoch": 1.3457721317983327,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048262731463178884,
      "loss": 5.2089,
      "mean_token_accuracy": 0.16168544739484786,
      "num_tokens": 29726223.0,
      "step": 13560
    },
    {
      "entropy": 5.388166189193726,
      "epoch": 1.3462683604605001,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004826136600217121,
      "loss": 5.1215,
      "mean_token_accuracy": 0.16803857535123826,
      "num_tokens": 29735850.0,
      "step": 13565
    },
    {
      "entropy": 5.4601874351501465,
      "epoch": 1.3467645891226678,
      "grad_norm": 1.1875,
      "learning_rate": 0.000482600000263264,
      "loss": 5.2993,
      "mean_token_accuracy": 0.15464717000722886,
      "num_tokens": 29747381.0,
      "step": 13570
    },
    {
      "entropy": 5.512057304382324,
      "epoch": 1.3472608177848353,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004825863353567832,
      "loss": 5.3442,
      "mean_token_accuracy": 0.15404024124145507,
      "num_tokens": 29758440.0,
      "step": 13575
    },
    {
      "entropy": 5.414356374740601,
      "epoch": 1.3477570464470028,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048257266530260865,
      "loss": 5.1814,
      "mean_token_accuracy": 0.15866454392671586,
      "num_tokens": 29770606.0,
      "step": 13580
    },
    {
      "entropy": 5.387361192703247,
      "epoch": 1.3482532751091703,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004825589901010792,
      "loss": 5.2579,
      "mean_token_accuracy": 0.1595859557390213,
      "num_tokens": 29781469.0,
      "step": 13585
    },
    {
      "entropy": 5.469564390182495,
      "epoch": 1.3487495037713377,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004825453097525341,
      "loss": 5.2973,
      "mean_token_accuracy": 0.15766896456480026,
      "num_tokens": 29792847.0,
      "step": 13590
    },
    {
      "entropy": 5.453678703308105,
      "epoch": 1.3492457324335054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048253162425731253,
      "loss": 5.2254,
      "mean_token_accuracy": 0.15540353506803511,
      "num_tokens": 29802784.0,
      "step": 13595
    },
    {
      "entropy": 5.515305137634277,
      "epoch": 1.349741961095673,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004825179336157539,
      "loss": 5.2222,
      "mean_token_accuracy": 0.1643040344119072,
      "num_tokens": 29812849.0,
      "step": 13600
    },
    {
      "entropy": 5.421775054931641,
      "epoch": 1.3502381897578404,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048250423782819765,
      "loss": 5.2127,
      "mean_token_accuracy": 0.16120753288269044,
      "num_tokens": 29823495.0,
      "step": 13605
    },
    {
      "entropy": 5.405583810806275,
      "epoch": 1.350734418420008,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004824905368949834,
      "loss": 5.2619,
      "mean_token_accuracy": 0.15832506120204926,
      "num_tokens": 29835295.0,
      "step": 13610
    },
    {
      "entropy": 5.421481323242188,
      "epoch": 1.3512306470821756,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004824768308164509,
      "loss": 5.1894,
      "mean_token_accuracy": 0.1663385197520256,
      "num_tokens": 29845747.0,
      "step": 13615
    },
    {
      "entropy": 5.469134950637818,
      "epoch": 1.351726875744343,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004824631195929402,
      "loss": 5.2468,
      "mean_token_accuracy": 0.16497644037008286,
      "num_tokens": 29857111.0,
      "step": 13620
    },
    {
      "entropy": 5.395828628540039,
      "epoch": 1.3522231044065105,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004824494032247911,
      "loss": 5.2413,
      "mean_token_accuracy": 0.16224489510059356,
      "num_tokens": 29866919.0,
      "step": 13625
    },
    {
      "entropy": 5.408735227584839,
      "epoch": 1.352719333068678,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004824356817123439,
      "loss": 5.1904,
      "mean_token_accuracy": 0.16026580929756165,
      "num_tokens": 29878631.0,
      "step": 13630
    },
    {
      "entropy": 5.442997312545776,
      "epoch": 1.3532155617308455,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048242195505593865,
      "loss": 5.2125,
      "mean_token_accuracy": 0.1656864896416664,
      "num_tokens": 29890809.0,
      "step": 13635
    },
    {
      "entropy": 5.482135581970215,
      "epoch": 1.3537117903930131,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048240822325591596,
      "loss": 5.3021,
      "mean_token_accuracy": 0.16234626471996308,
      "num_tokens": 29901789.0,
      "step": 13640
    },
    {
      "entropy": 5.4467644691467285,
      "epoch": 1.3542080190551806,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004823944863126162,
      "loss": 5.2298,
      "mean_token_accuracy": 0.15847614854574205,
      "num_tokens": 29912576.0,
      "step": 13645
    },
    {
      "entropy": 5.474138164520264,
      "epoch": 1.354704247717348,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004823807442263801,
      "loss": 5.307,
      "mean_token_accuracy": 0.15411677062511445,
      "num_tokens": 29924076.0,
      "step": 13650
    },
    {
      "entropy": 5.4127952575683596,
      "epoch": 1.3552004763795158,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048236699699754845,
      "loss": 5.2241,
      "mean_token_accuracy": 0.163233582675457,
      "num_tokens": 29935854.0,
      "step": 13655
    },
    {
      "entropy": 5.463515424728394,
      "epoch": 1.3556967050416833,
      "grad_norm": 1.015625,
      "learning_rate": 0.000482353244626462,
      "loss": 5.1903,
      "mean_token_accuracy": 0.15704114437103273,
      "num_tokens": 29946389.0,
      "step": 13660
    },
    {
      "entropy": 5.43990912437439,
      "epoch": 1.3561929337038507,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048233948711346196,
      "loss": 5.2465,
      "mean_token_accuracy": 0.151458241045475,
      "num_tokens": 29958170.0,
      "step": 13665
    },
    {
      "entropy": 5.387790632247925,
      "epoch": 1.3566891623660182,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048232572445888945,
      "loss": 5.147,
      "mean_token_accuracy": 0.16378505080938338,
      "num_tokens": 29969067.0,
      "step": 13670
    },
    {
      "entropy": 5.4705534934997555,
      "epoch": 1.3571853910281857,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004823119566630857,
      "loss": 5.2865,
      "mean_token_accuracy": 0.1534193530678749,
      "num_tokens": 29979727.0,
      "step": 13675
    },
    {
      "entropy": 5.490022277832031,
      "epoch": 1.3576816196903534,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048229818372639216,
      "loss": 5.2444,
      "mean_token_accuracy": 0.16055328994989396,
      "num_tokens": 29989397.0,
      "step": 13680
    },
    {
      "entropy": 5.330842733383179,
      "epoch": 1.3581778483525209,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048228440564915026,
      "loss": 5.1662,
      "mean_token_accuracy": 0.15625654011964799,
      "num_tokens": 29999415.0,
      "step": 13685
    },
    {
      "entropy": 5.403040361404419,
      "epoch": 1.3586740770146883,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048227062243170186,
      "loss": 5.2101,
      "mean_token_accuracy": 0.16178561598062516,
      "num_tokens": 30010277.0,
      "step": 13690
    },
    {
      "entropy": 5.377931404113769,
      "epoch": 1.3591703056768558,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004822568340743886,
      "loss": 5.115,
      "mean_token_accuracy": 0.16588497310876846,
      "num_tokens": 30020684.0,
      "step": 13695
    },
    {
      "entropy": 5.500359344482422,
      "epoch": 1.3596665343390235,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004822430405775525,
      "loss": 5.3251,
      "mean_token_accuracy": 0.15306295901536943,
      "num_tokens": 30030916.0,
      "step": 13700
    },
    {
      "entropy": 5.4920717716217045,
      "epoch": 1.360162763001191,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004822292419415355,
      "loss": 5.245,
      "mean_token_accuracy": 0.15852225571870804,
      "num_tokens": 30041140.0,
      "step": 13705
    },
    {
      "entropy": 5.4024505615234375,
      "epoch": 1.3606589916633585,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048221543816667984,
      "loss": 5.1899,
      "mean_token_accuracy": 0.16754062920808793,
      "num_tokens": 30052765.0,
      "step": 13710
    },
    {
      "entropy": 5.4366247177124025,
      "epoch": 1.361155220325526,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048220162925332786,
      "loss": 5.229,
      "mean_token_accuracy": 0.15959538221359254,
      "num_tokens": 30062511.0,
      "step": 13715
    },
    {
      "entropy": 5.529403448104858,
      "epoch": 1.3616514489876934,
      "grad_norm": 1.25,
      "learning_rate": 0.000482187815201822,
      "loss": 5.319,
      "mean_token_accuracy": 0.15796130299568176,
      "num_tokens": 30073404.0,
      "step": 13720
    },
    {
      "entropy": 5.3799402713775635,
      "epoch": 1.362147677649861,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048217399601250464,
      "loss": 5.1645,
      "mean_token_accuracy": 0.16428885310888292,
      "num_tokens": 30084978.0,
      "step": 13725
    },
    {
      "entropy": 5.359703063964844,
      "epoch": 1.3626439063120286,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004821601716857187,
      "loss": 5.2336,
      "mean_token_accuracy": 0.16382355242967606,
      "num_tokens": 30094498.0,
      "step": 13730
    },
    {
      "entropy": 5.430843830108643,
      "epoch": 1.363140134974196,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048214634222180686,
      "loss": 5.2616,
      "mean_token_accuracy": 0.16363436877727508,
      "num_tokens": 30104545.0,
      "step": 13735
    },
    {
      "entropy": 5.553267288208008,
      "epoch": 1.3636363636363638,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004821325076211121,
      "loss": 5.3171,
      "mean_token_accuracy": 0.15792741924524306,
      "num_tokens": 30114863.0,
      "step": 13740
    },
    {
      "entropy": 5.496372318267822,
      "epoch": 1.3641325922985312,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048211866788397753,
      "loss": 5.2355,
      "mean_token_accuracy": 0.1519894555211067,
      "num_tokens": 30125101.0,
      "step": 13745
    },
    {
      "entropy": 5.437227964401245,
      "epoch": 1.3646288209606987,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004821048230107462,
      "loss": 5.3051,
      "mean_token_accuracy": 0.1604721412062645,
      "num_tokens": 30135081.0,
      "step": 13750
    },
    {
      "entropy": 5.5649329662323,
      "epoch": 1.3651250496228662,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004820909730017616,
      "loss": 5.3235,
      "mean_token_accuracy": 0.1510474294424057,
      "num_tokens": 30147169.0,
      "step": 13755
    },
    {
      "entropy": 5.449578905105591,
      "epoch": 1.3656212782850337,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048207711785736707,
      "loss": 5.1381,
      "mean_token_accuracy": 0.16325948238372803,
      "num_tokens": 30157523.0,
      "step": 13760
    },
    {
      "entropy": 5.404944372177124,
      "epoch": 1.3661175069472011,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048206325757790624,
      "loss": 5.2159,
      "mean_token_accuracy": 0.16048824787139893,
      "num_tokens": 30168313.0,
      "step": 13765
    },
    {
      "entropy": 5.359929084777832,
      "epoch": 1.3666137356093688,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048204939216372285,
      "loss": 5.1799,
      "mean_token_accuracy": 0.16079125851392745,
      "num_tokens": 30179173.0,
      "step": 13770
    },
    {
      "entropy": 5.4698409080505375,
      "epoch": 1.3671099642715363,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004820355216151607,
      "loss": 5.2544,
      "mean_token_accuracy": 0.1573948308825493,
      "num_tokens": 30191666.0,
      "step": 13775
    },
    {
      "entropy": 5.485439920425415,
      "epoch": 1.3676061929337038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048202164593256363,
      "loss": 5.2177,
      "mean_token_accuracy": 0.15971429497003556,
      "num_tokens": 30202752.0,
      "step": 13780
    },
    {
      "entropy": 5.414976119995117,
      "epoch": 1.3681024215958715,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048200776511627585,
      "loss": 5.2432,
      "mean_token_accuracy": 0.15923550724983215,
      "num_tokens": 30213127.0,
      "step": 13785
    },
    {
      "entropy": 5.442012643814087,
      "epoch": 1.368598650258039,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004819938791666416,
      "loss": 5.268,
      "mean_token_accuracy": 0.1593529224395752,
      "num_tokens": 30224782.0,
      "step": 13790
    },
    {
      "entropy": 5.473060083389282,
      "epoch": 1.3690948789202064,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048197998808400524,
      "loss": 5.282,
      "mean_token_accuracy": 0.1625913366675377,
      "num_tokens": 30235129.0,
      "step": 13795
    },
    {
      "entropy": 5.473362636566162,
      "epoch": 1.369591107582374,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004819660918687111,
      "loss": 5.2908,
      "mean_token_accuracy": 0.15510911643505096,
      "num_tokens": 30247304.0,
      "step": 13800
    },
    {
      "entropy": 5.4014424800872805,
      "epoch": 1.3700873362445414,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004819521905211039,
      "loss": 5.1706,
      "mean_token_accuracy": 0.16428121775388718,
      "num_tokens": 30257900.0,
      "step": 13805
    },
    {
      "entropy": 5.510814237594604,
      "epoch": 1.370583564906709,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048193828404152826,
      "loss": 5.3085,
      "mean_token_accuracy": 0.15991924107074737,
      "num_tokens": 30269741.0,
      "step": 13810
    },
    {
      "entropy": 5.4953656673431395,
      "epoch": 1.3710797935688765,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004819243724303291,
      "loss": 5.2937,
      "mean_token_accuracy": 0.15457017049193383,
      "num_tokens": 30279382.0,
      "step": 13815
    },
    {
      "entropy": 5.4547172546386715,
      "epoch": 1.371576022231044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048191045568785144,
      "loss": 5.2623,
      "mean_token_accuracy": 0.16125618666410446,
      "num_tokens": 30289837.0,
      "step": 13820
    },
    {
      "entropy": 5.391208982467651,
      "epoch": 1.3720722508932117,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048189653381444024,
      "loss": 5.2229,
      "mean_token_accuracy": 0.16227136105298995,
      "num_tokens": 30302138.0,
      "step": 13825
    },
    {
      "entropy": 5.479448652267456,
      "epoch": 1.3725684795553792,
      "grad_norm": 1.125,
      "learning_rate": 0.0004818826068104409,
      "loss": 5.2867,
      "mean_token_accuracy": 0.16138470470905303,
      "num_tokens": 30313623.0,
      "step": 13830
    },
    {
      "entropy": 5.479711246490479,
      "epoch": 1.3730647082175467,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004818686746761987,
      "loss": 5.2429,
      "mean_token_accuracy": 0.15650436282157898,
      "num_tokens": 30324533.0,
      "step": 13835
    },
    {
      "entropy": 5.500314521789551,
      "epoch": 1.3735609368797141,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004818547374120591,
      "loss": 5.3307,
      "mean_token_accuracy": 0.1521144241094589,
      "num_tokens": 30335819.0,
      "step": 13840
    },
    {
      "entropy": 5.432929039001465,
      "epoch": 1.3740571655418816,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004818407950183678,
      "loss": 5.2264,
      "mean_token_accuracy": 0.1644834890961647,
      "num_tokens": 30345947.0,
      "step": 13845
    },
    {
      "entropy": 5.363990211486817,
      "epoch": 1.374553394204049,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004818268474954704,
      "loss": 5.2007,
      "mean_token_accuracy": 0.1606699764728546,
      "num_tokens": 30356972.0,
      "step": 13850
    },
    {
      "entropy": 5.511010646820068,
      "epoch": 1.3750496228662168,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004818128948437129,
      "loss": 5.2834,
      "mean_token_accuracy": 0.15604110509157182,
      "num_tokens": 30368414.0,
      "step": 13855
    },
    {
      "entropy": 5.472829723358155,
      "epoch": 1.3755458515283843,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004817989370634412,
      "loss": 5.2097,
      "mean_token_accuracy": 0.16205856651067735,
      "num_tokens": 30379141.0,
      "step": 13860
    },
    {
      "entropy": 5.408936214447022,
      "epoch": 1.3760420801905517,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048178497415500153,
      "loss": 5.1648,
      "mean_token_accuracy": 0.16866973042488098,
      "num_tokens": 30388985.0,
      "step": 13865
    },
    {
      "entropy": 5.406170463562011,
      "epoch": 1.3765383088527194,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004817710061187401,
      "loss": 5.2051,
      "mean_token_accuracy": 0.1689336195588112,
      "num_tokens": 30399599.0,
      "step": 13870
    },
    {
      "entropy": 5.439347934722901,
      "epoch": 1.377034537514887,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048175703295500323,
      "loss": 5.2772,
      "mean_token_accuracy": 0.16349248439073563,
      "num_tokens": 30411143.0,
      "step": 13875
    },
    {
      "entropy": 5.46564564704895,
      "epoch": 1.3775307661770544,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048174305466413755,
      "loss": 5.2799,
      "mean_token_accuracy": 0.15885397493839265,
      "num_tokens": 30422339.0,
      "step": 13880
    },
    {
      "entropy": 5.403530263900757,
      "epoch": 1.3780269948392219,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048172907124648947,
      "loss": 5.1893,
      "mean_token_accuracy": 0.16385374814271927,
      "num_tokens": 30432588.0,
      "step": 13885
    },
    {
      "entropy": 5.464610815048218,
      "epoch": 1.3785232235013893,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004817150827024059,
      "loss": 5.2698,
      "mean_token_accuracy": 0.16011369377374648,
      "num_tokens": 30442272.0,
      "step": 13890
    },
    {
      "entropy": 5.405564832687378,
      "epoch": 1.379019452163557,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004817010890322338,
      "loss": 5.1885,
      "mean_token_accuracy": 0.16222198754549028,
      "num_tokens": 30453142.0,
      "step": 13895
    },
    {
      "entropy": 5.414975595474243,
      "epoch": 1.3795156808257245,
      "grad_norm": 1.125,
      "learning_rate": 0.00048168709023632006,
      "loss": 5.2889,
      "mean_token_accuracy": 0.15799435824155808,
      "num_tokens": 30465136.0,
      "step": 13900
    },
    {
      "entropy": 5.507431316375732,
      "epoch": 1.380011909487892,
      "grad_norm": 1.25,
      "learning_rate": 0.00048167308631501184,
      "loss": 5.202,
      "mean_token_accuracy": 0.16517867147922516,
      "num_tokens": 30476511.0,
      "step": 13905
    },
    {
      "entropy": 5.47611927986145,
      "epoch": 1.3805081381500597,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048165907726865643,
      "loss": 5.3198,
      "mean_token_accuracy": 0.15479688793420793,
      "num_tokens": 30486547.0,
      "step": 13910
    },
    {
      "entropy": 5.478510427474975,
      "epoch": 1.3810043668122272,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004816450630976012,
      "loss": 5.2687,
      "mean_token_accuracy": 0.16752270013093948,
      "num_tokens": 30496163.0,
      "step": 13915
    },
    {
      "entropy": 5.429570579528809,
      "epoch": 1.3815005954743946,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004816310438021937,
      "loss": 5.2745,
      "mean_token_accuracy": 0.1640395849943161,
      "num_tokens": 30506979.0,
      "step": 13920
    },
    {
      "entropy": 5.442856645584106,
      "epoch": 1.381996824136562,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048161701938278156,
      "loss": 5.3034,
      "mean_token_accuracy": 0.15072427988052367,
      "num_tokens": 30517492.0,
      "step": 13925
    },
    {
      "entropy": 5.425680303573609,
      "epoch": 1.3824930527987296,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048160298983971256,
      "loss": 5.1838,
      "mean_token_accuracy": 0.16577876061201097,
      "num_tokens": 30528614.0,
      "step": 13930
    },
    {
      "entropy": 5.407075977325439,
      "epoch": 1.382989281460897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004815889551733346,
      "loss": 5.2845,
      "mean_token_accuracy": 0.15926803946495055,
      "num_tokens": 30539479.0,
      "step": 13935
    },
    {
      "entropy": 5.496245718002319,
      "epoch": 1.3834855101230648,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048157491538399573,
      "loss": 5.2977,
      "mean_token_accuracy": 0.15827858000993728,
      "num_tokens": 30549909.0,
      "step": 13940
    },
    {
      "entropy": 5.455719804763794,
      "epoch": 1.3839817387852322,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048156087047204407,
      "loss": 5.1913,
      "mean_token_accuracy": 0.1606842204928398,
      "num_tokens": 30562265.0,
      "step": 13945
    },
    {
      "entropy": 5.453008985519409,
      "epoch": 1.3844779674473997,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004815468204378279,
      "loss": 5.2116,
      "mean_token_accuracy": 0.1590519830584526,
      "num_tokens": 30572622.0,
      "step": 13950
    },
    {
      "entropy": 5.3937078475952145,
      "epoch": 1.3849741961095674,
      "grad_norm": 1.125,
      "learning_rate": 0.0004815327652816957,
      "loss": 5.1906,
      "mean_token_accuracy": 0.16617387533187866,
      "num_tokens": 30583300.0,
      "step": 13955
    },
    {
      "entropy": 5.394505596160888,
      "epoch": 1.3854704247717349,
      "grad_norm": 1.125,
      "learning_rate": 0.0004815187050039959,
      "loss": 5.1645,
      "mean_token_accuracy": 0.16272280663251876,
      "num_tokens": 30594460.0,
      "step": 13960
    },
    {
      "entropy": 5.425444030761719,
      "epoch": 1.3859666534339024,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048150463960507724,
      "loss": 5.206,
      "mean_token_accuracy": 0.16154352724552154,
      "num_tokens": 30604658.0,
      "step": 13965
    },
    {
      "entropy": 5.46236629486084,
      "epoch": 1.3864628820960698,
      "grad_norm": 1.0,
      "learning_rate": 0.00048149056908528855,
      "loss": 5.2083,
      "mean_token_accuracy": 0.15810299068689346,
      "num_tokens": 30616151.0,
      "step": 13970
    },
    {
      "entropy": 5.458207464218139,
      "epoch": 1.3869591107582373,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004814764934449786,
      "loss": 5.2618,
      "mean_token_accuracy": 0.16189065873622893,
      "num_tokens": 30627560.0,
      "step": 13975
    },
    {
      "entropy": 5.367296314239502,
      "epoch": 1.387455339420405,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048146241268449665,
      "loss": 5.2384,
      "mean_token_accuracy": 0.1608617216348648,
      "num_tokens": 30638490.0,
      "step": 13980
    },
    {
      "entropy": 5.458796262741089,
      "epoch": 1.3879515680825725,
      "grad_norm": 1.125,
      "learning_rate": 0.0004814483268041917,
      "loss": 5.3418,
      "mean_token_accuracy": 0.1550677239894867,
      "num_tokens": 30649501.0,
      "step": 13985
    },
    {
      "entropy": 5.498704195022583,
      "epoch": 1.38844779674474,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004814342358044131,
      "loss": 5.2173,
      "mean_token_accuracy": 0.159960974752903,
      "num_tokens": 30660479.0,
      "step": 13990
    },
    {
      "entropy": 5.595318984985352,
      "epoch": 1.3889440254069074,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048142013968551026,
      "loss": 5.3987,
      "mean_token_accuracy": 0.14971231818199157,
      "num_tokens": 30672015.0,
      "step": 13995
    },
    {
      "entropy": 5.386664867401123,
      "epoch": 1.3894402540690751,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048140603844783273,
      "loss": 5.2518,
      "mean_token_accuracy": 0.15733387172222138,
      "num_tokens": 30682612.0,
      "step": 14000
    },
    {
      "entropy": 5.378702020645141,
      "epoch": 1.3899364827312426,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004813919320917303,
      "loss": 5.1833,
      "mean_token_accuracy": 0.15952119827270508,
      "num_tokens": 30693289.0,
      "step": 14005
    },
    {
      "entropy": 5.444607591629028,
      "epoch": 1.39043271139341,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048137782061755265,
      "loss": 5.2225,
      "mean_token_accuracy": 0.1639940544962883,
      "num_tokens": 30705346.0,
      "step": 14010
    },
    {
      "entropy": 5.405619382858276,
      "epoch": 1.3909289400555775,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048136370402564974,
      "loss": 5.2206,
      "mean_token_accuracy": 0.1584940955042839,
      "num_tokens": 30715968.0,
      "step": 14015
    },
    {
      "entropy": 5.419011878967285,
      "epoch": 1.391425168717745,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004813495823163716,
      "loss": 5.1935,
      "mean_token_accuracy": 0.1632399931550026,
      "num_tokens": 30726225.0,
      "step": 14020
    },
    {
      "entropy": 5.362982082366943,
      "epoch": 1.3919213973799127,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004813354554900685,
      "loss": 5.1536,
      "mean_token_accuracy": 0.1633608803153038,
      "num_tokens": 30735603.0,
      "step": 14025
    },
    {
      "entropy": 5.386462926864624,
      "epoch": 1.3924176260420802,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004813213235470907,
      "loss": 5.2222,
      "mean_token_accuracy": 0.1558401256799698,
      "num_tokens": 30747414.0,
      "step": 14030
    },
    {
      "entropy": 5.51989974975586,
      "epoch": 1.3929138547042477,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004813071864877887,
      "loss": 5.2705,
      "mean_token_accuracy": 0.15844389349222182,
      "num_tokens": 30758389.0,
      "step": 14035
    },
    {
      "entropy": 5.43652753829956,
      "epoch": 1.3934100833664154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000481293044312513,
      "loss": 5.2817,
      "mean_token_accuracy": 0.15899729877710342,
      "num_tokens": 30769484.0,
      "step": 14040
    },
    {
      "entropy": 5.349252510070801,
      "epoch": 1.3939063120285828,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048127889702161435,
      "loss": 5.202,
      "mean_token_accuracy": 0.16596420705318451,
      "num_tokens": 30779775.0,
      "step": 14045
    },
    {
      "entropy": 5.455591535568237,
      "epoch": 1.3944025406907503,
      "grad_norm": 1.125,
      "learning_rate": 0.0004812647446154435,
      "loss": 5.3136,
      "mean_token_accuracy": 0.1548160955309868,
      "num_tokens": 30791456.0,
      "step": 14050
    },
    {
      "entropy": 5.416250371932984,
      "epoch": 1.3948987693529178,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004812505870943515,
      "loss": 5.1213,
      "mean_token_accuracy": 0.16653395742177962,
      "num_tokens": 30801659.0,
      "step": 14055
    },
    {
      "entropy": 5.5321601867675785,
      "epoch": 1.3953949980150853,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004812364244586893,
      "loss": 5.3659,
      "mean_token_accuracy": 0.15562427937984466,
      "num_tokens": 30813389.0,
      "step": 14060
    },
    {
      "entropy": 5.422713279724121,
      "epoch": 1.395891226677253,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004812222567088082,
      "loss": 5.1126,
      "mean_token_accuracy": 0.16218218952417374,
      "num_tokens": 30823725.0,
      "step": 14065
    },
    {
      "entropy": 5.359988594055176,
      "epoch": 1.3963874553394204,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004812080838450596,
      "loss": 5.1474,
      "mean_token_accuracy": 0.1616002321243286,
      "num_tokens": 30834448.0,
      "step": 14070
    },
    {
      "entropy": 5.425362873077392,
      "epoch": 1.396883684001588,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004811939058677947,
      "loss": 5.1905,
      "mean_token_accuracy": 0.1605086088180542,
      "num_tokens": 30845891.0,
      "step": 14075
    },
    {
      "entropy": 5.396380281448364,
      "epoch": 1.3973799126637554,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004811797227773654,
      "loss": 5.1424,
      "mean_token_accuracy": 0.16636328548192977,
      "num_tokens": 30855170.0,
      "step": 14080
    },
    {
      "entropy": 5.377769899368286,
      "epoch": 1.397876141325923,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048116553457412317,
      "loss": 5.2183,
      "mean_token_accuracy": 0.1622815325856209,
      "num_tokens": 30866527.0,
      "step": 14085
    },
    {
      "entropy": 5.387840318679809,
      "epoch": 1.3983723699880906,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048115134125841993,
      "loss": 5.1833,
      "mean_token_accuracy": 0.16223891973495483,
      "num_tokens": 30876292.0,
      "step": 14090
    },
    {
      "entropy": 5.459110498428345,
      "epoch": 1.398868598650258,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004811371428306076,
      "loss": 5.1588,
      "mean_token_accuracy": 0.1635096177458763,
      "num_tokens": 30887440.0,
      "step": 14095
    },
    {
      "entropy": 5.421774291992188,
      "epoch": 1.3993648273124255,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004811229392910384,
      "loss": 5.1998,
      "mean_token_accuracy": 0.16290010511875153,
      "num_tokens": 30897925.0,
      "step": 14100
    },
    {
      "entropy": 5.380305004119873,
      "epoch": 1.399861055974593,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004811087306400644,
      "loss": 5.2365,
      "mean_token_accuracy": 0.16661897748708726,
      "num_tokens": 30908413.0,
      "step": 14105
    },
    {
      "entropy": 5.461961507797241,
      "epoch": 1.4003572846367607,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048109451687803804,
      "loss": 5.2509,
      "mean_token_accuracy": 0.15921526551246643,
      "num_tokens": 30919149.0,
      "step": 14110
    },
    {
      "entropy": 5.471443557739258,
      "epoch": 1.4008535132989282,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004810802980053117,
      "loss": 5.2813,
      "mean_token_accuracy": 0.15965394973754882,
      "num_tokens": 30929883.0,
      "step": 14115
    },
    {
      "entropy": 5.417294883728028,
      "epoch": 1.4013497419610956,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048106607402223815,
      "loss": 5.2034,
      "mean_token_accuracy": 0.16000442951917648,
      "num_tokens": 30941358.0,
      "step": 14120
    },
    {
      "entropy": 5.452350711822509,
      "epoch": 1.4018459706232633,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048105184492916995,
      "loss": 5.2011,
      "mean_token_accuracy": 0.1588554412126541,
      "num_tokens": 30953289.0,
      "step": 14125
    },
    {
      "entropy": 5.384692525863647,
      "epoch": 1.4023421992854308,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048103761072645993,
      "loss": 5.1327,
      "mean_token_accuracy": 0.16611771583557128,
      "num_tokens": 30964304.0,
      "step": 14130
    },
    {
      "entropy": 5.380402374267578,
      "epoch": 1.4028384279475983,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048102337141446116,
      "loss": 5.1585,
      "mean_token_accuracy": 0.16854921132326126,
      "num_tokens": 30974840.0,
      "step": 14135
    },
    {
      "entropy": 5.407678747177124,
      "epoch": 1.4033346566097658,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048100912699352666,
      "loss": 5.2336,
      "mean_token_accuracy": 0.16353936344385148,
      "num_tokens": 30985805.0,
      "step": 14140
    },
    {
      "entropy": 5.410429048538208,
      "epoch": 1.4038308852719332,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004809948774640098,
      "loss": 5.229,
      "mean_token_accuracy": 0.16078313142061235,
      "num_tokens": 30998174.0,
      "step": 14145
    },
    {
      "entropy": 5.482632160186768,
      "epoch": 1.4043271139341007,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004809806228262639,
      "loss": 5.2404,
      "mean_token_accuracy": 0.15747760385274887,
      "num_tokens": 31010050.0,
      "step": 14150
    },
    {
      "entropy": 5.590401029586792,
      "epoch": 1.4048233425962684,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048096636308064236,
      "loss": 5.3593,
      "mean_token_accuracy": 0.15138811767101287,
      "num_tokens": 31023069.0,
      "step": 14155
    },
    {
      "entropy": 5.4560894012451175,
      "epoch": 1.4053195712584359,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004809520982274989,
      "loss": 5.1559,
      "mean_token_accuracy": 0.16391359716653825,
      "num_tokens": 31033512.0,
      "step": 14160
    },
    {
      "entropy": 5.426138925552368,
      "epoch": 1.4058157999206033,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048093782826718706,
      "loss": 5.1994,
      "mean_token_accuracy": 0.15686285942792894,
      "num_tokens": 31045731.0,
      "step": 14165
    },
    {
      "entropy": 5.390913629531861,
      "epoch": 1.406312028582771,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048092355320006086,
      "loss": 5.1199,
      "mean_token_accuracy": 0.16922507882118226,
      "num_tokens": 31055443.0,
      "step": 14170
    },
    {
      "entropy": 5.379008197784424,
      "epoch": 1.4068082572449385,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004809092730264744,
      "loss": 5.2106,
      "mean_token_accuracy": 0.16574379652738572,
      "num_tokens": 31067575.0,
      "step": 14175
    },
    {
      "entropy": 5.433275604248047,
      "epoch": 1.407304485907106,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048089498774678155,
      "loss": 5.2944,
      "mean_token_accuracy": 0.1605135753750801,
      "num_tokens": 31078198.0,
      "step": 14180
    },
    {
      "entropy": 5.401867961883545,
      "epoch": 1.4078007145692735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048088069736133664,
      "loss": 5.1877,
      "mean_token_accuracy": 0.16151152998209,
      "num_tokens": 31089961.0,
      "step": 14185
    },
    {
      "entropy": 5.417603635787964,
      "epoch": 1.408296943231441,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004808664018704941,
      "loss": 5.2087,
      "mean_token_accuracy": 0.16039336621761321,
      "num_tokens": 31100534.0,
      "step": 14190
    },
    {
      "entropy": 5.528165102005005,
      "epoch": 1.4087931718936086,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004808521012746085,
      "loss": 5.237,
      "mean_token_accuracy": 0.16515662968158723,
      "num_tokens": 31112016.0,
      "step": 14195
    },
    {
      "entropy": 5.389935827255249,
      "epoch": 1.4092894005557761,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048083779557403424,
      "loss": 5.2682,
      "mean_token_accuracy": 0.1601576402783394,
      "num_tokens": 31123659.0,
      "step": 14200
    },
    {
      "entropy": 5.421218490600586,
      "epoch": 1.4097856292179436,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004808234847691262,
      "loss": 5.2007,
      "mean_token_accuracy": 0.15902014076709747,
      "num_tokens": 31133872.0,
      "step": 14205
    },
    {
      "entropy": 5.498111629486084,
      "epoch": 1.4102818578801113,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004808091688602393,
      "loss": 5.2562,
      "mean_token_accuracy": 0.1601032167673111,
      "num_tokens": 31145834.0,
      "step": 14210
    },
    {
      "entropy": 5.462192726135254,
      "epoch": 1.4107780865422788,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048079484784772854,
      "loss": 5.317,
      "mean_token_accuracy": 0.1569316565990448,
      "num_tokens": 31157037.0,
      "step": 14215
    },
    {
      "entropy": 5.403896617889404,
      "epoch": 1.4112743152044462,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048078052173194886,
      "loss": 5.2137,
      "mean_token_accuracy": 0.16394178420305253,
      "num_tokens": 31166812.0,
      "step": 14220
    },
    {
      "entropy": 5.422975969314575,
      "epoch": 1.4117705438666137,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004807661905132557,
      "loss": 5.1963,
      "mean_token_accuracy": 0.15653181895613671,
      "num_tokens": 31177162.0,
      "step": 14225
    },
    {
      "entropy": 5.449989175796508,
      "epoch": 1.4122667725287812,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004807518541920045,
      "loss": 5.2307,
      "mean_token_accuracy": 0.15988135933876038,
      "num_tokens": 31187333.0,
      "step": 14230
    },
    {
      "entropy": 5.4929811477661135,
      "epoch": 1.4127630011909487,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048073751276855056,
      "loss": 5.2267,
      "mean_token_accuracy": 0.15594532042741777,
      "num_tokens": 31197869.0,
      "step": 14235
    },
    {
      "entropy": 5.446146154403687,
      "epoch": 1.4132592298531164,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004807231662432497,
      "loss": 5.258,
      "mean_token_accuracy": 0.1561680555343628,
      "num_tokens": 31208199.0,
      "step": 14240
    },
    {
      "entropy": 5.419759225845337,
      "epoch": 1.4137554585152838,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004807088146164575,
      "loss": 5.2199,
      "mean_token_accuracy": 0.1594805061817169,
      "num_tokens": 31218903.0,
      "step": 14245
    },
    {
      "entropy": 5.441708087921143,
      "epoch": 1.4142516871774513,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004806944578885301,
      "loss": 5.2427,
      "mean_token_accuracy": 0.1554356411099434,
      "num_tokens": 31231146.0,
      "step": 14250
    },
    {
      "entropy": 5.4596914768219,
      "epoch": 1.414747915839619,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048068009605982333,
      "loss": 5.1507,
      "mean_token_accuracy": 0.16351498365402223,
      "num_tokens": 31240980.0,
      "step": 14255
    },
    {
      "entropy": 5.401647281646729,
      "epoch": 1.4152441445017865,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048066572913069334,
      "loss": 5.2003,
      "mean_token_accuracy": 0.15954401791095735,
      "num_tokens": 31252458.0,
      "step": 14260
    },
    {
      "entropy": 5.4116058349609375,
      "epoch": 1.415740373163954,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004806513571014965,
      "loss": 5.203,
      "mean_token_accuracy": 0.15940373837947847,
      "num_tokens": 31262979.0,
      "step": 14265
    },
    {
      "entropy": 5.427302980422974,
      "epoch": 1.4162366018261214,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004806369799725891,
      "loss": 5.1939,
      "mean_token_accuracy": 0.1598379999399185,
      "num_tokens": 31274811.0,
      "step": 14270
    },
    {
      "entropy": 5.39881386756897,
      "epoch": 1.416732830488289,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004806225977443277,
      "loss": 5.1404,
      "mean_token_accuracy": 0.16844291239976883,
      "num_tokens": 31285573.0,
      "step": 14275
    },
    {
      "entropy": 5.4774298667907715,
      "epoch": 1.4172290591504566,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004806082104170689,
      "loss": 5.3196,
      "mean_token_accuracy": 0.1539553016424179,
      "num_tokens": 31296993.0,
      "step": 14280
    },
    {
      "entropy": 5.428158617019653,
      "epoch": 1.417725287812624,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048059381799116967,
      "loss": 5.2109,
      "mean_token_accuracy": 0.16175659149885177,
      "num_tokens": 31307217.0,
      "step": 14285
    },
    {
      "entropy": 5.423924160003662,
      "epoch": 1.4182215164747916,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004805794204669867,
      "loss": 5.2336,
      "mean_token_accuracy": 0.15793718099594117,
      "num_tokens": 31317939.0,
      "step": 14290
    },
    {
      "entropy": 5.486760044097901,
      "epoch": 1.418717745136959,
      "grad_norm": 1.125,
      "learning_rate": 0.00048056501784487707,
      "loss": 5.2086,
      "mean_token_accuracy": 0.15825916975736617,
      "num_tokens": 31327731.0,
      "step": 14295
    },
    {
      "entropy": 5.5644560813903805,
      "epoch": 1.4192139737991267,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000480550610125198,
      "loss": 5.3881,
      "mean_token_accuracy": 0.14580497816205024,
      "num_tokens": 31339388.0,
      "step": 14300
    },
    {
      "entropy": 5.400137996673584,
      "epoch": 1.4197102024612942,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004805361973083066,
      "loss": 5.0856,
      "mean_token_accuracy": 0.16489316523075104,
      "num_tokens": 31350186.0,
      "step": 14305
    },
    {
      "entropy": 5.413147687911987,
      "epoch": 1.4202064311234617,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004805217793945605,
      "loss": 5.2383,
      "mean_token_accuracy": 0.15491159856319428,
      "num_tokens": 31361219.0,
      "step": 14310
    },
    {
      "entropy": 5.393682765960693,
      "epoch": 1.4207026597856292,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004805073563843171,
      "loss": 5.2603,
      "mean_token_accuracy": 0.16031864881515503,
      "num_tokens": 31372820.0,
      "step": 14315
    },
    {
      "entropy": 5.42013201713562,
      "epoch": 1.4211988884477966,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048049292827793413,
      "loss": 5.1449,
      "mean_token_accuracy": 0.16678092926740645,
      "num_tokens": 31382779.0,
      "step": 14320
    },
    {
      "entropy": 5.49399209022522,
      "epoch": 1.4216951171099643,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004804784950757693,
      "loss": 5.3746,
      "mean_token_accuracy": 0.15169049054384232,
      "num_tokens": 31394872.0,
      "step": 14325
    },
    {
      "entropy": 5.459486198425293,
      "epoch": 1.4221913457721318,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048046405677818057,
      "loss": 5.254,
      "mean_token_accuracy": 0.1603461965918541,
      "num_tokens": 31405767.0,
      "step": 14330
    },
    {
      "entropy": 5.5042613506317135,
      "epoch": 1.4226875744342993,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048044961338552595,
      "loss": 5.2537,
      "mean_token_accuracy": 0.15982651561498643,
      "num_tokens": 31417255.0,
      "step": 14335
    },
    {
      "entropy": 5.439549160003662,
      "epoch": 1.423183803096467,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004804351648981636,
      "loss": 5.2143,
      "mean_token_accuracy": 0.16539685875177385,
      "num_tokens": 31428494.0,
      "step": 14340
    },
    {
      "entropy": 5.398505735397339,
      "epoch": 1.4236800317586344,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004804207113164519,
      "loss": 5.1325,
      "mean_token_accuracy": 0.1651623860001564,
      "num_tokens": 31439221.0,
      "step": 14345
    },
    {
      "entropy": 5.436392593383789,
      "epoch": 1.424176260420802,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048040625264074915,
      "loss": 5.203,
      "mean_token_accuracy": 0.16182088255882263,
      "num_tokens": 31450575.0,
      "step": 14350
    },
    {
      "entropy": 5.307959461212159,
      "epoch": 1.4246724890829694,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048039178887141404,
      "loss": 5.2032,
      "mean_token_accuracy": 0.16718610972166062,
      "num_tokens": 31462168.0,
      "step": 14355
    },
    {
      "entropy": 5.440174531936646,
      "epoch": 1.4251687177451369,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048037732000880503,
      "loss": 5.2128,
      "mean_token_accuracy": 0.1636338025331497,
      "num_tokens": 31472619.0,
      "step": 14360
    },
    {
      "entropy": 5.474051141738892,
      "epoch": 1.4256649464073046,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048036284605328106,
      "loss": 5.3053,
      "mean_token_accuracy": 0.1597226768732071,
      "num_tokens": 31485246.0,
      "step": 14365
    },
    {
      "entropy": 5.44502067565918,
      "epoch": 1.426161175069472,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004803483670052011,
      "loss": 5.2097,
      "mean_token_accuracy": 0.15525606721639634,
      "num_tokens": 31496258.0,
      "step": 14370
    },
    {
      "entropy": 5.445867919921875,
      "epoch": 1.4266574037316395,
      "grad_norm": 1.078125,
      "learning_rate": 0.000480333882864924,
      "loss": 5.1916,
      "mean_token_accuracy": 0.1593363657593727,
      "num_tokens": 31506671.0,
      "step": 14375
    },
    {
      "entropy": 5.460860776901245,
      "epoch": 1.427153632393807,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004803193936328092,
      "loss": 5.2338,
      "mean_token_accuracy": 0.15348103940486907,
      "num_tokens": 31517001.0,
      "step": 14380
    },
    {
      "entropy": 5.423148965835571,
      "epoch": 1.4276498610559747,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004803048993092158,
      "loss": 5.21,
      "mean_token_accuracy": 0.1647414520382881,
      "num_tokens": 31529649.0,
      "step": 14385
    },
    {
      "entropy": 5.348645162582398,
      "epoch": 1.4281460897181422,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048029039989450317,
      "loss": 5.1629,
      "mean_token_accuracy": 0.16381987929344177,
      "num_tokens": 31540083.0,
      "step": 14390
    },
    {
      "entropy": 5.478333187103272,
      "epoch": 1.4286423183803096,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048027589538903115,
      "loss": 5.2611,
      "mean_token_accuracy": 0.1624650701880455,
      "num_tokens": 31551400.0,
      "step": 14395
    },
    {
      "entropy": 5.4759268283844,
      "epoch": 1.4291385470424771,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004802613857931592,
      "loss": 5.2588,
      "mean_token_accuracy": 0.15395909994840623,
      "num_tokens": 31562430.0,
      "step": 14400
    },
    {
      "entropy": 5.494983911514282,
      "epoch": 1.4296347757046446,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004802468711072472,
      "loss": 5.2899,
      "mean_token_accuracy": 0.16409255415201188,
      "num_tokens": 31573173.0,
      "step": 14405
    },
    {
      "entropy": 5.455746507644653,
      "epoch": 1.4301310043668123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000480232351331655,
      "loss": 5.2443,
      "mean_token_accuracy": 0.15737870782613755,
      "num_tokens": 31584072.0,
      "step": 14410
    },
    {
      "entropy": 5.413278245925904,
      "epoch": 1.4306272330289798,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048021782646674277,
      "loss": 5.1716,
      "mean_token_accuracy": 0.1632256418466568,
      "num_tokens": 31594993.0,
      "step": 14415
    },
    {
      "entropy": 5.389691162109375,
      "epoch": 1.4311234616911472,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048020329651287064,
      "loss": 5.1931,
      "mean_token_accuracy": 0.16735554337501526,
      "num_tokens": 31604284.0,
      "step": 14420
    },
    {
      "entropy": 5.400984859466552,
      "epoch": 1.431619690353315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000480188761470399,
      "loss": 5.2841,
      "mean_token_accuracy": 0.16176337897777557,
      "num_tokens": 31616188.0,
      "step": 14425
    },
    {
      "entropy": 5.5156656265258786,
      "epoch": 1.4321159190154824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004801742213396881,
      "loss": 5.2341,
      "mean_token_accuracy": 0.16108967065811158,
      "num_tokens": 31626606.0,
      "step": 14430
    },
    {
      "entropy": 5.394476890563965,
      "epoch": 1.4326121476776499,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048015967612109875,
      "loss": 5.1522,
      "mean_token_accuracy": 0.1705331861972809,
      "num_tokens": 31638050.0,
      "step": 14435
    },
    {
      "entropy": 5.405538606643677,
      "epoch": 1.4331083763398174,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048014512581499136,
      "loss": 5.1718,
      "mean_token_accuracy": 0.16590820848941804,
      "num_tokens": 31648451.0,
      "step": 14440
    },
    {
      "entropy": 5.410956621170044,
      "epoch": 1.4336046050019848,
      "grad_norm": 1.09375,
      "learning_rate": 0.000480130570421727,
      "loss": 5.201,
      "mean_token_accuracy": 0.15809377133846284,
      "num_tokens": 31659310.0,
      "step": 14445
    },
    {
      "entropy": 5.426698970794678,
      "epoch": 1.4341008336641523,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004801160099416665,
      "loss": 5.2432,
      "mean_token_accuracy": 0.15986751019954681,
      "num_tokens": 31669594.0,
      "step": 14450
    },
    {
      "entropy": 5.438290643692016,
      "epoch": 1.43459706232632,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004801014443751709,
      "loss": 5.2256,
      "mean_token_accuracy": 0.1589008703827858,
      "num_tokens": 31682034.0,
      "step": 14455
    },
    {
      "entropy": 5.486562776565552,
      "epoch": 1.4350932909884875,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004800868737226015,
      "loss": 5.2603,
      "mean_token_accuracy": 0.16005049645900726,
      "num_tokens": 31692390.0,
      "step": 14460
    },
    {
      "entropy": 5.444645500183105,
      "epoch": 1.435589519650655,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048007229798431946,
      "loss": 5.2248,
      "mean_token_accuracy": 0.16389539688825608,
      "num_tokens": 31703690.0,
      "step": 14465
    },
    {
      "entropy": 5.43157205581665,
      "epoch": 1.4360857483128227,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004800577171606864,
      "loss": 5.2507,
      "mean_token_accuracy": 0.1609525963664055,
      "num_tokens": 31714655.0,
      "step": 14470
    },
    {
      "entropy": 5.372096157073974,
      "epoch": 1.4365819769749901,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004800431312520638,
      "loss": 5.151,
      "mean_token_accuracy": 0.16648461371660234,
      "num_tokens": 31727022.0,
      "step": 14475
    },
    {
      "entropy": 5.423698139190674,
      "epoch": 1.4370782056371576,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004800285402588134,
      "loss": 5.1676,
      "mean_token_accuracy": 0.166916623711586,
      "num_tokens": 31738071.0,
      "step": 14480
    },
    {
      "entropy": 5.480141687393188,
      "epoch": 1.437574434299325,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004800139441812969,
      "loss": 5.3092,
      "mean_token_accuracy": 0.1535298451781273,
      "num_tokens": 31749342.0,
      "step": 14485
    },
    {
      "entropy": 5.447306823730469,
      "epoch": 1.4380706629614926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047999934301987644,
      "loss": 5.2698,
      "mean_token_accuracy": 0.15799569934606553,
      "num_tokens": 31761218.0,
      "step": 14490
    },
    {
      "entropy": 5.462802505493164,
      "epoch": 1.4385668916236602,
      "grad_norm": 1.0625,
      "learning_rate": 0.000479984736774914,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1580069348216057,
      "num_tokens": 31773111.0,
      "step": 14495
    },
    {
      "entropy": 5.471288537979126,
      "epoch": 1.4390631202858277,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047997012544677176,
      "loss": 5.2329,
      "mean_token_accuracy": 0.16104327291250228,
      "num_tokens": 31782948.0,
      "step": 14500
    },
    {
      "entropy": 5.4544731140136715,
      "epoch": 1.4395593489479952,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004799555090358121,
      "loss": 5.2311,
      "mean_token_accuracy": 0.16058211773633957,
      "num_tokens": 31793992.0,
      "step": 14505
    },
    {
      "entropy": 5.494948959350586,
      "epoch": 1.440055577610163,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004799408875423975,
      "loss": 5.2428,
      "mean_token_accuracy": 0.15467620342969896,
      "num_tokens": 31804830.0,
      "step": 14510
    },
    {
      "entropy": 5.388610935211181,
      "epoch": 1.4405518062723304,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004799262609668904,
      "loss": 5.2246,
      "mean_token_accuracy": 0.16143872141838073,
      "num_tokens": 31817200.0,
      "step": 14515
    },
    {
      "entropy": 5.427373313903809,
      "epoch": 1.4410480349344978,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004799116293096537,
      "loss": 5.2827,
      "mean_token_accuracy": 0.16483641266822815,
      "num_tokens": 31828226.0,
      "step": 14520
    },
    {
      "entropy": 5.438805294036865,
      "epoch": 1.4415442635966653,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004798969925710501,
      "loss": 5.1496,
      "mean_token_accuracy": 0.17313910722732545,
      "num_tokens": 31839635.0,
      "step": 14525
    },
    {
      "entropy": 5.431429672241211,
      "epoch": 1.4420404922588328,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047988235075144254,
      "loss": 5.1859,
      "mean_token_accuracy": 0.16815268844366074,
      "num_tokens": 31850333.0,
      "step": 14530
    },
    {
      "entropy": 5.428141832351685,
      "epoch": 1.4425367209210003,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004798677038511942,
      "loss": 5.1802,
      "mean_token_accuracy": 0.15753020495176315,
      "num_tokens": 31861002.0,
      "step": 14535
    },
    {
      "entropy": 5.449025249481201,
      "epoch": 1.443032949583168,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004798530518706683,
      "loss": 5.252,
      "mean_token_accuracy": 0.15666510313749313,
      "num_tokens": 31871713.0,
      "step": 14540
    },
    {
      "entropy": 5.4455958843231205,
      "epoch": 1.4435291782453354,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004798383948102281,
      "loss": 5.1753,
      "mean_token_accuracy": 0.163969948887825,
      "num_tokens": 31882381.0,
      "step": 14545
    },
    {
      "entropy": 5.494309616088867,
      "epoch": 1.444025406907503,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004798237326702371,
      "loss": 5.2345,
      "mean_token_accuracy": 0.1626094773411751,
      "num_tokens": 31892347.0,
      "step": 14550
    },
    {
      "entropy": 5.517892169952392,
      "epoch": 1.4445216355696706,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047980906545105884,
      "loss": 5.3282,
      "mean_token_accuracy": 0.15321285277605057,
      "num_tokens": 31902751.0,
      "step": 14555
    },
    {
      "entropy": 5.381084060668945,
      "epoch": 1.445017864231838,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004797943931530571,
      "loss": 5.1776,
      "mean_token_accuracy": 0.1645873248577118,
      "num_tokens": 31913294.0,
      "step": 14560
    },
    {
      "entropy": 5.3995965957641605,
      "epoch": 1.4455140928940056,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004797797157765958,
      "loss": 5.3064,
      "mean_token_accuracy": 0.1633098304271698,
      "num_tokens": 31925678.0,
      "step": 14565
    },
    {
      "entropy": 5.435469484329223,
      "epoch": 1.446010321556173,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004797650333220387,
      "loss": 5.1126,
      "mean_token_accuracy": 0.16574485450983048,
      "num_tokens": 31934578.0,
      "step": 14570
    },
    {
      "entropy": 5.394409704208374,
      "epoch": 1.4465065502183405,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047975034578975006,
      "loss": 5.1856,
      "mean_token_accuracy": 0.15886692106723785,
      "num_tokens": 31945482.0,
      "step": 14575
    },
    {
      "entropy": 5.44382529258728,
      "epoch": 1.4470027788805082,
      "grad_norm": 1.125,
      "learning_rate": 0.000479735653180094,
      "loss": 5.3007,
      "mean_token_accuracy": 0.16387143433094026,
      "num_tokens": 31957303.0,
      "step": 14580
    },
    {
      "entropy": 5.46829137802124,
      "epoch": 1.4474990075426757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004797209554934349,
      "loss": 5.2096,
      "mean_token_accuracy": 0.16066322326660157,
      "num_tokens": 31968601.0,
      "step": 14585
    },
    {
      "entropy": 5.433897304534912,
      "epoch": 1.4479952362048432,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004797062527301373,
      "loss": 5.2455,
      "mean_token_accuracy": 0.15600375533103944,
      "num_tokens": 31978987.0,
      "step": 14590
    },
    {
      "entropy": 5.408560180664063,
      "epoch": 1.4484914648670106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004796915448905657,
      "loss": 5.2113,
      "mean_token_accuracy": 0.1619571715593338,
      "num_tokens": 31990950.0,
      "step": 14595
    },
    {
      "entropy": 5.49525408744812,
      "epoch": 1.4489876935291783,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047967683197508486,
      "loss": 5.2414,
      "mean_token_accuracy": 0.15757601112127304,
      "num_tokens": 32001795.0,
      "step": 14600
    },
    {
      "entropy": 5.434155416488648,
      "epoch": 1.4494839221913458,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047966211398405966,
      "loss": 5.2558,
      "mean_token_accuracy": 0.15163022875785828,
      "num_tokens": 32011775.0,
      "step": 14605
    },
    {
      "entropy": 5.552326440811157,
      "epoch": 1.4499801508535133,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000479647390917855,
      "loss": 5.3423,
      "mean_token_accuracy": 0.1472148358821869,
      "num_tokens": 32022674.0,
      "step": 14610
    },
    {
      "entropy": 5.508969974517822,
      "epoch": 1.4504763795156808,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000479632662776836,
      "loss": 5.2855,
      "mean_token_accuracy": 0.16320230662822724,
      "num_tokens": 32033383.0,
      "step": 14615
    },
    {
      "entropy": 5.438214588165283,
      "epoch": 1.4509726081778482,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000479617929561368,
      "loss": 5.187,
      "mean_token_accuracy": 0.1668942466378212,
      "num_tokens": 32044066.0,
      "step": 14620
    },
    {
      "entropy": 5.419362115859985,
      "epoch": 1.451468836840016,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004796031912718162,
      "loss": 5.2687,
      "mean_token_accuracy": 0.15063661485910415,
      "num_tokens": 32054650.0,
      "step": 14625
    },
    {
      "entropy": 5.529458665847779,
      "epoch": 1.4519650655021834,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047958844790854606,
      "loss": 5.2638,
      "mean_token_accuracy": 0.15852388143539428,
      "num_tokens": 32065786.0,
      "step": 14630
    },
    {
      "entropy": 5.504126119613647,
      "epoch": 1.4524612941643509,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047957369947192337,
      "loss": 5.2161,
      "mean_token_accuracy": 0.1624716490507126,
      "num_tokens": 32076326.0,
      "step": 14635
    },
    {
      "entropy": 5.396923494338989,
      "epoch": 1.4529575228265186,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004795589459623136,
      "loss": 5.1661,
      "mean_token_accuracy": 0.16039752811193467,
      "num_tokens": 32086915.0,
      "step": 14640
    },
    {
      "entropy": 5.462643003463745,
      "epoch": 1.453453751488686,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047954418738008295,
      "loss": 5.1374,
      "mean_token_accuracy": 0.1657378852367401,
      "num_tokens": 32096614.0,
      "step": 14645
    },
    {
      "entropy": 5.455061340332032,
      "epoch": 1.4539499801508535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004795294237255971,
      "loss": 5.2257,
      "mean_token_accuracy": 0.16869989037513733,
      "num_tokens": 32107721.0,
      "step": 14650
    },
    {
      "entropy": 5.406118488311767,
      "epoch": 1.454446208813021,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047951465499922234,
      "loss": 5.2296,
      "mean_token_accuracy": 0.1571524113416672,
      "num_tokens": 32116621.0,
      "step": 14655
    },
    {
      "entropy": 5.51029691696167,
      "epoch": 1.4549424374751885,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004794998812013248,
      "loss": 5.2628,
      "mean_token_accuracy": 0.15729925781488419,
      "num_tokens": 32126536.0,
      "step": 14660
    },
    {
      "entropy": 5.523024272918701,
      "epoch": 1.4554386661373562,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047948510233227084,
      "loss": 5.3141,
      "mean_token_accuracy": 0.15293849855661393,
      "num_tokens": 32137059.0,
      "step": 14665
    },
    {
      "entropy": 5.4094470024108885,
      "epoch": 1.4559348947995236,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004794703183924269,
      "loss": 5.2532,
      "mean_token_accuracy": 0.16183771044015885,
      "num_tokens": 32148186.0,
      "step": 14670
    },
    {
      "entropy": 5.503705930709839,
      "epoch": 1.4564311234616911,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004794555293821597,
      "loss": 5.2776,
      "mean_token_accuracy": 0.15234410315752028,
      "num_tokens": 32158244.0,
      "step": 14675
    },
    {
      "entropy": 5.421056032180786,
      "epoch": 1.4569273521238586,
      "grad_norm": 1.09375,
      "learning_rate": 0.000479440735301836,
      "loss": 5.1811,
      "mean_token_accuracy": 0.1667837083339691,
      "num_tokens": 32168404.0,
      "step": 14680
    },
    {
      "entropy": 5.408515739440918,
      "epoch": 1.4574235807860263,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004794259361518226,
      "loss": 5.1428,
      "mean_token_accuracy": 0.16685988008975983,
      "num_tokens": 32178744.0,
      "step": 14685
    },
    {
      "entropy": 5.441531372070313,
      "epoch": 1.4579198094481938,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047941113193248643,
      "loss": 5.2797,
      "mean_token_accuracy": 0.15279127955436705,
      "num_tokens": 32189948.0,
      "step": 14690
    },
    {
      "entropy": 5.452954816818237,
      "epoch": 1.4584160381103612,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004793963226441947,
      "loss": 5.2606,
      "mean_token_accuracy": 0.1541539803147316,
      "num_tokens": 32200979.0,
      "step": 14695
    },
    {
      "entropy": 5.394747591018676,
      "epoch": 1.4589122667725287,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047938150828731455,
      "loss": 5.2334,
      "mean_token_accuracy": 0.16226961016654967,
      "num_tokens": 32211569.0,
      "step": 14700
    },
    {
      "entropy": 5.435095930099488,
      "epoch": 1.4594084954346962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004793666888622134,
      "loss": 5.2044,
      "mean_token_accuracy": 0.1652999773621559,
      "num_tokens": 32222429.0,
      "step": 14705
    },
    {
      "entropy": 5.476523208618164,
      "epoch": 1.459904724096864,
      "grad_norm": 1.125,
      "learning_rate": 0.0004793518643692587,
      "loss": 5.2982,
      "mean_token_accuracy": 0.15165069550275803,
      "num_tokens": 32234332.0,
      "step": 14710
    },
    {
      "entropy": 5.546288585662841,
      "epoch": 1.4604009527590314,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004793370348088182,
      "loss": 5.326,
      "mean_token_accuracy": 0.1638160988688469,
      "num_tokens": 32244798.0,
      "step": 14715
    },
    {
      "entropy": 5.470380163192749,
      "epoch": 1.4608971814211988,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004793222001812595,
      "loss": 5.2223,
      "mean_token_accuracy": 0.16118253469467164,
      "num_tokens": 32257746.0,
      "step": 14720
    },
    {
      "entropy": 5.47119174003601,
      "epoch": 1.4613934100833665,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047930736048695045,
      "loss": 5.3228,
      "mean_token_accuracy": 0.15251221209764482,
      "num_tokens": 32270692.0,
      "step": 14725
    },
    {
      "entropy": 5.436778116226196,
      "epoch": 1.461889638745534,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004792925157262592,
      "loss": 5.2155,
      "mean_token_accuracy": 0.16005224511027336,
      "num_tokens": 32281350.0,
      "step": 14730
    },
    {
      "entropy": 5.441813373565674,
      "epoch": 1.4623858674077015,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004792776658995537,
      "loss": 5.2502,
      "mean_token_accuracy": 0.15605368018150328,
      "num_tokens": 32293182.0,
      "step": 14735
    },
    {
      "entropy": 5.425389957427979,
      "epoch": 1.462882096069869,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004792628110072023,
      "loss": 5.2207,
      "mean_token_accuracy": 0.16067558377981186,
      "num_tokens": 32304903.0,
      "step": 14740
    },
    {
      "entropy": 5.553899192810059,
      "epoch": 1.4633783247320364,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004792479510495734,
      "loss": 5.389,
      "mean_token_accuracy": 0.16006532311439514,
      "num_tokens": 32316074.0,
      "step": 14745
    },
    {
      "entropy": 5.447405195236206,
      "epoch": 1.463874553394204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047923308602703535,
      "loss": 5.2074,
      "mean_token_accuracy": 0.16557188481092452,
      "num_tokens": 32327594.0,
      "step": 14750
    },
    {
      "entropy": 5.411319255828857,
      "epoch": 1.4643707820563716,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047921821593995684,
      "loss": 5.2329,
      "mean_token_accuracy": 0.16105389446020127,
      "num_tokens": 32339224.0,
      "step": 14755
    },
    {
      "entropy": 5.4453453540802,
      "epoch": 1.464867010718539,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004792033407887066,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1633899465203285,
      "num_tokens": 32349451.0,
      "step": 14760
    },
    {
      "entropy": 5.514544820785522,
      "epoch": 1.4653632393807066,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004791884605736536,
      "loss": 5.2961,
      "mean_token_accuracy": 0.15245563983917237,
      "num_tokens": 32361177.0,
      "step": 14765
    },
    {
      "entropy": 5.439993667602539,
      "epoch": 1.4658594680428743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047917357529516676,
      "loss": 5.2022,
      "mean_token_accuracy": 0.15349031388759612,
      "num_tokens": 32372031.0,
      "step": 14770
    },
    {
      "entropy": 5.423114776611328,
      "epoch": 1.4663556967050417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004791586849536152,
      "loss": 5.1365,
      "mean_token_accuracy": 0.16914646029472352,
      "num_tokens": 32383936.0,
      "step": 14775
    },
    {
      "entropy": 5.472107791900635,
      "epoch": 1.4668519253672092,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004791437895493682,
      "loss": 5.2841,
      "mean_token_accuracy": 0.16094436198472978,
      "num_tokens": 32394283.0,
      "step": 14780
    },
    {
      "entropy": 5.460607624053955,
      "epoch": 1.4673481540293767,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047912888908279507,
      "loss": 5.2762,
      "mean_token_accuracy": 0.15670128613710405,
      "num_tokens": 32405549.0,
      "step": 14785
    },
    {
      "entropy": 5.418529033660889,
      "epoch": 1.4678443826915442,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004791139835542653,
      "loss": 5.1914,
      "mean_token_accuracy": 0.1605245590209961,
      "num_tokens": 32414896.0,
      "step": 14790
    },
    {
      "entropy": 5.458948040008545,
      "epoch": 1.4683406113537119,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047909907296414864,
      "loss": 5.2797,
      "mean_token_accuracy": 0.16252537518739701,
      "num_tokens": 32426863.0,
      "step": 14795
    },
    {
      "entropy": 5.425500535964966,
      "epoch": 1.4688368400158793,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047908415731281475,
      "loss": 5.2248,
      "mean_token_accuracy": 0.16299201548099518,
      "num_tokens": 32437703.0,
      "step": 14800
    },
    {
      "entropy": 5.4683380126953125,
      "epoch": 1.4693330686780468,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047906923660063344,
      "loss": 5.273,
      "mean_token_accuracy": 0.16166049093008042,
      "num_tokens": 32448517.0,
      "step": 14805
    },
    {
      "entropy": 5.44162425994873,
      "epoch": 1.4698292973402145,
      "grad_norm": 1.140625,
      "learning_rate": 0.000479054310827975,
      "loss": 5.248,
      "mean_token_accuracy": 0.1625928148627281,
      "num_tokens": 32458782.0,
      "step": 14810
    },
    {
      "entropy": 5.422355508804321,
      "epoch": 1.470325526002382,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047903937999520913,
      "loss": 5.1255,
      "mean_token_accuracy": 0.1641959235072136,
      "num_tokens": 32470113.0,
      "step": 14815
    },
    {
      "entropy": 5.452918910980225,
      "epoch": 1.4708217546645495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004790244441027064,
      "loss": 5.2035,
      "mean_token_accuracy": 0.16456764489412307,
      "num_tokens": 32482455.0,
      "step": 14820
    },
    {
      "entropy": 5.478126764297485,
      "epoch": 1.471317983326717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004790095031508371,
      "loss": 5.2529,
      "mean_token_accuracy": 0.16193103939294815,
      "num_tokens": 32493536.0,
      "step": 14825
    },
    {
      "entropy": 5.398959541320801,
      "epoch": 1.4718142119888844,
      "grad_norm": 1.0,
      "learning_rate": 0.00047899455713997164,
      "loss": 5.1715,
      "mean_token_accuracy": 0.16413404494524003,
      "num_tokens": 32503991.0,
      "step": 14830
    },
    {
      "entropy": 5.414902830123902,
      "epoch": 1.4723104406510519,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047897960607048084,
      "loss": 5.2431,
      "mean_token_accuracy": 0.1585626557469368,
      "num_tokens": 32515474.0,
      "step": 14835
    },
    {
      "entropy": 5.417044258117675,
      "epoch": 1.4728066693132196,
      "grad_norm": 1.0,
      "learning_rate": 0.0004789646499427353,
      "loss": 5.1598,
      "mean_token_accuracy": 0.164191073179245,
      "num_tokens": 32526751.0,
      "step": 14840
    },
    {
      "entropy": 5.48064866065979,
      "epoch": 1.473302897975387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004789496887571059,
      "loss": 5.3292,
      "mean_token_accuracy": 0.15698872804641723,
      "num_tokens": 32538191.0,
      "step": 14845
    },
    {
      "entropy": 5.460055875778198,
      "epoch": 1.4737991266375545,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047893472251396373,
      "loss": 5.2577,
      "mean_token_accuracy": 0.16326451748609544,
      "num_tokens": 32548841.0,
      "step": 14850
    },
    {
      "entropy": 5.435880184173584,
      "epoch": 1.4742953552997222,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047891975121368,
      "loss": 5.1777,
      "mean_token_accuracy": 0.1593521997332573,
      "num_tokens": 32558599.0,
      "step": 14855
    },
    {
      "entropy": 5.431378316879273,
      "epoch": 1.4747915839618897,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047890477485662563,
      "loss": 5.1555,
      "mean_token_accuracy": 0.16730142533779144,
      "num_tokens": 32568641.0,
      "step": 14860
    },
    {
      "entropy": 5.435833692550659,
      "epoch": 1.4752878126240572,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047888979344317236,
      "loss": 5.25,
      "mean_token_accuracy": 0.15751945823431016,
      "num_tokens": 32578924.0,
      "step": 14865
    },
    {
      "entropy": 5.469187068939209,
      "epoch": 1.4757840412862246,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047887480697369155,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1674180120229721,
      "num_tokens": 32590233.0,
      "step": 14870
    },
    {
      "entropy": 5.458964204788208,
      "epoch": 1.4762802699483921,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047885981544855475,
      "loss": 5.2448,
      "mean_token_accuracy": 0.15713003575801848,
      "num_tokens": 32600782.0,
      "step": 14875
    },
    {
      "entropy": 5.390079307556152,
      "epoch": 1.4767764986105598,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004788448188681339,
      "loss": 5.2123,
      "mean_token_accuracy": 0.16712939590215684,
      "num_tokens": 32612611.0,
      "step": 14880
    },
    {
      "entropy": 5.379263830184937,
      "epoch": 1.4772727272727273,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004788298172328007,
      "loss": 5.3061,
      "mean_token_accuracy": 0.16256429702043534,
      "num_tokens": 32623436.0,
      "step": 14885
    },
    {
      "entropy": 5.410916996002197,
      "epoch": 1.4777689559348948,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047881481054292736,
      "loss": 5.1608,
      "mean_token_accuracy": 0.16267303228378296,
      "num_tokens": 32633356.0,
      "step": 14890
    },
    {
      "entropy": 5.437669134140014,
      "epoch": 1.4782651845970625,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047879979879888584,
      "loss": 5.2127,
      "mean_token_accuracy": 0.1668834090232849,
      "num_tokens": 32643927.0,
      "step": 14895
    },
    {
      "entropy": 5.38257474899292,
      "epoch": 1.47876141325923,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047878478200104845,
      "loss": 5.1947,
      "mean_token_accuracy": 0.16661451160907745,
      "num_tokens": 32655086.0,
      "step": 14900
    },
    {
      "entropy": 5.4189940929412845,
      "epoch": 1.4792576419213974,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004787697601497876,
      "loss": 5.1376,
      "mean_token_accuracy": 0.16446703672409058,
      "num_tokens": 32665298.0,
      "step": 14905
    },
    {
      "entropy": 5.422038936614991,
      "epoch": 1.479753870583565,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004787547332454758,
      "loss": 5.1788,
      "mean_token_accuracy": 0.15813643783330916,
      "num_tokens": 32676717.0,
      "step": 14910
    },
    {
      "entropy": 5.466410303115845,
      "epoch": 1.4802500992457324,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047873970128848567,
      "loss": 5.245,
      "mean_token_accuracy": 0.15536442250013352,
      "num_tokens": 32686886.0,
      "step": 14915
    },
    {
      "entropy": 5.418141889572143,
      "epoch": 1.4807463279078998,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004787246642791899,
      "loss": 5.1733,
      "mean_token_accuracy": 0.1685820698738098,
      "num_tokens": 32698631.0,
      "step": 14920
    },
    {
      "entropy": 5.418970155715942,
      "epoch": 1.4812425565700675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047870962221796147,
      "loss": 5.1983,
      "mean_token_accuracy": 0.15671693235635759,
      "num_tokens": 32710672.0,
      "step": 14925
    },
    {
      "entropy": 5.439627742767334,
      "epoch": 1.481738785232235,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004786945751051734,
      "loss": 5.2432,
      "mean_token_accuracy": 0.16453288197517396,
      "num_tokens": 32720858.0,
      "step": 14930
    },
    {
      "entropy": 5.423089027404785,
      "epoch": 1.4822350138944025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004786795229411988,
      "loss": 5.1773,
      "mean_token_accuracy": 0.16653541028499602,
      "num_tokens": 32731592.0,
      "step": 14935
    },
    {
      "entropy": 5.477849769592285,
      "epoch": 1.4827312425565702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047866446572641084,
      "loss": 5.2379,
      "mean_token_accuracy": 0.157359379529953,
      "num_tokens": 32742777.0,
      "step": 14940
    },
    {
      "entropy": 5.422476720809937,
      "epoch": 1.4832274712187377,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004786494034611831,
      "loss": 5.1734,
      "mean_token_accuracy": 0.1648246705532074,
      "num_tokens": 32753035.0,
      "step": 14945
    },
    {
      "entropy": 5.432260894775391,
      "epoch": 1.4837236998809051,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000478634336145889,
      "loss": 5.2028,
      "mean_token_accuracy": 0.158352592587471,
      "num_tokens": 32764318.0,
      "step": 14950
    },
    {
      "entropy": 5.421120500564575,
      "epoch": 1.4842199285430726,
      "grad_norm": 1.25,
      "learning_rate": 0.00047861926378090197,
      "loss": 5.2196,
      "mean_token_accuracy": 0.15650428235530853,
      "num_tokens": 32775521.0,
      "step": 14955
    },
    {
      "entropy": 5.465495252609253,
      "epoch": 1.48471615720524,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004786041863665961,
      "loss": 5.2404,
      "mean_token_accuracy": 0.16316316574811934,
      "num_tokens": 32787101.0,
      "step": 14960
    },
    {
      "entropy": 5.4019969463348385,
      "epoch": 1.4852123858674078,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004785891039033451,
      "loss": 5.2359,
      "mean_token_accuracy": 0.16132086515426636,
      "num_tokens": 32798446.0,
      "step": 14965
    },
    {
      "entropy": 5.433573770523071,
      "epoch": 1.4857086145295753,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000478574016391523,
      "loss": 5.2463,
      "mean_token_accuracy": 0.1536371350288391,
      "num_tokens": 32809636.0,
      "step": 14970
    },
    {
      "entropy": 5.416846752166748,
      "epoch": 1.4862048431917427,
      "grad_norm": 1.0,
      "learning_rate": 0.000478558923831504,
      "loss": 5.1405,
      "mean_token_accuracy": 0.1603006348013878,
      "num_tokens": 32820132.0,
      "step": 14975
    },
    {
      "entropy": 5.452705240249633,
      "epoch": 1.4867010718539102,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047854382622366235,
      "loss": 5.2388,
      "mean_token_accuracy": 0.1649303175508976,
      "num_tokens": 32830720.0,
      "step": 14980
    },
    {
      "entropy": 5.376061820983887,
      "epoch": 1.487197300516078,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047852872356837236,
      "loss": 5.1659,
      "mean_token_accuracy": 0.16734691262245177,
      "num_tokens": 32841621.0,
      "step": 14985
    },
    {
      "entropy": 5.464379405975341,
      "epoch": 1.4876935291782454,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004785136158660085,
      "loss": 5.312,
      "mean_token_accuracy": 0.1534930095076561,
      "num_tokens": 32852074.0,
      "step": 14990
    },
    {
      "entropy": 5.506190490722656,
      "epoch": 1.4881897578404129,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047849850311694567,
      "loss": 5.2757,
      "mean_token_accuracy": 0.16292567923665047,
      "num_tokens": 32862556.0,
      "step": 14995
    },
    {
      "entropy": 5.445034122467041,
      "epoch": 1.4886859865025803,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047848338532155833,
      "loss": 5.1892,
      "mean_token_accuracy": 0.16786503195762634,
      "num_tokens": 32874407.0,
      "step": 15000
    },
    {
      "epoch": 1.4886859865025803,
      "eval_entropy": 5.303544319222619,
      "eval_loss": 5.357519626617432,
      "eval_mean_token_accuracy": 0.16005599613897334,
      "eval_num_tokens": 32874407.0,
      "eval_runtime": 26.7258,
      "eval_samples_per_second": 1259.795,
      "eval_steps_per_second": 157.488,
      "step": 15000
    },
    {
      "entropy": 5.446721696853638,
      "epoch": 1.4891822151647478,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047846826248022155,
      "loss": 5.2442,
      "mean_token_accuracy": 0.15665974617004394,
      "num_tokens": 32884986.0,
      "step": 15005
    },
    {
      "entropy": 5.455188131332397,
      "epoch": 1.4896784438269155,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047845313459331034,
      "loss": 5.2705,
      "mean_token_accuracy": 0.15842320770025253,
      "num_tokens": 32896016.0,
      "step": 15010
    },
    {
      "entropy": 5.383634233474732,
      "epoch": 1.490174672489083,
      "grad_norm": 1.125,
      "learning_rate": 0.00047843800166119974,
      "loss": 5.1893,
      "mean_token_accuracy": 0.1659022569656372,
      "num_tokens": 32905648.0,
      "step": 15015
    },
    {
      "entropy": 5.458200263977051,
      "epoch": 1.4906709011512504,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000478422863684265,
      "loss": 5.2475,
      "mean_token_accuracy": 0.16063673049211502,
      "num_tokens": 32917103.0,
      "step": 15020
    },
    {
      "entropy": 5.486766767501831,
      "epoch": 1.4911671298134181,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047840772066288167,
      "loss": 5.1977,
      "mean_token_accuracy": 0.15895269066095352,
      "num_tokens": 32928741.0,
      "step": 15025
    },
    {
      "entropy": 5.405528545379639,
      "epoch": 1.4916633584755856,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004783925725974252,
      "loss": 5.2448,
      "mean_token_accuracy": 0.15696649998426437,
      "num_tokens": 32939292.0,
      "step": 15030
    },
    {
      "entropy": 5.473295879364014,
      "epoch": 1.492159587137753,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004783774194882712,
      "loss": 5.236,
      "mean_token_accuracy": 0.16180628687143325,
      "num_tokens": 32949940.0,
      "step": 15035
    },
    {
      "entropy": 5.415271186828614,
      "epoch": 1.4926558157999206,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004783622613357954,
      "loss": 5.1976,
      "mean_token_accuracy": 0.16416143774986267,
      "num_tokens": 32961222.0,
      "step": 15040
    },
    {
      "entropy": 5.3601900100708,
      "epoch": 1.493152044462088,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004783470981403737,
      "loss": 5.1913,
      "mean_token_accuracy": 0.16611864417791367,
      "num_tokens": 32972095.0,
      "step": 15045
    },
    {
      "entropy": 5.454072618484497,
      "epoch": 1.4936482731242555,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004783319299023821,
      "loss": 5.2322,
      "mean_token_accuracy": 0.16713867336511612,
      "num_tokens": 32981834.0,
      "step": 15050
    },
    {
      "entropy": 5.5320725440979,
      "epoch": 1.4941445017864232,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004783167566221969,
      "loss": 5.3122,
      "mean_token_accuracy": 0.15757208913564683,
      "num_tokens": 32994347.0,
      "step": 15055
    },
    {
      "entropy": 5.517721891403198,
      "epoch": 1.4946407304485907,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047830157830019423,
      "loss": 5.2767,
      "mean_token_accuracy": 0.1630987137556076,
      "num_tokens": 33004780.0,
      "step": 15060
    },
    {
      "entropy": 5.477479887008667,
      "epoch": 1.4951369591107582,
      "grad_norm": 1.125,
      "learning_rate": 0.0004782863949367505,
      "loss": 5.2402,
      "mean_token_accuracy": 0.15652129501104356,
      "num_tokens": 33015697.0,
      "step": 15065
    },
    {
      "entropy": 5.344119739532471,
      "epoch": 1.4956331877729259,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004782712065322421,
      "loss": 5.2227,
      "mean_token_accuracy": 0.1545237645506859,
      "num_tokens": 33026563.0,
      "step": 15070
    },
    {
      "entropy": 5.464089727401733,
      "epoch": 1.4961294164350933,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004782560130870459,
      "loss": 5.2129,
      "mean_token_accuracy": 0.15596314072608947,
      "num_tokens": 33037290.0,
      "step": 15075
    },
    {
      "entropy": 5.4877630233764645,
      "epoch": 1.4966256450972608,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047824081460153846,
      "loss": 5.1624,
      "mean_token_accuracy": 0.1593530535697937,
      "num_tokens": 33047199.0,
      "step": 15080
    },
    {
      "entropy": 5.379214286804199,
      "epoch": 1.4971218737594283,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004782256110760969,
      "loss": 5.1699,
      "mean_token_accuracy": 0.16123603731393815,
      "num_tokens": 33057253.0,
      "step": 15085
    },
    {
      "entropy": 5.479037666320801,
      "epoch": 1.4976181024215958,
      "grad_norm": 1.140625,
      "learning_rate": 0.000478210402511098,
      "loss": 5.3076,
      "mean_token_accuracy": 0.15774478763341904,
      "num_tokens": 33068228.0,
      "step": 15090
    },
    {
      "entropy": 5.519403266906738,
      "epoch": 1.4981143310837635,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004781951889069189,
      "loss": 5.2956,
      "mean_token_accuracy": 0.15769878923892974,
      "num_tokens": 33079285.0,
      "step": 15095
    },
    {
      "entropy": 5.3933428764343265,
      "epoch": 1.498610559745931,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047817997026393714,
      "loss": 5.1447,
      "mean_token_accuracy": 0.16598886102437974,
      "num_tokens": 33091253.0,
      "step": 15100
    },
    {
      "entropy": 5.348390054702759,
      "epoch": 1.4991067884080984,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047816474658252985,
      "loss": 5.1685,
      "mean_token_accuracy": 0.16121202260255812,
      "num_tokens": 33102492.0,
      "step": 15105
    },
    {
      "entropy": 5.374969530105591,
      "epoch": 1.499603017070266,
      "grad_norm": 1.125,
      "learning_rate": 0.00047814951786307467,
      "loss": 5.1877,
      "mean_token_accuracy": 0.16293865442276,
      "num_tokens": 33115038.0,
      "step": 15110
    },
    {
      "entropy": 5.43083233833313,
      "epoch": 1.5000992457324336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004781342841059492,
      "loss": 5.1627,
      "mean_token_accuracy": 0.16533697694540023,
      "num_tokens": 33124470.0,
      "step": 15115
    },
    {
      "entropy": 5.396132516860962,
      "epoch": 1.500595474394601,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004781190453115311,
      "loss": 5.1999,
      "mean_token_accuracy": 0.16049351990222932,
      "num_tokens": 33135994.0,
      "step": 15120
    },
    {
      "entropy": 5.43891716003418,
      "epoch": 1.5010917030567685,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047810380148019847,
      "loss": 5.2032,
      "mean_token_accuracy": 0.16616733223199845,
      "num_tokens": 33145783.0,
      "step": 15125
    },
    {
      "entropy": 5.448919343948364,
      "epoch": 1.501587931718936,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047808855261232915,
      "loss": 5.3166,
      "mean_token_accuracy": 0.15991536974906922,
      "num_tokens": 33156404.0,
      "step": 15130
    },
    {
      "entropy": 5.395834159851074,
      "epoch": 1.5020841603811035,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047807329870830134,
      "loss": 5.3028,
      "mean_token_accuracy": 0.16115988045930862,
      "num_tokens": 33168226.0,
      "step": 15135
    },
    {
      "entropy": 5.421104621887207,
      "epoch": 1.5025803890432712,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004780580397684933,
      "loss": 5.2102,
      "mean_token_accuracy": 0.15856048166751863,
      "num_tokens": 33178265.0,
      "step": 15140
    },
    {
      "entropy": 5.4498780250549315,
      "epoch": 1.5030766177054387,
      "grad_norm": 1.0,
      "learning_rate": 0.0004780427757932835,
      "loss": 5.1654,
      "mean_token_accuracy": 0.16681558638811111,
      "num_tokens": 33188992.0,
      "step": 15145
    },
    {
      "entropy": 5.442611360549927,
      "epoch": 1.5035728463676064,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004780275067830503,
      "loss": 5.2347,
      "mean_token_accuracy": 0.16235430985689164,
      "num_tokens": 33199424.0,
      "step": 15150
    },
    {
      "entropy": 5.388536214828491,
      "epoch": 1.5040690750297738,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004780122327381725,
      "loss": 5.1743,
      "mean_token_accuracy": 0.16855715811252595,
      "num_tokens": 33210900.0,
      "step": 15155
    },
    {
      "entropy": 5.472215986251831,
      "epoch": 1.5045653036919413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047799695365902865,
      "loss": 5.2448,
      "mean_token_accuracy": 0.15824507176876068,
      "num_tokens": 33220796.0,
      "step": 15160
    },
    {
      "entropy": 5.4401780128479,
      "epoch": 1.5050615323541088,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004779816695459979,
      "loss": 5.2364,
      "mean_token_accuracy": 0.16059010326862336,
      "num_tokens": 33233454.0,
      "step": 15165
    },
    {
      "entropy": 5.427543354034424,
      "epoch": 1.5055577610162763,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004779663803994591,
      "loss": 5.1396,
      "mean_token_accuracy": 0.16548582017421723,
      "num_tokens": 33243574.0,
      "step": 15170
    },
    {
      "entropy": 5.382674932479858,
      "epoch": 1.5060539896784437,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004779510862197914,
      "loss": 5.184,
      "mean_token_accuracy": 0.16102853566408157,
      "num_tokens": 33256298.0,
      "step": 15175
    },
    {
      "entropy": 5.433162975311279,
      "epoch": 1.5065502183406112,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004779357870073741,
      "loss": 5.2966,
      "mean_token_accuracy": 0.1583479568362236,
      "num_tokens": 33268089.0,
      "step": 15180
    },
    {
      "entropy": 5.449942493438721,
      "epoch": 1.507046447002779,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047792048276258653,
      "loss": 5.2044,
      "mean_token_accuracy": 0.15479422360658646,
      "num_tokens": 33279955.0,
      "step": 15185
    },
    {
      "entropy": 5.422054815292358,
      "epoch": 1.5075426756649464,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004779051734858083,
      "loss": 5.1821,
      "mean_token_accuracy": 0.1619126558303833,
      "num_tokens": 33291470.0,
      "step": 15190
    },
    {
      "entropy": 5.4718828201293945,
      "epoch": 1.508038904327114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047788985917741894,
      "loss": 5.2374,
      "mean_token_accuracy": 0.16582146435976028,
      "num_tokens": 33304169.0,
      "step": 15195
    },
    {
      "entropy": 5.3124449253082275,
      "epoch": 1.5085351329892815,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047787453983779826,
      "loss": 5.0665,
      "mean_token_accuracy": 0.17236597687005997,
      "num_tokens": 33314436.0,
      "step": 15200
    },
    {
      "entropy": 5.401573514938354,
      "epoch": 1.509031361651449,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004778592154673262,
      "loss": 5.302,
      "mean_token_accuracy": 0.15696002691984176,
      "num_tokens": 33325361.0,
      "step": 15205
    },
    {
      "entropy": 5.3749768257141115,
      "epoch": 1.5095275903136165,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047784388606638263,
      "loss": 5.1732,
      "mean_token_accuracy": 0.15769533067941666,
      "num_tokens": 33336644.0,
      "step": 15210
    },
    {
      "entropy": 5.43075156211853,
      "epoch": 1.510023818975784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004778285516353478,
      "loss": 5.2223,
      "mean_token_accuracy": 0.1577630877494812,
      "num_tokens": 33348418.0,
      "step": 15215
    },
    {
      "entropy": 5.437948703765869,
      "epoch": 1.5105200476379514,
      "grad_norm": 1.0,
      "learning_rate": 0.00047781321217460205,
      "loss": 5.1666,
      "mean_token_accuracy": 0.16776464134454727,
      "num_tokens": 33360615.0,
      "step": 15220
    },
    {
      "entropy": 5.4100230693817135,
      "epoch": 1.5110162763001191,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047779786768452554,
      "loss": 5.1612,
      "mean_token_accuracy": 0.15920688658952714,
      "num_tokens": 33372030.0,
      "step": 15225
    },
    {
      "entropy": 5.3952337265014645,
      "epoch": 1.5115125049622866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047778251816549895,
      "loss": 5.1874,
      "mean_token_accuracy": 0.161589577794075,
      "num_tokens": 33382834.0,
      "step": 15230
    },
    {
      "entropy": 5.3972001552581785,
      "epoch": 1.512008733624454,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047776716361790284,
      "loss": 5.2156,
      "mean_token_accuracy": 0.169775553047657,
      "num_tokens": 33393250.0,
      "step": 15235
    },
    {
      "entropy": 5.422188663482666,
      "epoch": 1.5125049622866218,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047775180404211804,
      "loss": 5.1563,
      "mean_token_accuracy": 0.17075642496347426,
      "num_tokens": 33402778.0,
      "step": 15240
    },
    {
      "entropy": 5.475852918624878,
      "epoch": 1.5130011909487893,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004777364394385254,
      "loss": 5.2371,
      "mean_token_accuracy": 0.16123313158750535,
      "num_tokens": 33413565.0,
      "step": 15245
    },
    {
      "entropy": 5.450520610809326,
      "epoch": 1.5134974196109567,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047772106980750593,
      "loss": 5.168,
      "mean_token_accuracy": 0.15918315649032594,
      "num_tokens": 33424108.0,
      "step": 15250
    },
    {
      "entropy": 5.410372304916382,
      "epoch": 1.5139936482731242,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004777056951494407,
      "loss": 5.2815,
      "mean_token_accuracy": 0.1558677151799202,
      "num_tokens": 33435683.0,
      "step": 15255
    },
    {
      "entropy": 5.427628993988037,
      "epoch": 1.5144898769352917,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000477690315464711,
      "loss": 5.2287,
      "mean_token_accuracy": 0.1604921728372574,
      "num_tokens": 33446267.0,
      "step": 15260
    },
    {
      "entropy": 5.429253959655762,
      "epoch": 1.5149861055974592,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047767493075369827,
      "loss": 5.139,
      "mean_token_accuracy": 0.1776985615491867,
      "num_tokens": 33457901.0,
      "step": 15265
    },
    {
      "entropy": 5.495711946487427,
      "epoch": 1.5154823342596269,
      "grad_norm": 1.0625,
      "learning_rate": 0.000477659541016784,
      "loss": 5.3164,
      "mean_token_accuracy": 0.15869105309247972,
      "num_tokens": 33469215.0,
      "step": 15270
    },
    {
      "entropy": 5.453813743591309,
      "epoch": 1.5159785629217943,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047764414625434983,
      "loss": 5.16,
      "mean_token_accuracy": 0.16676609367132186,
      "num_tokens": 33480384.0,
      "step": 15275
    },
    {
      "entropy": 5.445078039169312,
      "epoch": 1.516474791583962,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004776287464667774,
      "loss": 5.1846,
      "mean_token_accuracy": 0.1673949398100376,
      "num_tokens": 33491404.0,
      "step": 15280
    },
    {
      "entropy": 5.525157499313354,
      "epoch": 1.5169710202461295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047761334165444876,
      "loss": 5.3085,
      "mean_token_accuracy": 0.14923258423805236,
      "num_tokens": 33501888.0,
      "step": 15285
    },
    {
      "entropy": 5.463419246673584,
      "epoch": 1.517467248908297,
      "grad_norm": 1.0,
      "learning_rate": 0.00047759793181774586,
      "loss": 5.2435,
      "mean_token_accuracy": 0.16156346797943116,
      "num_tokens": 33513274.0,
      "step": 15290
    },
    {
      "entropy": 5.425913095474243,
      "epoch": 1.5179634775704645,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047758251695705073,
      "loss": 5.2265,
      "mean_token_accuracy": 0.1602153077721596,
      "num_tokens": 33524188.0,
      "step": 15295
    },
    {
      "entropy": 5.455765295028686,
      "epoch": 1.518459706232632,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047756709707274576,
      "loss": 5.2942,
      "mean_token_accuracy": 0.15821214467287065,
      "num_tokens": 33535207.0,
      "step": 15300
    },
    {
      "entropy": 5.461170196533203,
      "epoch": 1.5189559348947994,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004775516721652133,
      "loss": 5.2395,
      "mean_token_accuracy": 0.1567886918783188,
      "num_tokens": 33546154.0,
      "step": 15305
    },
    {
      "entropy": 5.4443933963775635,
      "epoch": 1.519452163556967,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004775362422348358,
      "loss": 5.1437,
      "mean_token_accuracy": 0.16915892213582992,
      "num_tokens": 33555365.0,
      "step": 15310
    },
    {
      "entropy": 5.424424028396606,
      "epoch": 1.5199483922191346,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047752080728199586,
      "loss": 5.2172,
      "mean_token_accuracy": 0.1573648601770401,
      "num_tokens": 33566316.0,
      "step": 15315
    },
    {
      "entropy": 5.433236789703369,
      "epoch": 1.520444620881302,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004775053673070764,
      "loss": 5.1578,
      "mean_token_accuracy": 0.16895272433757783,
      "num_tokens": 33576549.0,
      "step": 15320
    },
    {
      "entropy": 5.412270641326904,
      "epoch": 1.5209408495434698,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004774899223104601,
      "loss": 5.2157,
      "mean_token_accuracy": 0.16570321917533876,
      "num_tokens": 33586681.0,
      "step": 15325
    },
    {
      "entropy": 5.555965757369995,
      "epoch": 1.5214370782056372,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004774744722925301,
      "loss": 5.3409,
      "mean_token_accuracy": 0.15516527891159057,
      "num_tokens": 33599201.0,
      "step": 15330
    },
    {
      "entropy": 5.4898459911346436,
      "epoch": 1.5219333068678047,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004774590172536695,
      "loss": 5.1907,
      "mean_token_accuracy": 0.16346394270658493,
      "num_tokens": 33610217.0,
      "step": 15335
    },
    {
      "entropy": 5.433169794082642,
      "epoch": 1.5224295355299722,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047744355719426146,
      "loss": 5.2398,
      "mean_token_accuracy": 0.1604839891195297,
      "num_tokens": 33620586.0,
      "step": 15340
    },
    {
      "entropy": 5.458272933959961,
      "epoch": 1.5229257641921397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047742809211468953,
      "loss": 5.2079,
      "mean_token_accuracy": 0.16237540394067765,
      "num_tokens": 33630712.0,
      "step": 15345
    },
    {
      "entropy": 5.4211705207824705,
      "epoch": 1.5234219928543071,
      "grad_norm": 1.078125,
      "learning_rate": 0.000477412622015337,
      "loss": 5.1557,
      "mean_token_accuracy": 0.1625997990369797,
      "num_tokens": 33640367.0,
      "step": 15350
    },
    {
      "entropy": 5.452194309234619,
      "epoch": 1.5239182215164748,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004773971468965876,
      "loss": 5.2607,
      "mean_token_accuracy": 0.15671809315681456,
      "num_tokens": 33651816.0,
      "step": 15355
    },
    {
      "entropy": 5.399330759048462,
      "epoch": 1.5244144501786423,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047738166675882516,
      "loss": 5.1311,
      "mean_token_accuracy": 0.16616967767477037,
      "num_tokens": 33663661.0,
      "step": 15360
    },
    {
      "entropy": 5.365200757980347,
      "epoch": 1.52491067884081,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004773661816024335,
      "loss": 5.0474,
      "mean_token_accuracy": 0.17197935730218888,
      "num_tokens": 33673670.0,
      "step": 15365
    },
    {
      "entropy": 5.399232482910156,
      "epoch": 1.5254069075029775,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004773506914277965,
      "loss": 5.2931,
      "mean_token_accuracy": 0.15727135241031648,
      "num_tokens": 33685434.0,
      "step": 15370
    },
    {
      "entropy": 5.449458360671997,
      "epoch": 1.525903136165145,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047733519623529853,
      "loss": 5.2364,
      "mean_token_accuracy": 0.16373806148767472,
      "num_tokens": 33696751.0,
      "step": 15375
    },
    {
      "entropy": 5.440271663665771,
      "epoch": 1.5263993648273124,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004773196960253236,
      "loss": 5.1767,
      "mean_token_accuracy": 0.16412730515003204,
      "num_tokens": 33708141.0,
      "step": 15380
    },
    {
      "entropy": 5.362249755859375,
      "epoch": 1.52689559348948,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004773041907982561,
      "loss": 5.0381,
      "mean_token_accuracy": 0.1734695017337799,
      "num_tokens": 33719219.0,
      "step": 15385
    },
    {
      "entropy": 5.444176292419433,
      "epoch": 1.5273918221516474,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047728868055448064,
      "loss": 5.1949,
      "mean_token_accuracy": 0.16440460681915284,
      "num_tokens": 33730493.0,
      "step": 15390
    },
    {
      "entropy": 5.462137651443482,
      "epoch": 1.5278880508138148,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004772731652943818,
      "loss": 5.238,
      "mean_token_accuracy": 0.1567269653081894,
      "num_tokens": 33740273.0,
      "step": 15395
    },
    {
      "entropy": 5.4346436023712155,
      "epoch": 1.5283842794759825,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004772576450183444,
      "loss": 5.2152,
      "mean_token_accuracy": 0.1648477092385292,
      "num_tokens": 33752162.0,
      "step": 15400
    },
    {
      "entropy": 5.427360582351684,
      "epoch": 1.52888050813815,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047724211972675315,
      "loss": 5.1655,
      "mean_token_accuracy": 0.15785127878189087,
      "num_tokens": 33763139.0,
      "step": 15405
    },
    {
      "entropy": 5.3990795612335205,
      "epoch": 1.5293767368003177,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004772265894199931,
      "loss": 5.2246,
      "mean_token_accuracy": 0.16230090260505675,
      "num_tokens": 33774641.0,
      "step": 15410
    },
    {
      "entropy": 5.480242109298706,
      "epoch": 1.5298729654624852,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004772110540984494,
      "loss": 5.2164,
      "mean_token_accuracy": 0.16483995467424392,
      "num_tokens": 33785644.0,
      "step": 15415
    },
    {
      "entropy": 5.545241069793701,
      "epoch": 1.5303691941246527,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004771955137625073,
      "loss": 5.3432,
      "mean_token_accuracy": 0.15191099792718887,
      "num_tokens": 33797880.0,
      "step": 15420
    },
    {
      "entropy": 5.350762462615966,
      "epoch": 1.5308654227868201,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047717996841255224,
      "loss": 5.1569,
      "mean_token_accuracy": 0.16467146575450897,
      "num_tokens": 33808782.0,
      "step": 15425
    },
    {
      "entropy": 5.394152069091797,
      "epoch": 1.5313616514489876,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047716441804896955,
      "loss": 5.1953,
      "mean_token_accuracy": 0.16922770142555238,
      "num_tokens": 33818483.0,
      "step": 15430
    },
    {
      "entropy": 5.419955635070801,
      "epoch": 1.531857880111155,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004771488626721449,
      "loss": 5.1836,
      "mean_token_accuracy": 0.162690931558609,
      "num_tokens": 33828590.0,
      "step": 15435
    },
    {
      "entropy": 5.4699453830719,
      "epoch": 1.5323541087733228,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047713330228246406,
      "loss": 5.2889,
      "mean_token_accuracy": 0.16116743236780168,
      "num_tokens": 33838978.0,
      "step": 15440
    },
    {
      "entropy": 5.4819855213165285,
      "epoch": 1.5328503374354903,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004771177368803129,
      "loss": 5.2093,
      "mean_token_accuracy": 0.16332343816757203,
      "num_tokens": 33851244.0,
      "step": 15445
    },
    {
      "entropy": 5.4326482772827145,
      "epoch": 1.533346566097658,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047710216646607733,
      "loss": 5.1088,
      "mean_token_accuracy": 0.16946526020765304,
      "num_tokens": 33861666.0,
      "step": 15450
    },
    {
      "entropy": 5.429335832595825,
      "epoch": 1.5338427947598254,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047708659104014363,
      "loss": 5.183,
      "mean_token_accuracy": 0.16721709072589874,
      "num_tokens": 33872416.0,
      "step": 15455
    },
    {
      "entropy": 5.363960409164429,
      "epoch": 1.534339023421993,
      "grad_norm": 1.125,
      "learning_rate": 0.0004770710106028978,
      "loss": 5.1697,
      "mean_token_accuracy": 0.16559816747903824,
      "num_tokens": 33881808.0,
      "step": 15460
    },
    {
      "entropy": 5.463426733016968,
      "epoch": 1.5348352520841604,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047705542515472647,
      "loss": 5.2378,
      "mean_token_accuracy": 0.16246891766786575,
      "num_tokens": 33892925.0,
      "step": 15465
    },
    {
      "entropy": 5.4612470149993895,
      "epoch": 1.5353314807463279,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004770398346960159,
      "loss": 5.2052,
      "mean_token_accuracy": 0.15585589408874512,
      "num_tokens": 33903702.0,
      "step": 15470
    },
    {
      "entropy": 5.490330219268799,
      "epoch": 1.5358277094084953,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047702423922715284,
      "loss": 5.2804,
      "mean_token_accuracy": 0.15879511833190918,
      "num_tokens": 33917233.0,
      "step": 15475
    },
    {
      "entropy": 5.492999601364136,
      "epoch": 1.5363239380706628,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004770086387485239,
      "loss": 5.2328,
      "mean_token_accuracy": 0.16183273792266845,
      "num_tokens": 33928362.0,
      "step": 15480
    },
    {
      "entropy": 5.511008834838867,
      "epoch": 1.5368201667328305,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000476993033260516,
      "loss": 5.2649,
      "mean_token_accuracy": 0.16127101927995682,
      "num_tokens": 33938624.0,
      "step": 15485
    },
    {
      "entropy": 5.375752544403076,
      "epoch": 1.537316395394998,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004769774227635162,
      "loss": 5.1513,
      "mean_token_accuracy": 0.16260193139314652,
      "num_tokens": 33950058.0,
      "step": 15490
    },
    {
      "entropy": 5.492281150817871,
      "epoch": 1.5378126240571657,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004769618072579115,
      "loss": 5.2638,
      "mean_token_accuracy": 0.15887387245893478,
      "num_tokens": 33962404.0,
      "step": 15495
    },
    {
      "entropy": 5.496834468841553,
      "epoch": 1.5383088527193332,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047694618674408917,
      "loss": 5.206,
      "mean_token_accuracy": 0.16177328377962114,
      "num_tokens": 33973095.0,
      "step": 15500
    },
    {
      "entropy": 5.392947340011597,
      "epoch": 1.5388050813815006,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047693056122243655,
      "loss": 5.225,
      "mean_token_accuracy": 0.15969944447278978,
      "num_tokens": 33984152.0,
      "step": 15505
    },
    {
      "entropy": 5.443367147445679,
      "epoch": 1.539301310043668,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047691493069334124,
      "loss": 5.2246,
      "mean_token_accuracy": 0.16181586533784867,
      "num_tokens": 33994854.0,
      "step": 15510
    },
    {
      "entropy": 5.485648107528687,
      "epoch": 1.5397975387058356,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047689929515719066,
      "loss": 5.2374,
      "mean_token_accuracy": 0.16289484947919847,
      "num_tokens": 34006573.0,
      "step": 15515
    },
    {
      "entropy": 5.454595422744751,
      "epoch": 1.540293767368003,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047688365461437263,
      "loss": 5.1585,
      "mean_token_accuracy": 0.16113646626472472,
      "num_tokens": 34016537.0,
      "step": 15520
    },
    {
      "entropy": 5.409765100479126,
      "epoch": 1.5407899960301707,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047686800906527493,
      "loss": 5.201,
      "mean_token_accuracy": 0.1714401587843895,
      "num_tokens": 34026825.0,
      "step": 15525
    },
    {
      "entropy": 5.482147693634033,
      "epoch": 1.5412862246923382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004768523585102856,
      "loss": 5.1858,
      "mean_token_accuracy": 0.1696465089917183,
      "num_tokens": 34036786.0,
      "step": 15530
    },
    {
      "entropy": 5.451251125335693,
      "epoch": 1.541782453354506,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004768367029497928,
      "loss": 5.1938,
      "mean_token_accuracy": 0.16580987870693206,
      "num_tokens": 34046784.0,
      "step": 15535
    },
    {
      "entropy": 5.452660417556762,
      "epoch": 1.5422786820166734,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047682104238418464,
      "loss": 5.2365,
      "mean_token_accuracy": 0.16390419900417327,
      "num_tokens": 34058250.0,
      "step": 15540
    },
    {
      "entropy": 5.370761919021606,
      "epoch": 1.5427749106788409,
      "grad_norm": 1.125,
      "learning_rate": 0.00047680537681384956,
      "loss": 5.1811,
      "mean_token_accuracy": 0.16614935994148256,
      "num_tokens": 34069969.0,
      "step": 15545
    },
    {
      "entropy": 5.464462375640869,
      "epoch": 1.5432711393410083,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004767897062391759,
      "loss": 5.3042,
      "mean_token_accuracy": 0.15810180231928825,
      "num_tokens": 34080817.0,
      "step": 15550
    },
    {
      "entropy": 5.4380738735198975,
      "epoch": 1.5437673680031758,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004767740306605524,
      "loss": 5.2075,
      "mean_token_accuracy": 0.1595169261097908,
      "num_tokens": 34091138.0,
      "step": 15555
    },
    {
      "entropy": 5.422082471847534,
      "epoch": 1.5442635966653433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004767583500783677,
      "loss": 5.162,
      "mean_token_accuracy": 0.16570256948471068,
      "num_tokens": 34102035.0,
      "step": 15560
    },
    {
      "entropy": 5.412417602539063,
      "epoch": 1.5447598253275108,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047674266449301074,
      "loss": 5.2027,
      "mean_token_accuracy": 0.1637146458029747,
      "num_tokens": 34112533.0,
      "step": 15565
    },
    {
      "entropy": 5.401985549926758,
      "epoch": 1.5452560539896785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047672697390487034,
      "loss": 5.2059,
      "mean_token_accuracy": 0.16969087570905686,
      "num_tokens": 34124196.0,
      "step": 15570
    },
    {
      "entropy": 5.427552795410156,
      "epoch": 1.545752282651846,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004767112783143357,
      "loss": 5.1875,
      "mean_token_accuracy": 0.16745086908340454,
      "num_tokens": 34135635.0,
      "step": 15575
    },
    {
      "entropy": 5.443451261520385,
      "epoch": 1.5462485113140136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000476695577721796,
      "loss": 5.1311,
      "mean_token_accuracy": 0.16368705332279204,
      "num_tokens": 34146202.0,
      "step": 15580
    },
    {
      "entropy": 5.392278623580933,
      "epoch": 1.5467447399761811,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047667987212764065,
      "loss": 5.175,
      "mean_token_accuracy": 0.16482080966234208,
      "num_tokens": 34157354.0,
      "step": 15585
    },
    {
      "entropy": 5.351127576828003,
      "epoch": 1.5472409686383486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047666416153225894,
      "loss": 5.2182,
      "mean_token_accuracy": 0.15922287851572037,
      "num_tokens": 34168563.0,
      "step": 15590
    },
    {
      "entropy": 5.417463684082032,
      "epoch": 1.547737197300516,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004766484459360406,
      "loss": 5.1573,
      "mean_token_accuracy": 0.16358204036951066,
      "num_tokens": 34178604.0,
      "step": 15595
    },
    {
      "entropy": 5.491041898727417,
      "epoch": 1.5482334259626835,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047663272533937537,
      "loss": 5.2005,
      "mean_token_accuracy": 0.16533066034317018,
      "num_tokens": 34190321.0,
      "step": 15600
    },
    {
      "entropy": 5.394814443588257,
      "epoch": 1.548729654624851,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047661699974265294,
      "loss": 5.2122,
      "mean_token_accuracy": 0.16280404925346376,
      "num_tokens": 34201519.0,
      "step": 15605
    },
    {
      "entropy": 5.396702671051026,
      "epoch": 1.5492258832870187,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047660126914626344,
      "loss": 5.1547,
      "mean_token_accuracy": 0.16249047815799714,
      "num_tokens": 34212710.0,
      "step": 15610
    },
    {
      "entropy": 5.495629215240479,
      "epoch": 1.5497221119491862,
      "grad_norm": 1.125,
      "learning_rate": 0.0004765855335505969,
      "loss": 5.2855,
      "mean_token_accuracy": 0.1618147909641266,
      "num_tokens": 34223187.0,
      "step": 15615
    },
    {
      "entropy": 5.473035097122192,
      "epoch": 1.5502183406113537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004765697929560434,
      "loss": 5.2215,
      "mean_token_accuracy": 0.16943839192390442,
      "num_tokens": 34234530.0,
      "step": 15620
    },
    {
      "entropy": 5.437847042083741,
      "epoch": 1.5507145692735214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047655404736299354,
      "loss": 5.2261,
      "mean_token_accuracy": 0.15874495804309846,
      "num_tokens": 34245003.0,
      "step": 15625
    },
    {
      "entropy": 5.411380624771118,
      "epoch": 1.5512107979356888,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047653829677183745,
      "loss": 5.2022,
      "mean_token_accuracy": 0.16575342267751694,
      "num_tokens": 34255630.0,
      "step": 15630
    },
    {
      "entropy": 5.439237403869629,
      "epoch": 1.5517070265978563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047652254118296596,
      "loss": 5.1944,
      "mean_token_accuracy": 0.16493301391601561,
      "num_tokens": 34266716.0,
      "step": 15635
    },
    {
      "entropy": 5.440914535522461,
      "epoch": 1.5522032552600238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004765067805967697,
      "loss": 5.2124,
      "mean_token_accuracy": 0.16261286437511444,
      "num_tokens": 34278905.0,
      "step": 15640
    },
    {
      "entropy": 5.438565969467163,
      "epoch": 1.5526994839221913,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004764910150136395,
      "loss": 5.1579,
      "mean_token_accuracy": 0.16449419558048248,
      "num_tokens": 34289688.0,
      "step": 15645
    },
    {
      "entropy": 5.447380781173706,
      "epoch": 1.5531957125843587,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047647524443396625,
      "loss": 5.2136,
      "mean_token_accuracy": 0.1600058689713478,
      "num_tokens": 34300964.0,
      "step": 15650
    },
    {
      "entropy": 5.424160099029541,
      "epoch": 1.5536919412465264,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047645946885814105,
      "loss": 5.198,
      "mean_token_accuracy": 0.16325591802597045,
      "num_tokens": 34311435.0,
      "step": 15655
    },
    {
      "entropy": 5.430169534683228,
      "epoch": 1.554188169908694,
      "grad_norm": 1.125,
      "learning_rate": 0.0004764436882865552,
      "loss": 5.2059,
      "mean_token_accuracy": 0.1583005368709564,
      "num_tokens": 34321742.0,
      "step": 15660
    },
    {
      "entropy": 5.459180116653442,
      "epoch": 1.5546843985708616,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004764279027195999,
      "loss": 5.2608,
      "mean_token_accuracy": 0.15546346306800843,
      "num_tokens": 34333796.0,
      "step": 15665
    },
    {
      "entropy": 5.619548273086548,
      "epoch": 1.555180627233029,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047641211215766673,
      "loss": 5.3185,
      "mean_token_accuracy": 0.1616634711623192,
      "num_tokens": 34344521.0,
      "step": 15670
    },
    {
      "entropy": 5.393112468719482,
      "epoch": 1.5556768558951966,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004763963166011471,
      "loss": 5.1964,
      "mean_token_accuracy": 0.16805144399404526,
      "num_tokens": 34355629.0,
      "step": 15675
    },
    {
      "entropy": 5.372286319732666,
      "epoch": 1.556173084557364,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004763805160504328,
      "loss": 5.2496,
      "mean_token_accuracy": 0.16343512684106826,
      "num_tokens": 34366935.0,
      "step": 15680
    },
    {
      "entropy": 5.417632246017456,
      "epoch": 1.5566693132195315,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004763647105059157,
      "loss": 5.1628,
      "mean_token_accuracy": 0.15801329910755157,
      "num_tokens": 34378355.0,
      "step": 15685
    },
    {
      "entropy": 5.412997102737426,
      "epoch": 1.557165541881699,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004763488999679877,
      "loss": 5.1105,
      "mean_token_accuracy": 0.16227585971355438,
      "num_tokens": 34388811.0,
      "step": 15690
    },
    {
      "entropy": 5.513126611709595,
      "epoch": 1.5576617705438665,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047633308443704076,
      "loss": 5.2386,
      "mean_token_accuracy": 0.1560899317264557,
      "num_tokens": 34399905.0,
      "step": 15695
    },
    {
      "entropy": 5.370826053619385,
      "epoch": 1.5581579992060341,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047631726391346717,
      "loss": 5.118,
      "mean_token_accuracy": 0.16351733207702637,
      "num_tokens": 34410338.0,
      "step": 15700
    },
    {
      "entropy": 5.45331072807312,
      "epoch": 1.5586542278682016,
      "grad_norm": 1.125,
      "learning_rate": 0.0004763014383976593,
      "loss": 5.2829,
      "mean_token_accuracy": 0.16049907952547074,
      "num_tokens": 34421204.0,
      "step": 15705
    },
    {
      "entropy": 5.393855047225952,
      "epoch": 1.5591504565303693,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047628560789000945,
      "loss": 5.1069,
      "mean_token_accuracy": 0.17336554676294327,
      "num_tokens": 34431002.0,
      "step": 15710
    },
    {
      "entropy": 5.370547437667847,
      "epoch": 1.5596466851925368,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047626977239091027,
      "loss": 5.1411,
      "mean_token_accuracy": 0.16525082439184188,
      "num_tokens": 34442446.0,
      "step": 15715
    },
    {
      "entropy": 5.405730962753296,
      "epoch": 1.5601429138547043,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004762539319007545,
      "loss": 5.1191,
      "mean_token_accuracy": 0.16598850935697557,
      "num_tokens": 34453590.0,
      "step": 15720
    },
    {
      "entropy": 5.476835775375366,
      "epoch": 1.5606391425168717,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004762380864199348,
      "loss": 5.2569,
      "mean_token_accuracy": 0.16224393546581267,
      "num_tokens": 34465614.0,
      "step": 15725
    },
    {
      "entropy": 5.507569646835327,
      "epoch": 1.5611353711790392,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004762222359488443,
      "loss": 5.3472,
      "mean_token_accuracy": 0.15399719998240471,
      "num_tokens": 34476616.0,
      "step": 15730
    },
    {
      "entropy": 5.453872728347778,
      "epoch": 1.5616315998412067,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004762063804878758,
      "loss": 5.1415,
      "mean_token_accuracy": 0.1604670763015747,
      "num_tokens": 34486260.0,
      "step": 15735
    },
    {
      "entropy": 5.422682666778565,
      "epoch": 1.5621278285033744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047619052003742277,
      "loss": 5.2288,
      "mean_token_accuracy": 0.16447885185480118,
      "num_tokens": 34496392.0,
      "step": 15740
    },
    {
      "entropy": 5.429277229309082,
      "epoch": 1.5626240571655419,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004761746545978782,
      "loss": 5.2036,
      "mean_token_accuracy": 0.16769365668296815,
      "num_tokens": 34507218.0,
      "step": 15745
    },
    {
      "entropy": 5.504792308807373,
      "epoch": 1.5631202858277096,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004761587841696358,
      "loss": 5.2704,
      "mean_token_accuracy": 0.15601086542010306,
      "num_tokens": 34518888.0,
      "step": 15750
    },
    {
      "entropy": 5.4458136558532715,
      "epoch": 1.563616514489877,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000476142908753089,
      "loss": 5.1849,
      "mean_token_accuracy": 0.16326580941677094,
      "num_tokens": 34529865.0,
      "step": 15755
    },
    {
      "entropy": 5.4703858375549315,
      "epoch": 1.5641127431520445,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004761270283486314,
      "loss": 5.2615,
      "mean_token_accuracy": 0.15775466859340667,
      "num_tokens": 34542541.0,
      "step": 15760
    },
    {
      "entropy": 5.48168306350708,
      "epoch": 1.564608971814212,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047611114295665697,
      "loss": 5.2482,
      "mean_token_accuracy": 0.16499602049589157,
      "num_tokens": 34553056.0,
      "step": 15765
    },
    {
      "entropy": 5.4970938205719,
      "epoch": 1.5651052004763795,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047609525257755967,
      "loss": 5.2365,
      "mean_token_accuracy": 0.15997283160686493,
      "num_tokens": 34563344.0,
      "step": 15770
    },
    {
      "entropy": 5.44418773651123,
      "epoch": 1.565601429138547,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004760793572117333,
      "loss": 5.1298,
      "mean_token_accuracy": 0.16773202866315842,
      "num_tokens": 34574320.0,
      "step": 15775
    },
    {
      "entropy": 5.392736625671387,
      "epoch": 1.5660976578007144,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004760634568595721,
      "loss": 5.2031,
      "mean_token_accuracy": 0.15500207245349884,
      "num_tokens": 34584728.0,
      "step": 15780
    },
    {
      "entropy": 5.473801898956299,
      "epoch": 1.5665938864628821,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004760475515214704,
      "loss": 5.3123,
      "mean_token_accuracy": 0.15665165781974794,
      "num_tokens": 34596395.0,
      "step": 15785
    },
    {
      "entropy": 5.4012809753417965,
      "epoch": 1.5670901151250496,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004760316411978228,
      "loss": 5.2042,
      "mean_token_accuracy": 0.1580704152584076,
      "num_tokens": 34607950.0,
      "step": 15790
    },
    {
      "entropy": 5.433672523498535,
      "epoch": 1.5675863437872173,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047601572588902363,
      "loss": 5.2064,
      "mean_token_accuracy": 0.16906123608350754,
      "num_tokens": 34617640.0,
      "step": 15795
    },
    {
      "entropy": 5.44781527519226,
      "epoch": 1.5680825724493848,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047599980559546757,
      "loss": 5.2482,
      "mean_token_accuracy": 0.15641605406999587,
      "num_tokens": 34629475.0,
      "step": 15800
    },
    {
      "entropy": 5.41588282585144,
      "epoch": 1.5685788011115522,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004759838803175495,
      "loss": 5.1818,
      "mean_token_accuracy": 0.1588212490081787,
      "num_tokens": 34641423.0,
      "step": 15805
    },
    {
      "entropy": 5.5271978855133055,
      "epoch": 1.5690750297737197,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047596795005566423,
      "loss": 5.2422,
      "mean_token_accuracy": 0.1571370117366314,
      "num_tokens": 34651894.0,
      "step": 15810
    },
    {
      "entropy": 5.477277994155884,
      "epoch": 1.5695712584358872,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047595201481020685,
      "loss": 5.2374,
      "mean_token_accuracy": 0.1620178371667862,
      "num_tokens": 34664177.0,
      "step": 15815
    },
    {
      "entropy": 5.4406365871429445,
      "epoch": 1.5700674870980547,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047593607458157257,
      "loss": 5.1983,
      "mean_token_accuracy": 0.16601175665855408,
      "num_tokens": 34674850.0,
      "step": 15820
    },
    {
      "entropy": 5.429003953933716,
      "epoch": 1.5705637157602224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047592012937015647,
      "loss": 5.2259,
      "mean_token_accuracy": 0.1642740398645401,
      "num_tokens": 34685719.0,
      "step": 15825
    },
    {
      "entropy": 5.470280599594116,
      "epoch": 1.5710599444223898,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047590417917635423,
      "loss": 5.2238,
      "mean_token_accuracy": 0.1587255284190178,
      "num_tokens": 34697261.0,
      "step": 15830
    },
    {
      "entropy": 5.442282676696777,
      "epoch": 1.5715561730845575,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004758882240005613,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1637085869908333,
      "num_tokens": 34708196.0,
      "step": 15835
    },
    {
      "entropy": 5.456255292892456,
      "epoch": 1.572052401746725,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047587226384317323,
      "loss": 5.251,
      "mean_token_accuracy": 0.16009379476308822,
      "num_tokens": 34718855.0,
      "step": 15840
    },
    {
      "entropy": 5.475618457794189,
      "epoch": 1.5725486304088925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047585629870458584,
      "loss": 5.1764,
      "mean_token_accuracy": 0.16853090077638627,
      "num_tokens": 34731137.0,
      "step": 15845
    },
    {
      "entropy": 5.440131664276123,
      "epoch": 1.57304485907106,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004758403285851951,
      "loss": 5.1372,
      "mean_token_accuracy": 0.1688038483262062,
      "num_tokens": 34741628.0,
      "step": 15850
    },
    {
      "entropy": 5.381322860717773,
      "epoch": 1.5735410877332274,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047582435348539695,
      "loss": 5.217,
      "mean_token_accuracy": 0.16351741254329683,
      "num_tokens": 34752984.0,
      "step": 15855
    },
    {
      "entropy": 5.461823558807373,
      "epoch": 1.574037316395395,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004758083734055875,
      "loss": 5.2902,
      "mean_token_accuracy": 0.15422503650188446,
      "num_tokens": 34764363.0,
      "step": 15860
    },
    {
      "entropy": 5.478481960296631,
      "epoch": 1.5745335450575624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004757923883461632,
      "loss": 5.2383,
      "mean_token_accuracy": 0.15731287747621536,
      "num_tokens": 34776289.0,
      "step": 15865
    },
    {
      "entropy": 5.4630931377410885,
      "epoch": 1.57502977371973,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047577639830752034,
      "loss": 5.2082,
      "mean_token_accuracy": 0.16364870518445968,
      "num_tokens": 34786760.0,
      "step": 15870
    },
    {
      "entropy": 5.408983278274536,
      "epoch": 1.5755260023818976,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047576040329005536,
      "loss": 5.188,
      "mean_token_accuracy": 0.16306786388158798,
      "num_tokens": 34797737.0,
      "step": 15875
    },
    {
      "entropy": 5.4296458721160885,
      "epoch": 1.5760222310440652,
      "grad_norm": 1.0625,
      "learning_rate": 0.000475744403294165,
      "loss": 5.2204,
      "mean_token_accuracy": 0.1633307680487633,
      "num_tokens": 34809650.0,
      "step": 15880
    },
    {
      "entropy": 5.530984401702881,
      "epoch": 1.5765184597062327,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047572839832024606,
      "loss": 5.246,
      "mean_token_accuracy": 0.15678427666425704,
      "num_tokens": 34820171.0,
      "step": 15885
    },
    {
      "entropy": 5.367771768569947,
      "epoch": 1.5770146883684002,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047571238836869537,
      "loss": 5.1016,
      "mean_token_accuracy": 0.1716366708278656,
      "num_tokens": 34831286.0,
      "step": 15890
    },
    {
      "entropy": 5.332991409301758,
      "epoch": 1.5775109170305677,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004756963734399099,
      "loss": 5.1239,
      "mean_token_accuracy": 0.17053066492080687,
      "num_tokens": 34841767.0,
      "step": 15895
    },
    {
      "entropy": 5.438945007324219,
      "epoch": 1.5780071456927351,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047568035353428683,
      "loss": 5.2281,
      "mean_token_accuracy": 0.16056728959083558,
      "num_tokens": 34852644.0,
      "step": 15900
    },
    {
      "entropy": 5.519192790985107,
      "epoch": 1.5785033743549026,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047566432865222343,
      "loss": 5.2094,
      "mean_token_accuracy": 0.16914456635713576,
      "num_tokens": 34863605.0,
      "step": 15905
    },
    {
      "entropy": 5.381642198562622,
      "epoch": 1.5789996030170703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047564829879411706,
      "loss": 5.1727,
      "mean_token_accuracy": 0.16675966531038283,
      "num_tokens": 34874594.0,
      "step": 15910
    },
    {
      "entropy": 5.429270505905151,
      "epoch": 1.5794958316792378,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004756322639603653,
      "loss": 5.2644,
      "mean_token_accuracy": 0.16047297418117523,
      "num_tokens": 34884627.0,
      "step": 15915
    },
    {
      "entropy": 5.546004819869995,
      "epoch": 1.5799920603414053,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047561622415136565,
      "loss": 5.2457,
      "mean_token_accuracy": 0.16672524511814119,
      "num_tokens": 34895679.0,
      "step": 15920
    },
    {
      "entropy": 5.415398359298706,
      "epoch": 1.580488289003573,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047560017936751597,
      "loss": 5.1593,
      "mean_token_accuracy": 0.16720156371593475,
      "num_tokens": 34906811.0,
      "step": 15925
    },
    {
      "entropy": 5.400812149047852,
      "epoch": 1.5809845176657404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004755841296092141,
      "loss": 5.2052,
      "mean_token_accuracy": 0.16012502759695052,
      "num_tokens": 34918655.0,
      "step": 15930
    },
    {
      "entropy": 5.420860528945923,
      "epoch": 1.581480746327908,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047556807487685797,
      "loss": 5.1879,
      "mean_token_accuracy": 0.16344253122806549,
      "num_tokens": 34931034.0,
      "step": 15935
    },
    {
      "entropy": 5.479482173919678,
      "epoch": 1.5819769749900754,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004755520151708458,
      "loss": 5.1436,
      "mean_token_accuracy": 0.16672358214855193,
      "num_tokens": 34939707.0,
      "step": 15940
    },
    {
      "entropy": 5.411362981796264,
      "epoch": 1.5824732036522429,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004755359504915758,
      "loss": 5.2658,
      "mean_token_accuracy": 0.156767837703228,
      "num_tokens": 34949237.0,
      "step": 15945
    },
    {
      "entropy": 5.491056680679321,
      "epoch": 1.5829694323144103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004755198808394464,
      "loss": 5.2928,
      "mean_token_accuracy": 0.16406552493572235,
      "num_tokens": 34960380.0,
      "step": 15950
    },
    {
      "entropy": 5.466914176940918,
      "epoch": 1.583465660976578,
      "grad_norm": 1.046875,
      "learning_rate": 0.000475503806214856,
      "loss": 5.2453,
      "mean_token_accuracy": 0.16110083162784578,
      "num_tokens": 34971699.0,
      "step": 15955
    },
    {
      "entropy": 5.462534236907959,
      "epoch": 1.5839618896387455,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004754877266182033,
      "loss": 5.2049,
      "mean_token_accuracy": 0.16689397096633912,
      "num_tokens": 34982333.0,
      "step": 15960
    },
    {
      "entropy": 5.443154811859131,
      "epoch": 1.5844581183009132,
      "grad_norm": 1.109375,
      "learning_rate": 0.000475471642049887,
      "loss": 5.2405,
      "mean_token_accuracy": 0.1609290823340416,
      "num_tokens": 34993277.0,
      "step": 15965
    },
    {
      "entropy": 5.455015516281128,
      "epoch": 1.5849543469630807,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004754555525103058,
      "loss": 5.1801,
      "mean_token_accuracy": 0.16444890797138215,
      "num_tokens": 35004252.0,
      "step": 15970
    },
    {
      "entropy": 5.39490270614624,
      "epoch": 1.5854505756252482,
      "grad_norm": 1.078125,
      "learning_rate": 0.000475439457999859,
      "loss": 5.1734,
      "mean_token_accuracy": 0.16916212439537048,
      "num_tokens": 35016974.0,
      "step": 15975
    },
    {
      "entropy": 5.400368785858154,
      "epoch": 1.5859468042874156,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047542335851894553,
      "loss": 5.2534,
      "mean_token_accuracy": 0.1606230691075325,
      "num_tokens": 35029017.0,
      "step": 15980
    },
    {
      "entropy": 5.486992406845093,
      "epoch": 1.586443032949583,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047540725406796454,
      "loss": 5.2852,
      "mean_token_accuracy": 0.16176501363515855,
      "num_tokens": 35040239.0,
      "step": 15985
    },
    {
      "entropy": 5.495342159271241,
      "epoch": 1.5869392616117506,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004753911446473156,
      "loss": 5.1876,
      "mean_token_accuracy": 0.15998111963272094,
      "num_tokens": 35051426.0,
      "step": 15990
    },
    {
      "entropy": 5.3526914596557615,
      "epoch": 1.587435490273918,
      "grad_norm": 1.15625,
      "learning_rate": 0.000475375030257398,
      "loss": 5.1663,
      "mean_token_accuracy": 0.16089655607938766,
      "num_tokens": 35061777.0,
      "step": 15995
    },
    {
      "entropy": 5.417815351486206,
      "epoch": 1.5879317189360858,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004753589108986114,
      "loss": 5.1196,
      "mean_token_accuracy": 0.176739864051342,
      "num_tokens": 35071822.0,
      "step": 16000
    },
    {
      "entropy": 5.443265438079834,
      "epoch": 1.5884279475982532,
      "grad_norm": 1.125,
      "learning_rate": 0.00047534278657135565,
      "loss": 5.2127,
      "mean_token_accuracy": 0.16318365633487703,
      "num_tokens": 35081900.0,
      "step": 16005
    },
    {
      "entropy": 5.4573578357696535,
      "epoch": 1.588924176260421,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004753266572760304,
      "loss": 5.264,
      "mean_token_accuracy": 0.1591874048113823,
      "num_tokens": 35092866.0,
      "step": 16010
    },
    {
      "entropy": 5.627674674987793,
      "epoch": 1.5894204049225884,
      "grad_norm": 1.125,
      "learning_rate": 0.00047531052301303577,
      "loss": 5.3728,
      "mean_token_accuracy": 0.15928667932748794,
      "num_tokens": 35103425.0,
      "step": 16015
    },
    {
      "entropy": 5.416684818267822,
      "epoch": 1.5899166335847559,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004752943837827718,
      "loss": 5.133,
      "mean_token_accuracy": 0.16403217911720275,
      "num_tokens": 35114620.0,
      "step": 16020
    },
    {
      "entropy": 5.4717213153839115,
      "epoch": 1.5904128622469234,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004752782395856387,
      "loss": 5.2451,
      "mean_token_accuracy": 0.1566806674003601,
      "num_tokens": 35125199.0,
      "step": 16025
    },
    {
      "entropy": 5.404940032958985,
      "epoch": 1.5909090909090908,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004752620904220368,
      "loss": 5.208,
      "mean_token_accuracy": 0.16773392558097838,
      "num_tokens": 35135892.0,
      "step": 16030
    },
    {
      "entropy": 5.417830991744995,
      "epoch": 1.5914053195712583,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004752459362923666,
      "loss": 5.204,
      "mean_token_accuracy": 0.16370364129543305,
      "num_tokens": 35146186.0,
      "step": 16035
    },
    {
      "entropy": 5.456462669372558,
      "epoch": 1.591901548233426,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004752297771970287,
      "loss": 5.2528,
      "mean_token_accuracy": 0.1563214361667633,
      "num_tokens": 35157261.0,
      "step": 16040
    },
    {
      "entropy": 5.469935274124145,
      "epoch": 1.5923977768955935,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047521361313642373,
      "loss": 5.1989,
      "mean_token_accuracy": 0.1661875620484352,
      "num_tokens": 35168508.0,
      "step": 16045
    },
    {
      "entropy": 5.419561195373535,
      "epoch": 1.5928940055577612,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004751974441109526,
      "loss": 5.1584,
      "mean_token_accuracy": 0.16643386483192443,
      "num_tokens": 35179444.0,
      "step": 16050
    },
    {
      "entropy": 5.3913195610046385,
      "epoch": 1.5933902342199286,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047518127012101624,
      "loss": 5.1736,
      "mean_token_accuracy": 0.16854743361473085,
      "num_tokens": 35189517.0,
      "step": 16055
    },
    {
      "entropy": 5.4189293384552,
      "epoch": 1.5938864628820961,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004751650911670157,
      "loss": 5.2082,
      "mean_token_accuracy": 0.16142287999391555,
      "num_tokens": 35200191.0,
      "step": 16060
    },
    {
      "entropy": 5.454503726959229,
      "epoch": 1.5943826915442636,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047514890724935236,
      "loss": 5.2319,
      "mean_token_accuracy": 0.15894406139850617,
      "num_tokens": 35211260.0,
      "step": 16065
    },
    {
      "entropy": 5.3598497867584225,
      "epoch": 1.594878920206431,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004751327183684272,
      "loss": 5.2028,
      "mean_token_accuracy": 0.16138967722654343,
      "num_tokens": 35222408.0,
      "step": 16070
    },
    {
      "entropy": 5.480382442474365,
      "epoch": 1.5953751488685985,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047511652452464196,
      "loss": 5.2435,
      "mean_token_accuracy": 0.16363485902547836,
      "num_tokens": 35232760.0,
      "step": 16075
    },
    {
      "entropy": 5.48322606086731,
      "epoch": 1.595871377530766,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047510032571839816,
      "loss": 5.2012,
      "mean_token_accuracy": 0.15964706391096115,
      "num_tokens": 35243849.0,
      "step": 16080
    },
    {
      "entropy": 5.449856090545654,
      "epoch": 1.5963676061929337,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004750841219500975,
      "loss": 5.2279,
      "mean_token_accuracy": 0.15955229848623276,
      "num_tokens": 35254962.0,
      "step": 16085
    },
    {
      "entropy": 5.417530059814453,
      "epoch": 1.5968638348551012,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004750679132201416,
      "loss": 5.2866,
      "mean_token_accuracy": 0.1580141916871071,
      "num_tokens": 35265895.0,
      "step": 16090
    },
    {
      "entropy": 5.55294246673584,
      "epoch": 1.597360063517269,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047505169952893276,
      "loss": 5.2894,
      "mean_token_accuracy": 0.16007387340068818,
      "num_tokens": 35277543.0,
      "step": 16095
    },
    {
      "entropy": 5.43766508102417,
      "epoch": 1.5978562921794364,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047503548087687267,
      "loss": 5.1985,
      "mean_token_accuracy": 0.16524655967950821,
      "num_tokens": 35289130.0,
      "step": 16100
    },
    {
      "entropy": 5.348242235183716,
      "epoch": 1.5983525208416038,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047501925726436376,
      "loss": 5.1463,
      "mean_token_accuracy": 0.1592506468296051,
      "num_tokens": 35298414.0,
      "step": 16105
    },
    {
      "entropy": 5.420316934585571,
      "epoch": 1.5988487495037713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004750030286918082,
      "loss": 5.2079,
      "mean_token_accuracy": 0.1636727675795555,
      "num_tokens": 35310629.0,
      "step": 16110
    },
    {
      "entropy": 5.494815874099731,
      "epoch": 1.5993449781659388,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004749867951596086,
      "loss": 5.254,
      "mean_token_accuracy": 0.15957536101341246,
      "num_tokens": 35321810.0,
      "step": 16115
    },
    {
      "entropy": 5.4291435241699215,
      "epoch": 1.5998412068281063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047497055666816736,
      "loss": 5.1504,
      "mean_token_accuracy": 0.16900261640548705,
      "num_tokens": 35333517.0,
      "step": 16120
    },
    {
      "entropy": 5.372884941101074,
      "epoch": 1.600337435490274,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004749543132178873,
      "loss": 5.0992,
      "mean_token_accuracy": 0.16981023401021958,
      "num_tokens": 35344170.0,
      "step": 16125
    },
    {
      "entropy": 5.455666255950928,
      "epoch": 1.6008336641524414,
      "grad_norm": 1.03125,
      "learning_rate": 0.000474938064809171,
      "loss": 5.2957,
      "mean_token_accuracy": 0.15465985387563705,
      "num_tokens": 35354963.0,
      "step": 16130
    },
    {
      "entropy": 5.453388023376465,
      "epoch": 1.6013298928146091,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004749218114424216,
      "loss": 5.2628,
      "mean_token_accuracy": 0.16259667426347732,
      "num_tokens": 35365716.0,
      "step": 16135
    },
    {
      "entropy": 5.42831654548645,
      "epoch": 1.6018261214767766,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047490555311804206,
      "loss": 5.1719,
      "mean_token_accuracy": 0.16351140141487122,
      "num_tokens": 35376113.0,
      "step": 16140
    },
    {
      "entropy": 5.3518537998199465,
      "epoch": 1.602322350138944,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047488928983643565,
      "loss": 5.1364,
      "mean_token_accuracy": 0.16477736085653305,
      "num_tokens": 35386178.0,
      "step": 16145
    },
    {
      "entropy": 5.354482889175415,
      "epoch": 1.6028185788011116,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047487302159800545,
      "loss": 5.0731,
      "mean_token_accuracy": 0.16901521533727645,
      "num_tokens": 35396874.0,
      "step": 16150
    },
    {
      "entropy": 5.422810888290405,
      "epoch": 1.603314807463279,
      "grad_norm": 1.03125,
      "learning_rate": 0.000474856748403155,
      "loss": 5.2889,
      "mean_token_accuracy": 0.15612497329711914,
      "num_tokens": 35409405.0,
      "step": 16155
    },
    {
      "entropy": 5.46998438835144,
      "epoch": 1.6038110361254465,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004748404702522879,
      "loss": 5.221,
      "mean_token_accuracy": 0.16292105466127396,
      "num_tokens": 35421449.0,
      "step": 16160
    },
    {
      "entropy": 5.41917405128479,
      "epoch": 1.604307264787614,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047482418714580774,
      "loss": 5.2709,
      "mean_token_accuracy": 0.15861179828643798,
      "num_tokens": 35434612.0,
      "step": 16165
    },
    {
      "entropy": 5.4153862476348875,
      "epoch": 1.6048034934497817,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004748078990841183,
      "loss": 5.1678,
      "mean_token_accuracy": 0.16966065019369125,
      "num_tokens": 35446016.0,
      "step": 16170
    },
    {
      "entropy": 5.527661657333374,
      "epoch": 1.6052997221119492,
      "grad_norm": 1.125,
      "learning_rate": 0.0004747916060676235,
      "loss": 5.2596,
      "mean_token_accuracy": 0.16141952723264694,
      "num_tokens": 35457405.0,
      "step": 16175
    },
    {
      "entropy": 5.438045072555542,
      "epoch": 1.6057959507741169,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047477530809672747,
      "loss": 5.2816,
      "mean_token_accuracy": 0.16160221695899962,
      "num_tokens": 35468361.0,
      "step": 16180
    },
    {
      "entropy": 5.418409395217895,
      "epoch": 1.6062921794362843,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004747590051718342,
      "loss": 5.1577,
      "mean_token_accuracy": 0.1630768209695816,
      "num_tokens": 35480092.0,
      "step": 16185
    },
    {
      "entropy": 5.445323181152344,
      "epoch": 1.6067884080984518,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047474269729334806,
      "loss": 5.2722,
      "mean_token_accuracy": 0.15766603797674178,
      "num_tokens": 35492131.0,
      "step": 16190
    },
    {
      "entropy": 5.490629243850708,
      "epoch": 1.6072846367606193,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047472638446167346,
      "loss": 5.227,
      "mean_token_accuracy": 0.16195226460695267,
      "num_tokens": 35503600.0,
      "step": 16195
    },
    {
      "entropy": 5.494294309616089,
      "epoch": 1.6077808654227868,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047471006667721484,
      "loss": 5.2803,
      "mean_token_accuracy": 0.1563917189836502,
      "num_tokens": 35514716.0,
      "step": 16200
    },
    {
      "entropy": 5.411047649383545,
      "epoch": 1.6082770940849542,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000474693743940377,
      "loss": 5.1952,
      "mean_token_accuracy": 0.16390678137540818,
      "num_tokens": 35525688.0,
      "step": 16205
    },
    {
      "entropy": 5.389076042175293,
      "epoch": 1.608773322747122,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047467741625156453,
      "loss": 5.2043,
      "mean_token_accuracy": 0.16315472722053528,
      "num_tokens": 35536391.0,
      "step": 16210
    },
    {
      "entropy": 5.386904239654541,
      "epoch": 1.6092695514092894,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047466108361118235,
      "loss": 5.1342,
      "mean_token_accuracy": 0.16991463154554368,
      "num_tokens": 35547016.0,
      "step": 16215
    },
    {
      "entropy": 5.448830318450928,
      "epoch": 1.6097657800714569,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047464474601963564,
      "loss": 5.2137,
      "mean_token_accuracy": 0.16141622066497802,
      "num_tokens": 35558402.0,
      "step": 16220
    },
    {
      "entropy": 5.453782367706299,
      "epoch": 1.6102620087336246,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004746284034773294,
      "loss": 5.2238,
      "mean_token_accuracy": 0.15994890108704568,
      "num_tokens": 35569779.0,
      "step": 16225
    },
    {
      "entropy": 5.439876461029053,
      "epoch": 1.610758237395792,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004746120559846689,
      "loss": 5.2289,
      "mean_token_accuracy": 0.1617952987551689,
      "num_tokens": 35580044.0,
      "step": 16230
    },
    {
      "entropy": 5.49529447555542,
      "epoch": 1.6112544660579595,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047459570354205954,
      "loss": 5.3042,
      "mean_token_accuracy": 0.1656623214483261,
      "num_tokens": 35591216.0,
      "step": 16235
    },
    {
      "entropy": 5.401050758361817,
      "epoch": 1.611750694720127,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047457934614990684,
      "loss": 5.124,
      "mean_token_accuracy": 0.16580240279436112,
      "num_tokens": 35602376.0,
      "step": 16240
    },
    {
      "entropy": 5.392791748046875,
      "epoch": 1.6122469233822945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004745629838086164,
      "loss": 5.1077,
      "mean_token_accuracy": 0.16858224123716353,
      "num_tokens": 35612184.0,
      "step": 16245
    },
    {
      "entropy": 5.430430603027344,
      "epoch": 1.612743152044462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047454661651859394,
      "loss": 5.2854,
      "mean_token_accuracy": 0.16054305136203767,
      "num_tokens": 35624864.0,
      "step": 16250
    },
    {
      "entropy": 5.462211561203003,
      "epoch": 1.6132393807066296,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004745302442802454,
      "loss": 5.2192,
      "mean_token_accuracy": 0.16479989588260652,
      "num_tokens": 35636492.0,
      "step": 16255
    },
    {
      "entropy": 5.394619989395141,
      "epoch": 1.6137356093687971,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004745138670939768,
      "loss": 5.1412,
      "mean_token_accuracy": 0.16660680770874023,
      "num_tokens": 35646831.0,
      "step": 16260
    },
    {
      "entropy": 5.461142492294312,
      "epoch": 1.6142318380309648,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004744974849601941,
      "loss": 5.2569,
      "mean_token_accuracy": 0.16264424324035645,
      "num_tokens": 35657703.0,
      "step": 16265
    },
    {
      "entropy": 5.293202018737793,
      "epoch": 1.6147280666931323,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004744810978793038,
      "loss": 5.0683,
      "mean_token_accuracy": 0.16895448565483093,
      "num_tokens": 35668334.0,
      "step": 16270
    },
    {
      "entropy": 5.390244579315185,
      "epoch": 1.6152242953552998,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047446470585171206,
      "loss": 5.2295,
      "mean_token_accuracy": 0.1606098085641861,
      "num_tokens": 35679796.0,
      "step": 16275
    },
    {
      "entropy": 5.41858139038086,
      "epoch": 1.6157205240174672,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047444830887782537,
      "loss": 5.1414,
      "mean_token_accuracy": 0.16069540083408357,
      "num_tokens": 35690825.0,
      "step": 16280
    },
    {
      "entropy": 5.524032306671143,
      "epoch": 1.6162167526796347,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004744319069580504,
      "loss": 5.2452,
      "mean_token_accuracy": 0.15706901699304582,
      "num_tokens": 35701824.0,
      "step": 16285
    },
    {
      "entropy": 5.458008098602295,
      "epoch": 1.6167129813418022,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047441550009279395,
      "loss": 5.1773,
      "mean_token_accuracy": 0.16412523984909058,
      "num_tokens": 35713250.0,
      "step": 16290
    },
    {
      "entropy": 5.3815819263458256,
      "epoch": 1.6172092100039697,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004743990882824627,
      "loss": 5.2386,
      "mean_token_accuracy": 0.1640092611312866,
      "num_tokens": 35725049.0,
      "step": 16295
    },
    {
      "entropy": 5.379100131988525,
      "epoch": 1.6177054386661374,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047438267152746376,
      "loss": 5.1649,
      "mean_token_accuracy": 0.16193425953388213,
      "num_tokens": 35734676.0,
      "step": 16300
    },
    {
      "entropy": 5.442473268508911,
      "epoch": 1.6182016673283048,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047436624982820423,
      "loss": 5.1786,
      "mean_token_accuracy": 0.1620963767170906,
      "num_tokens": 35743953.0,
      "step": 16305
    },
    {
      "entropy": 5.357157039642334,
      "epoch": 1.6186978959904725,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004743498231850913,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17033961415290833,
      "num_tokens": 35753149.0,
      "step": 16310
    },
    {
      "entropy": 5.369758462905883,
      "epoch": 1.61919412465264,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004743333915985323,
      "loss": 5.1895,
      "mean_token_accuracy": 0.16290192008018495,
      "num_tokens": 35763783.0,
      "step": 16315
    },
    {
      "entropy": 5.344868087768555,
      "epoch": 1.6196903533148075,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047431695506893476,
      "loss": 5.1411,
      "mean_token_accuracy": 0.17065437585115434,
      "num_tokens": 35774663.0,
      "step": 16320
    },
    {
      "entropy": 5.398631048202515,
      "epoch": 1.620186581976975,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047430051359670616,
      "loss": 5.284,
      "mean_token_accuracy": 0.15673630088567733,
      "num_tokens": 35786990.0,
      "step": 16325
    },
    {
      "entropy": 5.542551422119141,
      "epoch": 1.6206828106391424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004742840671822542,
      "loss": 5.2803,
      "mean_token_accuracy": 0.16352447420358657,
      "num_tokens": 35798818.0,
      "step": 16330
    },
    {
      "entropy": 5.504353475570679,
      "epoch": 1.62117903930131,
      "grad_norm": 1.171875,
      "learning_rate": 0.000474267615825987,
      "loss": 5.2244,
      "mean_token_accuracy": 0.16336674392223358,
      "num_tokens": 35808813.0,
      "step": 16335
    },
    {
      "entropy": 5.42971625328064,
      "epoch": 1.6216752679634776,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047425115952831213,
      "loss": 5.195,
      "mean_token_accuracy": 0.16537172496318817,
      "num_tokens": 35818664.0,
      "step": 16340
    },
    {
      "entropy": 5.345673513412476,
      "epoch": 1.622171496625645,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000474234698289638,
      "loss": 5.1422,
      "mean_token_accuracy": 0.17082065641880034,
      "num_tokens": 35828975.0,
      "step": 16345
    },
    {
      "entropy": 5.3451436996459964,
      "epoch": 1.6226677252878128,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004742182321103725,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1685357630252838,
      "num_tokens": 35839437.0,
      "step": 16350
    },
    {
      "entropy": 5.395840358734131,
      "epoch": 1.6231639539499803,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004742017609909242,
      "loss": 5.1169,
      "mean_token_accuracy": 0.1690753385424614,
      "num_tokens": 35850413.0,
      "step": 16355
    },
    {
      "entropy": 5.440779495239258,
      "epoch": 1.6236601826121477,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047418528493170157,
      "loss": 5.2119,
      "mean_token_accuracy": 0.16208771616220474,
      "num_tokens": 35862745.0,
      "step": 16360
    },
    {
      "entropy": 5.480452966690064,
      "epoch": 1.6241564112743152,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004741688039331129,
      "loss": 5.1859,
      "mean_token_accuracy": 0.1596340924501419,
      "num_tokens": 35873934.0,
      "step": 16365
    },
    {
      "entropy": 5.371326875686646,
      "epoch": 1.6246526399364827,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047415231799556725,
      "loss": 5.1146,
      "mean_token_accuracy": 0.16962513327598572,
      "num_tokens": 35884279.0,
      "step": 16370
    },
    {
      "entropy": 5.333006238937378,
      "epoch": 1.6251488685986502,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047413582711947305,
      "loss": 5.1693,
      "mean_token_accuracy": 0.1682824045419693,
      "num_tokens": 35894144.0,
      "step": 16375
    },
    {
      "entropy": 5.381010913848877,
      "epoch": 1.6256450972608176,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047411933130523954,
      "loss": 5.1617,
      "mean_token_accuracy": 0.16027027368545532,
      "num_tokens": 35903641.0,
      "step": 16380
    },
    {
      "entropy": 5.382118844985962,
      "epoch": 1.6261413259229853,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004741028305532757,
      "loss": 5.1378,
      "mean_token_accuracy": 0.16411058604717255,
      "num_tokens": 35913634.0,
      "step": 16385
    },
    {
      "entropy": 5.436276483535766,
      "epoch": 1.6266375545851528,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004740863248639907,
      "loss": 5.1751,
      "mean_token_accuracy": 0.16867901086807252,
      "num_tokens": 35923106.0,
      "step": 16390
    },
    {
      "entropy": 5.459074354171753,
      "epoch": 1.6271337832473205,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047406981423779376,
      "loss": 5.177,
      "mean_token_accuracy": 0.1626700848340988,
      "num_tokens": 35933709.0,
      "step": 16395
    },
    {
      "entropy": 5.377651596069336,
      "epoch": 1.627630011909488,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047405329867509447,
      "loss": 5.1928,
      "mean_token_accuracy": 0.16006845086812974,
      "num_tokens": 35944734.0,
      "step": 16400
    },
    {
      "entropy": 5.374727630615235,
      "epoch": 1.6281262405716554,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047403677817630226,
      "loss": 5.1568,
      "mean_token_accuracy": 0.16647227108478546,
      "num_tokens": 35956513.0,
      "step": 16405
    },
    {
      "entropy": 5.361451721191406,
      "epoch": 1.628622469233823,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004740202527418269,
      "loss": 5.1861,
      "mean_token_accuracy": 0.16157235205173492,
      "num_tokens": 35967096.0,
      "step": 16410
    },
    {
      "entropy": 5.44601993560791,
      "epoch": 1.6291186978959904,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000474003722372078,
      "loss": 5.2465,
      "mean_token_accuracy": 0.16381874084472656,
      "num_tokens": 35979010.0,
      "step": 16415
    },
    {
      "entropy": 5.386987161636353,
      "epoch": 1.6296149265581579,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047398718706746556,
      "loss": 5.1297,
      "mean_token_accuracy": 0.165200836956501,
      "num_tokens": 35989253.0,
      "step": 16420
    },
    {
      "entropy": 5.357050895690918,
      "epoch": 1.6301111552203256,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004739706468283998,
      "loss": 5.111,
      "mean_token_accuracy": 0.1680722638964653,
      "num_tokens": 35999659.0,
      "step": 16425
    },
    {
      "entropy": 5.378021144866944,
      "epoch": 1.630607383882493,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047395410165529075,
      "loss": 5.1825,
      "mean_token_accuracy": 0.16301301568746568,
      "num_tokens": 36010295.0,
      "step": 16430
    },
    {
      "entropy": 5.4220202445983885,
      "epoch": 1.6311036125446607,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047393755154854855,
      "loss": 5.2138,
      "mean_token_accuracy": 0.16122177392244338,
      "num_tokens": 36021575.0,
      "step": 16435
    },
    {
      "entropy": 5.463352012634277,
      "epoch": 1.6315998412068282,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004739209965085839,
      "loss": 5.2093,
      "mean_token_accuracy": 0.15525000393390656,
      "num_tokens": 36032764.0,
      "step": 16440
    },
    {
      "entropy": 5.448444461822509,
      "epoch": 1.6320960698689957,
      "grad_norm": 1.109375,
      "learning_rate": 0.000473904436535807,
      "loss": 5.1536,
      "mean_token_accuracy": 0.1715414211153984,
      "num_tokens": 36041864.0,
      "step": 16445
    },
    {
      "entropy": 5.2945116519927975,
      "epoch": 1.6325922985311632,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004738878716306288,
      "loss": 5.1452,
      "mean_token_accuracy": 0.16269672214984893,
      "num_tokens": 36051321.0,
      "step": 16450
    },
    {
      "entropy": 5.344901704788208,
      "epoch": 1.6330885271933306,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047387130179345996,
      "loss": 5.1251,
      "mean_token_accuracy": 0.16227182298898696,
      "num_tokens": 36061448.0,
      "step": 16455
    },
    {
      "entropy": 5.441643190383911,
      "epoch": 1.6335847558554981,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047385472702471133,
      "loss": 5.2461,
      "mean_token_accuracy": 0.16513296365737914,
      "num_tokens": 36071832.0,
      "step": 16460
    },
    {
      "entropy": 5.397855663299561,
      "epoch": 1.6340809845176656,
      "grad_norm": 1.046875,
      "learning_rate": 0.000473838147324794,
      "loss": 5.1522,
      "mean_token_accuracy": 0.16204911917448045,
      "num_tokens": 36082103.0,
      "step": 16465
    },
    {
      "entropy": 5.461338138580322,
      "epoch": 1.6345772131798333,
      "grad_norm": 1.125,
      "learning_rate": 0.000473821562694119,
      "loss": 5.2162,
      "mean_token_accuracy": 0.1584436148405075,
      "num_tokens": 36092063.0,
      "step": 16470
    },
    {
      "entropy": 5.437670469284058,
      "epoch": 1.6350734418420008,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004738049731330977,
      "loss": 5.1976,
      "mean_token_accuracy": 0.1577135682106018,
      "num_tokens": 36102573.0,
      "step": 16475
    },
    {
      "entropy": 5.444153642654419,
      "epoch": 1.6355696705041685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004737883786421414,
      "loss": 5.174,
      "mean_token_accuracy": 0.16354111433029175,
      "num_tokens": 36112677.0,
      "step": 16480
    },
    {
      "entropy": 5.410617923736572,
      "epoch": 1.636065899166336,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047377177922166175,
      "loss": 5.2356,
      "mean_token_accuracy": 0.16140222996473313,
      "num_tokens": 36123337.0,
      "step": 16485
    },
    {
      "entropy": 5.440532970428467,
      "epoch": 1.6365621278285034,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047375517487207024,
      "loss": 5.2127,
      "mean_token_accuracy": 0.16278964132070542,
      "num_tokens": 36134503.0,
      "step": 16490
    },
    {
      "entropy": 5.529685831069946,
      "epoch": 1.6370583564906709,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047373856559377876,
      "loss": 5.2673,
      "mean_token_accuracy": 0.16154926866292954,
      "num_tokens": 36145766.0,
      "step": 16495
    },
    {
      "entropy": 5.461714792251587,
      "epoch": 1.6375545851528384,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004737219513871989,
      "loss": 5.2153,
      "mean_token_accuracy": 0.16148438155651093,
      "num_tokens": 36156584.0,
      "step": 16500
    },
    {
      "entropy": 5.375883722305298,
      "epoch": 1.6380508138150058,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047370533225274305,
      "loss": 5.1916,
      "mean_token_accuracy": 0.1647510215640068,
      "num_tokens": 36166999.0,
      "step": 16505
    },
    {
      "entropy": 5.428827571868896,
      "epoch": 1.6385470424771735,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004736887081908231,
      "loss": 5.1789,
      "mean_token_accuracy": 0.16813623458147048,
      "num_tokens": 36178724.0,
      "step": 16510
    },
    {
      "entropy": 5.493850803375244,
      "epoch": 1.639043271139341,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004736720792018513,
      "loss": 5.2669,
      "mean_token_accuracy": 0.16175237596035003,
      "num_tokens": 36188912.0,
      "step": 16515
    },
    {
      "entropy": 5.40254020690918,
      "epoch": 1.6395394998015085,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004736554452862399,
      "loss": 5.1571,
      "mean_token_accuracy": 0.16721909642219543,
      "num_tokens": 36198525.0,
      "step": 16520
    },
    {
      "entropy": 5.392537355422974,
      "epoch": 1.6400357284636762,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047363880644440173,
      "loss": 5.1219,
      "mean_token_accuracy": 0.16717253774404525,
      "num_tokens": 36209384.0,
      "step": 16525
    },
    {
      "entropy": 5.441282844543457,
      "epoch": 1.6405319571258437,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047362216267674906,
      "loss": 5.2219,
      "mean_token_accuracy": 0.16171573847532272,
      "num_tokens": 36220597.0,
      "step": 16530
    },
    {
      "entropy": 5.473483657836914,
      "epoch": 1.6410281857880111,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004736055139836948,
      "loss": 5.2594,
      "mean_token_accuracy": 0.16234485059976578,
      "num_tokens": 36230317.0,
      "step": 16535
    },
    {
      "entropy": 5.54504976272583,
      "epoch": 1.6415244144501786,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047358886036565176,
      "loss": 5.3482,
      "mean_token_accuracy": 0.16090625822544097,
      "num_tokens": 36241645.0,
      "step": 16540
    },
    {
      "entropy": 5.454131174087524,
      "epoch": 1.642020643112346,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047357220182303285,
      "loss": 5.2763,
      "mean_token_accuracy": 0.15622056424617767,
      "num_tokens": 36253098.0,
      "step": 16545
    },
    {
      "entropy": 5.386646127700805,
      "epoch": 1.6425168717745136,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004735555383562513,
      "loss": 5.1781,
      "mean_token_accuracy": 0.16684114038944245,
      "num_tokens": 36264417.0,
      "step": 16550
    },
    {
      "entropy": 5.512826681137085,
      "epoch": 1.6430131004366813,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047353886996572014,
      "loss": 5.2715,
      "mean_token_accuracy": 0.15724946260452272,
      "num_tokens": 36276589.0,
      "step": 16555
    },
    {
      "entropy": 5.421932744979858,
      "epoch": 1.6435093290988487,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047352219665185293,
      "loss": 5.1895,
      "mean_token_accuracy": 0.16185040026903152,
      "num_tokens": 36287308.0,
      "step": 16560
    },
    {
      "entropy": 5.375375270843506,
      "epoch": 1.6440055577610164,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004735055184150629,
      "loss": 5.1994,
      "mean_token_accuracy": 0.16407883465290068,
      "num_tokens": 36298966.0,
      "step": 16565
    },
    {
      "entropy": 5.412043285369873,
      "epoch": 1.644501786423184,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047348883525576385,
      "loss": 5.2101,
      "mean_token_accuracy": 0.16135587990283967,
      "num_tokens": 36310498.0,
      "step": 16570
    },
    {
      "entropy": 5.474941396713257,
      "epoch": 1.6449980150853514,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004734721471743694,
      "loss": 5.264,
      "mean_token_accuracy": 0.16176439821720123,
      "num_tokens": 36321589.0,
      "step": 16575
    },
    {
      "entropy": 5.421678066253662,
      "epoch": 1.6454942437475188,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004734554541712933,
      "loss": 5.1881,
      "mean_token_accuracy": 0.15752718448638917,
      "num_tokens": 36333041.0,
      "step": 16580
    },
    {
      "entropy": 5.475110244750977,
      "epoch": 1.6459904724096863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004734387562469496,
      "loss": 5.2327,
      "mean_token_accuracy": 0.15645822137594223,
      "num_tokens": 36344225.0,
      "step": 16585
    },
    {
      "entropy": 5.431123208999634,
      "epoch": 1.6464867010718538,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047342205340175235,
      "loss": 5.1174,
      "mean_token_accuracy": 0.16633806228637696,
      "num_tokens": 36355524.0,
      "step": 16590
    },
    {
      "entropy": 5.392293691635132,
      "epoch": 1.6469829297340213,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047340534563611574,
      "loss": 5.1917,
      "mean_token_accuracy": 0.16680285483598709,
      "num_tokens": 36366495.0,
      "step": 16595
    },
    {
      "entropy": 5.396942567825318,
      "epoch": 1.647479158396189,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047338863295045414,
      "loss": 5.2034,
      "mean_token_accuracy": 0.16286242604255677,
      "num_tokens": 36378316.0,
      "step": 16600
    },
    {
      "entropy": 5.486587381362915,
      "epoch": 1.6479753870583564,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047337191534518186,
      "loss": 5.2742,
      "mean_token_accuracy": 0.15607527792453765,
      "num_tokens": 36390500.0,
      "step": 16605
    },
    {
      "entropy": 5.514648151397705,
      "epoch": 1.6484716157205241,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004733551928207135,
      "loss": 5.2169,
      "mean_token_accuracy": 0.1658637762069702,
      "num_tokens": 36403831.0,
      "step": 16610
    },
    {
      "entropy": 5.428684997558594,
      "epoch": 1.6489678443826916,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004733384653774639,
      "loss": 5.1794,
      "mean_token_accuracy": 0.16267184913158417,
      "num_tokens": 36414896.0,
      "step": 16615
    },
    {
      "entropy": 5.40056209564209,
      "epoch": 1.649464073044859,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047332173301584764,
      "loss": 5.1149,
      "mean_token_accuracy": 0.1660406619310379,
      "num_tokens": 36426260.0,
      "step": 16620
    },
    {
      "entropy": 5.349297714233399,
      "epoch": 1.6499603017070266,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047330499573627975,
      "loss": 5.0747,
      "mean_token_accuracy": 0.17279173582792282,
      "num_tokens": 36437125.0,
      "step": 16625
    },
    {
      "entropy": 5.39158296585083,
      "epoch": 1.650456530369194,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047328825353917526,
      "loss": 5.1199,
      "mean_token_accuracy": 0.1685706913471222,
      "num_tokens": 36447507.0,
      "step": 16630
    },
    {
      "entropy": 5.460553216934204,
      "epoch": 1.6509527590313615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047327150642494944,
      "loss": 5.3086,
      "mean_token_accuracy": 0.16450339406728745,
      "num_tokens": 36459134.0,
      "step": 16635
    },
    {
      "entropy": 5.4703747749328615,
      "epoch": 1.6514489876935292,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004732547543940174,
      "loss": 5.2151,
      "mean_token_accuracy": 0.1600191593170166,
      "num_tokens": 36469594.0,
      "step": 16640
    },
    {
      "entropy": 5.449242353439331,
      "epoch": 1.6519452163556967,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004732379974467947,
      "loss": 5.2169,
      "mean_token_accuracy": 0.16647123992443086,
      "num_tokens": 36480477.0,
      "step": 16645
    },
    {
      "entropy": 5.392209911346436,
      "epoch": 1.6524414450178644,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004732212355836968,
      "loss": 5.2027,
      "mean_token_accuracy": 0.1627037614583969,
      "num_tokens": 36491304.0,
      "step": 16650
    },
    {
      "entropy": 5.502873802185059,
      "epoch": 1.6529376736800319,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004732044688051395,
      "loss": 5.2835,
      "mean_token_accuracy": 0.15526345521211624,
      "num_tokens": 36502030.0,
      "step": 16655
    },
    {
      "entropy": 5.47953839302063,
      "epoch": 1.6534339023421993,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047318769711153833,
      "loss": 5.3018,
      "mean_token_accuracy": 0.15728846490383147,
      "num_tokens": 36514405.0,
      "step": 16660
    },
    {
      "entropy": 5.466570711135864,
      "epoch": 1.6539301310043668,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004731709205033094,
      "loss": 5.1504,
      "mean_token_accuracy": 0.15970703959465027,
      "num_tokens": 36524626.0,
      "step": 16665
    },
    {
      "entropy": 5.454094076156617,
      "epoch": 1.6544263596665343,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047315413898086863,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1706256940960884,
      "num_tokens": 36535340.0,
      "step": 16670
    },
    {
      "entropy": 5.468825912475586,
      "epoch": 1.6549225883287018,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004731373525446322,
      "loss": 5.1746,
      "mean_token_accuracy": 0.16439798027276992,
      "num_tokens": 36547189.0,
      "step": 16675
    },
    {
      "entropy": 5.42506628036499,
      "epoch": 1.6554188169908692,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004731205611950164,
      "loss": 5.2556,
      "mean_token_accuracy": 0.16359982937574385,
      "num_tokens": 36558177.0,
      "step": 16680
    },
    {
      "entropy": 5.378905487060547,
      "epoch": 1.655915045653037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047310376493243754,
      "loss": 5.0677,
      "mean_token_accuracy": 0.17196686714887618,
      "num_tokens": 36569559.0,
      "step": 16685
    },
    {
      "entropy": 5.418493223190308,
      "epoch": 1.6564112743152044,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004730869637573122,
      "loss": 5.1637,
      "mean_token_accuracy": 0.16660676151514053,
      "num_tokens": 36579617.0,
      "step": 16690
    },
    {
      "entropy": 5.377469062805176,
      "epoch": 1.656907502977372,
      "grad_norm": 1.125,
      "learning_rate": 0.00047307015767005705,
      "loss": 5.1435,
      "mean_token_accuracy": 0.1667794719338417,
      "num_tokens": 36590427.0,
      "step": 16695
    },
    {
      "entropy": 5.404322290420533,
      "epoch": 1.6574037316395396,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004730533466710888,
      "loss": 5.1392,
      "mean_token_accuracy": 0.16910175383090972,
      "num_tokens": 36602313.0,
      "step": 16700
    },
    {
      "entropy": 5.5056471824646,
      "epoch": 1.657899960301707,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004730365307608242,
      "loss": 5.2326,
      "mean_token_accuracy": 0.16425890773534774,
      "num_tokens": 36615191.0,
      "step": 16705
    },
    {
      "entropy": 5.37666425704956,
      "epoch": 1.6583961889638745,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047301970993968054,
      "loss": 5.1529,
      "mean_token_accuracy": 0.17499279528856276,
      "num_tokens": 36624692.0,
      "step": 16710
    },
    {
      "entropy": 5.416981220245361,
      "epoch": 1.658892417626042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047300288420807463,
      "loss": 5.2235,
      "mean_token_accuracy": 0.1565033331513405,
      "num_tokens": 36635961.0,
      "step": 16715
    },
    {
      "entropy": 5.379738426208496,
      "epoch": 1.6593886462882095,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047298605356642387,
      "loss": 5.1917,
      "mean_token_accuracy": 0.16286439448595047,
      "num_tokens": 36646710.0,
      "step": 16720
    },
    {
      "entropy": 5.429356384277344,
      "epoch": 1.6598848749503772,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047296921801514565,
      "loss": 5.2799,
      "mean_token_accuracy": 0.1615061119198799,
      "num_tokens": 36657168.0,
      "step": 16725
    },
    {
      "entropy": 5.483791446685791,
      "epoch": 1.6603811036125447,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004729523775546574,
      "loss": 5.2418,
      "mean_token_accuracy": 0.16363940238952637,
      "num_tokens": 36668428.0,
      "step": 16730
    },
    {
      "entropy": 5.454153776168823,
      "epoch": 1.6608773322747123,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004729355321853767,
      "loss": 5.2089,
      "mean_token_accuracy": 0.1679923191666603,
      "num_tokens": 36679361.0,
      "step": 16735
    },
    {
      "entropy": 5.492412900924682,
      "epoch": 1.6613735609368798,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004729186819077214,
      "loss": 5.2585,
      "mean_token_accuracy": 0.15860113650560378,
      "num_tokens": 36689480.0,
      "step": 16740
    },
    {
      "entropy": 5.521343040466308,
      "epoch": 1.6618697895990473,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004729018267221092,
      "loss": 5.3049,
      "mean_token_accuracy": 0.16136780977249146,
      "num_tokens": 36700420.0,
      "step": 16745
    },
    {
      "entropy": 5.483749151229858,
      "epoch": 1.6623660182612148,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004728849666289581,
      "loss": 5.2495,
      "mean_token_accuracy": 0.16050813645124434,
      "num_tokens": 36710106.0,
      "step": 16750
    },
    {
      "entropy": 5.442740535736084,
      "epoch": 1.6628622469233822,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047286810162868625,
      "loss": 5.1184,
      "mean_token_accuracy": 0.16795748472213745,
      "num_tokens": 36719630.0,
      "step": 16755
    },
    {
      "entropy": 5.351249551773071,
      "epoch": 1.6633584755855497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004728512317217119,
      "loss": 5.0938,
      "mean_token_accuracy": 0.16857063472270967,
      "num_tokens": 36729379.0,
      "step": 16760
    },
    {
      "entropy": 5.363901138305664,
      "epoch": 1.6638547042477172,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047283435690845336,
      "loss": 5.1533,
      "mean_token_accuracy": 0.1687484472990036,
      "num_tokens": 36740103.0,
      "step": 16765
    },
    {
      "entropy": 5.433605718612671,
      "epoch": 1.664350932909885,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047281747718932906,
      "loss": 5.2045,
      "mean_token_accuracy": 0.16792607307434082,
      "num_tokens": 36750928.0,
      "step": 16770
    },
    {
      "entropy": 5.403333282470703,
      "epoch": 1.6648471615720524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004728005925647575,
      "loss": 5.196,
      "mean_token_accuracy": 0.16072066873311996,
      "num_tokens": 36761415.0,
      "step": 16775
    },
    {
      "entropy": 5.352472352981567,
      "epoch": 1.66534339023422,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047278370303515744,
      "loss": 5.2206,
      "mean_token_accuracy": 0.1599453866481781,
      "num_tokens": 36772560.0,
      "step": 16780
    },
    {
      "entropy": 5.404735994338989,
      "epoch": 1.6658396188963875,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004727668086009478,
      "loss": 5.2371,
      "mean_token_accuracy": 0.16436617970466613,
      "num_tokens": 36783597.0,
      "step": 16785
    },
    {
      "entropy": 5.518607187271118,
      "epoch": 1.666335847558555,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047274990926254753,
      "loss": 5.3121,
      "mean_token_accuracy": 0.1561573013663292,
      "num_tokens": 36794219.0,
      "step": 16790
    },
    {
      "entropy": 5.403351926803589,
      "epoch": 1.6668320762207225,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004727330050203755,
      "loss": 5.1287,
      "mean_token_accuracy": 0.16844879537820817,
      "num_tokens": 36804156.0,
      "step": 16795
    },
    {
      "entropy": 5.356365299224853,
      "epoch": 1.66732830488289,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004727160958748511,
      "loss": 5.152,
      "mean_token_accuracy": 0.1690562143921852,
      "num_tokens": 36814414.0,
      "step": 16800
    },
    {
      "entropy": 5.399230813980102,
      "epoch": 1.6678245335450574,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047269918182639355,
      "loss": 5.1198,
      "mean_token_accuracy": 0.16304894536733627,
      "num_tokens": 36826173.0,
      "step": 16805
    },
    {
      "entropy": 5.433908987045288,
      "epoch": 1.6683207622072251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004726822628754222,
      "loss": 5.2399,
      "mean_token_accuracy": 0.16109345555305482,
      "num_tokens": 36837726.0,
      "step": 16810
    },
    {
      "entropy": 5.4375927448272705,
      "epoch": 1.6688169908693926,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047266533902235683,
      "loss": 5.1742,
      "mean_token_accuracy": 0.15874040275812148,
      "num_tokens": 36848605.0,
      "step": 16815
    },
    {
      "entropy": 5.447031116485595,
      "epoch": 1.66931321953156,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004726484102676169,
      "loss": 5.2035,
      "mean_token_accuracy": 0.16165889352560042,
      "num_tokens": 36858730.0,
      "step": 16820
    },
    {
      "entropy": 5.356701421737671,
      "epoch": 1.6698094481937278,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004726314766116223,
      "loss": 5.0624,
      "mean_token_accuracy": 0.17089584469795227,
      "num_tokens": 36868239.0,
      "step": 16825
    },
    {
      "entropy": 5.330888462066651,
      "epoch": 1.6703056768558953,
      "grad_norm": 1.234375,
      "learning_rate": 0.000472614538054793,
      "loss": 5.1555,
      "mean_token_accuracy": 0.16622113734483718,
      "num_tokens": 36878077.0,
      "step": 16830
    },
    {
      "entropy": 5.38208794593811,
      "epoch": 1.6708019055180627,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047259759459754893,
      "loss": 5.1865,
      "mean_token_accuracy": 0.1592635229229927,
      "num_tokens": 36888874.0,
      "step": 16835
    },
    {
      "entropy": 5.462038803100586,
      "epoch": 1.6712981341802302,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004725806462403103,
      "loss": 5.1964,
      "mean_token_accuracy": 0.16409918293356895,
      "num_tokens": 36898584.0,
      "step": 16840
    },
    {
      "entropy": 5.449842214584351,
      "epoch": 1.6717943628423977,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047256369298349736,
      "loss": 5.193,
      "mean_token_accuracy": 0.16655738204717635,
      "num_tokens": 36908662.0,
      "step": 16845
    },
    {
      "entropy": 5.3673053741455075,
      "epoch": 1.6722905915045652,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047254673482753065,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1649319812655449,
      "num_tokens": 36919482.0,
      "step": 16850
    },
    {
      "entropy": 5.391136360168457,
      "epoch": 1.6727868201667329,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047252977177283045,
      "loss": 5.1677,
      "mean_token_accuracy": 0.15829679518938064,
      "num_tokens": 36930470.0,
      "step": 16855
    },
    {
      "entropy": 5.433168506622314,
      "epoch": 1.6732830488289003,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047251280381981764,
      "loss": 5.1512,
      "mean_token_accuracy": 0.16839530169963837,
      "num_tokens": 36942837.0,
      "step": 16860
    },
    {
      "entropy": 5.477422666549683,
      "epoch": 1.673779277491068,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004724958309689129,
      "loss": 5.2429,
      "mean_token_accuracy": 0.15767752826213838,
      "num_tokens": 36952994.0,
      "step": 16865
    },
    {
      "entropy": 5.398792743682861,
      "epoch": 1.6742755061532355,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047247885322053705,
      "loss": 5.2106,
      "mean_token_accuracy": 0.16445887684822083,
      "num_tokens": 36963973.0,
      "step": 16870
    },
    {
      "entropy": 5.359536504745483,
      "epoch": 1.674771734815403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047246187057511127,
      "loss": 5.1434,
      "mean_token_accuracy": 0.17217618674039842,
      "num_tokens": 36975897.0,
      "step": 16875
    },
    {
      "entropy": 5.436277246475219,
      "epoch": 1.6752679634775705,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004724448830330565,
      "loss": 5.1705,
      "mean_token_accuracy": 0.16373747289180757,
      "num_tokens": 36986481.0,
      "step": 16880
    },
    {
      "entropy": 5.417337417602539,
      "epoch": 1.675764192139738,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047242789059479413,
      "loss": 5.2557,
      "mean_token_accuracy": 0.1584458902478218,
      "num_tokens": 36997863.0,
      "step": 16885
    },
    {
      "entropy": 5.448734712600708,
      "epoch": 1.6762604208019054,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047241089326074547,
      "loss": 5.2037,
      "mean_token_accuracy": 0.1591024786233902,
      "num_tokens": 37009294.0,
      "step": 16890
    },
    {
      "entropy": 5.458444833755493,
      "epoch": 1.6767566494640729,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004723938910313321,
      "loss": 5.1754,
      "mean_token_accuracy": 0.16689318865537645,
      "num_tokens": 37019748.0,
      "step": 16895
    },
    {
      "entropy": 5.449307680130005,
      "epoch": 1.6772528781262406,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047237688390697555,
      "loss": 5.2395,
      "mean_token_accuracy": 0.1666789874434471,
      "num_tokens": 37030903.0,
      "step": 16900
    },
    {
      "entropy": 5.4349593162536625,
      "epoch": 1.677749106788408,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004723598718880975,
      "loss": 5.1731,
      "mean_token_accuracy": 0.16585267931222916,
      "num_tokens": 37042551.0,
      "step": 16905
    },
    {
      "entropy": 5.463727426528931,
      "epoch": 1.6782453354505757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047234285497512005,
      "loss": 5.2331,
      "mean_token_accuracy": 0.15327072739601136,
      "num_tokens": 37053151.0,
      "step": 16910
    },
    {
      "entropy": 5.4736865043640135,
      "epoch": 1.6787415641127432,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047232583316846494,
      "loss": 5.2544,
      "mean_token_accuracy": 0.1604996621608734,
      "num_tokens": 37064421.0,
      "step": 16915
    },
    {
      "entropy": 5.439113712310791,
      "epoch": 1.6792377927749107,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004723088064685544,
      "loss": 5.1706,
      "mean_token_accuracy": 0.17021820098161697,
      "num_tokens": 37074758.0,
      "step": 16920
    },
    {
      "entropy": 5.446474361419678,
      "epoch": 1.6797340214370782,
      "grad_norm": 1.125,
      "learning_rate": 0.00047229177487581055,
      "loss": 5.1336,
      "mean_token_accuracy": 0.16528764367103577,
      "num_tokens": 37084898.0,
      "step": 16925
    },
    {
      "entropy": 5.441545629501343,
      "epoch": 1.6802302500992456,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047227473839065587,
      "loss": 5.2312,
      "mean_token_accuracy": 0.15742463618516922,
      "num_tokens": 37094366.0,
      "step": 16930
    },
    {
      "entropy": 5.522211313247681,
      "epoch": 1.6807264787614131,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047225769701351284,
      "loss": 5.2781,
      "mean_token_accuracy": 0.16385298520326613,
      "num_tokens": 37104965.0,
      "step": 16935
    },
    {
      "entropy": 5.390651988983154,
      "epoch": 1.6812227074235808,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004722406507448038,
      "loss": 5.1517,
      "mean_token_accuracy": 0.1657077044248581,
      "num_tokens": 37115622.0,
      "step": 16940
    },
    {
      "entropy": 5.48987979888916,
      "epoch": 1.6817189360857483,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047222359958495175,
      "loss": 5.2366,
      "mean_token_accuracy": 0.15927164554595946,
      "num_tokens": 37127234.0,
      "step": 16945
    },
    {
      "entropy": 5.454946279525757,
      "epoch": 1.682215164747916,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004722065435343794,
      "loss": 5.1186,
      "mean_token_accuracy": 0.1667145535349846,
      "num_tokens": 37136533.0,
      "step": 16950
    },
    {
      "entropy": 5.41376314163208,
      "epoch": 1.6827113934100835,
      "grad_norm": 1.125,
      "learning_rate": 0.0004721894825935097,
      "loss": 5.2067,
      "mean_token_accuracy": 0.1639291375875473,
      "num_tokens": 37147821.0,
      "step": 16955
    },
    {
      "entropy": 5.392738151550293,
      "epoch": 1.683207622072251,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004721724167627657,
      "loss": 5.2099,
      "mean_token_accuracy": 0.16529448032379152,
      "num_tokens": 37158929.0,
      "step": 16960
    },
    {
      "entropy": 5.482515859603882,
      "epoch": 1.6837038507344184,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004721553460425708,
      "loss": 5.2409,
      "mean_token_accuracy": 0.16545132398605347,
      "num_tokens": 37169817.0,
      "step": 16965
    },
    {
      "entropy": 5.455887222290039,
      "epoch": 1.684200079396586,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000472138270433348,
      "loss": 5.3237,
      "mean_token_accuracy": 0.1562894120812416,
      "num_tokens": 37181009.0,
      "step": 16970
    },
    {
      "entropy": 5.43373384475708,
      "epoch": 1.6846963080587534,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047212118993552097,
      "loss": 5.1537,
      "mean_token_accuracy": 0.16884590238332747,
      "num_tokens": 37192110.0,
      "step": 16975
    },
    {
      "entropy": 5.434031867980957,
      "epoch": 1.6851925367209208,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004721041045495131,
      "loss": 5.1903,
      "mean_token_accuracy": 0.16753554940223694,
      "num_tokens": 37203426.0,
      "step": 16980
    },
    {
      "entropy": 5.476783370971679,
      "epoch": 1.6856887653830885,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047208701427574817,
      "loss": 5.2832,
      "mean_token_accuracy": 0.1603301152586937,
      "num_tokens": 37215025.0,
      "step": 16985
    },
    {
      "entropy": 5.553802967071533,
      "epoch": 1.686184994045256,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047206991911465,
      "loss": 5.2112,
      "mean_token_accuracy": 0.165936703979969,
      "num_tokens": 37225653.0,
      "step": 16990
    },
    {
      "entropy": 5.429096031188965,
      "epoch": 1.6866812227074237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004720528190666425,
      "loss": 5.1853,
      "mean_token_accuracy": 0.16075737476348878,
      "num_tokens": 37237282.0,
      "step": 16995
    },
    {
      "entropy": 5.3427050590515135,
      "epoch": 1.6871774513695912,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047203571413214967,
      "loss": 5.141,
      "mean_token_accuracy": 0.169723080098629,
      "num_tokens": 37248476.0,
      "step": 17000
    },
    {
      "entropy": 5.419410705566406,
      "epoch": 1.6876736800317587,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004720186043115957,
      "loss": 5.1356,
      "mean_token_accuracy": 0.16248626559972762,
      "num_tokens": 37259354.0,
      "step": 17005
    },
    {
      "entropy": 5.422538232803345,
      "epoch": 1.6881699086939261,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004720014896054048,
      "loss": 5.2121,
      "mean_token_accuracy": 0.1627548009157181,
      "num_tokens": 37269895.0,
      "step": 17010
    },
    {
      "entropy": 5.388499975204468,
      "epoch": 1.6886661373560936,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004719843700140017,
      "loss": 5.1782,
      "mean_token_accuracy": 0.16746238023042678,
      "num_tokens": 37281201.0,
      "step": 17015
    },
    {
      "entropy": 5.397493839263916,
      "epoch": 1.689162366018261,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004719672455378105,
      "loss": 5.2014,
      "mean_token_accuracy": 0.16215396374464036,
      "num_tokens": 37293362.0,
      "step": 17020
    },
    {
      "entropy": 5.499426174163818,
      "epoch": 1.6896585946804288,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000471950116177256,
      "loss": 5.2184,
      "mean_token_accuracy": 0.15977409482002258,
      "num_tokens": 37304291.0,
      "step": 17025
    },
    {
      "entropy": 5.461057996749878,
      "epoch": 1.6901548233425963,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004719329819327631,
      "loss": 5.2049,
      "mean_token_accuracy": 0.15862593650817872,
      "num_tokens": 37315608.0,
      "step": 17030
    },
    {
      "entropy": 5.401481628417969,
      "epoch": 1.690651052004764,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004719158428047566,
      "loss": 5.1377,
      "mean_token_accuracy": 0.16156528443098067,
      "num_tokens": 37327559.0,
      "step": 17035
    },
    {
      "entropy": 5.421053409576416,
      "epoch": 1.6911472806669314,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004718986987936615,
      "loss": 5.1775,
      "mean_token_accuracy": 0.16881328523159028,
      "num_tokens": 37338101.0,
      "step": 17040
    },
    {
      "entropy": 5.474651718139649,
      "epoch": 1.691643509329099,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047188154989990296,
      "loss": 5.2707,
      "mean_token_accuracy": 0.15633052438497544,
      "num_tokens": 37349124.0,
      "step": 17045
    },
    {
      "entropy": 5.371481561660767,
      "epoch": 1.6921397379912664,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004718643961239062,
      "loss": 5.1377,
      "mean_token_accuracy": 0.16675345748662948,
      "num_tokens": 37360061.0,
      "step": 17050
    },
    {
      "entropy": 5.410939931869507,
      "epoch": 1.6926359666534339,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047184723746609666,
      "loss": 5.2019,
      "mean_token_accuracy": 0.1674695372581482,
      "num_tokens": 37371210.0,
      "step": 17055
    },
    {
      "entropy": 5.439813709259033,
      "epoch": 1.6931321953156013,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047183007392689977,
      "loss": 5.1553,
      "mean_token_accuracy": 0.16335569322109222,
      "num_tokens": 37382547.0,
      "step": 17060
    },
    {
      "entropy": 5.427677774429322,
      "epoch": 1.6936284239777688,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047181290550674115,
      "loss": 5.2019,
      "mean_token_accuracy": 0.16917244344949722,
      "num_tokens": 37392311.0,
      "step": 17065
    },
    {
      "entropy": 5.446359777450562,
      "epoch": 1.6941246526399365,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004717957322060466,
      "loss": 5.161,
      "mean_token_accuracy": 0.1672020897269249,
      "num_tokens": 37402626.0,
      "step": 17070
    },
    {
      "entropy": 5.398947477340698,
      "epoch": 1.694620881302104,
      "grad_norm": 1.125,
      "learning_rate": 0.0004717785540252419,
      "loss": 5.1513,
      "mean_token_accuracy": 0.1595374971628189,
      "num_tokens": 37412100.0,
      "step": 17075
    },
    {
      "entropy": 5.342575550079346,
      "epoch": 1.6951171099642717,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004717613709647532,
      "loss": 5.0593,
      "mean_token_accuracy": 0.16762058287858964,
      "num_tokens": 37421735.0,
      "step": 17080
    },
    {
      "entropy": 5.4850959300994875,
      "epoch": 1.6956133386264391,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047174418302500635,
      "loss": 5.2708,
      "mean_token_accuracy": 0.15292758792638778,
      "num_tokens": 37432179.0,
      "step": 17085
    },
    {
      "entropy": 5.427484130859375,
      "epoch": 1.6961095672886066,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004717269902064278,
      "loss": 5.1231,
      "mean_token_accuracy": 0.1757965162396431,
      "num_tokens": 37443276.0,
      "step": 17090
    },
    {
      "entropy": 5.426189756393432,
      "epoch": 1.696605795950774,
      "grad_norm": 1.125,
      "learning_rate": 0.0004717097925094438,
      "loss": 5.2471,
      "mean_token_accuracy": 0.15737387239933015,
      "num_tokens": 37454537.0,
      "step": 17095
    },
    {
      "entropy": 5.461902809143067,
      "epoch": 1.6971020246129416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004716925899344808,
      "loss": 5.1617,
      "mean_token_accuracy": 0.16117839366197587,
      "num_tokens": 37465493.0,
      "step": 17100
    },
    {
      "entropy": 5.382326221466064,
      "epoch": 1.697598253275109,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004716753824819654,
      "loss": 5.1426,
      "mean_token_accuracy": 0.16655945181846618,
      "num_tokens": 37475511.0,
      "step": 17105
    },
    {
      "entropy": 5.356410121917724,
      "epoch": 1.6980944819372767,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004716581701523243,
      "loss": 5.0756,
      "mean_token_accuracy": 0.164541395008564,
      "num_tokens": 37487046.0,
      "step": 17110
    },
    {
      "entropy": 5.42896728515625,
      "epoch": 1.6985907105994442,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047164095294598433,
      "loss": 5.2394,
      "mean_token_accuracy": 0.1629299759864807,
      "num_tokens": 37498398.0,
      "step": 17115
    },
    {
      "entropy": 5.441703510284424,
      "epoch": 1.699086939261612,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047162373086337243,
      "loss": 5.2295,
      "mean_token_accuracy": 0.15868429243564605,
      "num_tokens": 37509397.0,
      "step": 17120
    },
    {
      "entropy": 5.42757019996643,
      "epoch": 1.6995831679237794,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047160650390491574,
      "loss": 5.0952,
      "mean_token_accuracy": 0.16778693348169327,
      "num_tokens": 37519284.0,
      "step": 17125
    },
    {
      "entropy": 5.419823217391968,
      "epoch": 1.7000793965859469,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004715892720710414,
      "loss": 5.1987,
      "mean_token_accuracy": 0.166463965177536,
      "num_tokens": 37529699.0,
      "step": 17130
    },
    {
      "entropy": 5.486166095733642,
      "epoch": 1.7005756252481143,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004715720353621767,
      "loss": 5.2621,
      "mean_token_accuracy": 0.15633364915847778,
      "num_tokens": 37541756.0,
      "step": 17135
    },
    {
      "entropy": 5.375031471252441,
      "epoch": 1.7010718539102818,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047155479377874917,
      "loss": 5.1363,
      "mean_token_accuracy": 0.16827592253684998,
      "num_tokens": 37552693.0,
      "step": 17140
    },
    {
      "entropy": 5.388325023651123,
      "epoch": 1.7015680825724493,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047153754732118616,
      "loss": 5.1237,
      "mean_token_accuracy": 0.16656687557697297,
      "num_tokens": 37563826.0,
      "step": 17145
    },
    {
      "entropy": 5.39304666519165,
      "epoch": 1.7020643112346168,
      "grad_norm": 1.0,
      "learning_rate": 0.0004715202959899156,
      "loss": 5.1815,
      "mean_token_accuracy": 0.16388113051652908,
      "num_tokens": 37576645.0,
      "step": 17150
    },
    {
      "entropy": 5.311873292922973,
      "epoch": 1.7025605398967845,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004715030397853651,
      "loss": 5.0566,
      "mean_token_accuracy": 0.17210018038749694,
      "num_tokens": 37587467.0,
      "step": 17155
    },
    {
      "entropy": 5.438110399246216,
      "epoch": 1.703056768558952,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047148577870796264,
      "loss": 5.22,
      "mean_token_accuracy": 0.15876845568418502,
      "num_tokens": 37598475.0,
      "step": 17160
    },
    {
      "entropy": 5.487463092803955,
      "epoch": 1.7035529972211196,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004714685127581363,
      "loss": 5.3185,
      "mean_token_accuracy": 0.15719595551490784,
      "num_tokens": 37610260.0,
      "step": 17165
    },
    {
      "entropy": 5.434078073501587,
      "epoch": 1.704049225883287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047145124193631415,
      "loss": 5.2382,
      "mean_token_accuracy": 0.15974005311727524,
      "num_tokens": 37621129.0,
      "step": 17170
    },
    {
      "entropy": 5.428492879867553,
      "epoch": 1.7045454545454546,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047143396624292454,
      "loss": 5.1349,
      "mean_token_accuracy": 0.161228546500206,
      "num_tokens": 37633130.0,
      "step": 17175
    },
    {
      "entropy": 5.474985551834107,
      "epoch": 1.705041683207622,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047141668567839584,
      "loss": 5.2664,
      "mean_token_accuracy": 0.16134987622499466,
      "num_tokens": 37644247.0,
      "step": 17180
    },
    {
      "entropy": 5.428211116790772,
      "epoch": 1.7055379118697895,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004713994002431566,
      "loss": 5.1913,
      "mean_token_accuracy": 0.1662836879491806,
      "num_tokens": 37655602.0,
      "step": 17185
    },
    {
      "entropy": 5.330511569976807,
      "epoch": 1.706034140531957,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047138210993763534,
      "loss": 5.0972,
      "mean_token_accuracy": 0.16880395412445068,
      "num_tokens": 37667150.0,
      "step": 17190
    },
    {
      "entropy": 5.3791642665863035,
      "epoch": 1.7065303691941247,
      "grad_norm": 1.0,
      "learning_rate": 0.000471364814762261,
      "loss": 5.1205,
      "mean_token_accuracy": 0.17094438821077346,
      "num_tokens": 37677626.0,
      "step": 17195
    },
    {
      "entropy": 5.508015966415405,
      "epoch": 1.7070265978562922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004713475147174624,
      "loss": 5.2691,
      "mean_token_accuracy": 0.16215938627719878,
      "num_tokens": 37689050.0,
      "step": 17200
    },
    {
      "entropy": 5.468196630477905,
      "epoch": 1.7075228265184597,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004713302098036685,
      "loss": 5.1472,
      "mean_token_accuracy": 0.17862721532583237,
      "num_tokens": 37699190.0,
      "step": 17205
    },
    {
      "entropy": 5.300451469421387,
      "epoch": 1.7080190551806274,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004713129000213085,
      "loss": 5.1372,
      "mean_token_accuracy": 0.17085391134023667,
      "num_tokens": 37710465.0,
      "step": 17210
    },
    {
      "entropy": 5.381169843673706,
      "epoch": 1.7085152838427948,
      "grad_norm": 1.0,
      "learning_rate": 0.0004712955853708115,
      "loss": 5.2452,
      "mean_token_accuracy": 0.1644422307610512,
      "num_tokens": 37722426.0,
      "step": 17215
    },
    {
      "entropy": 5.375699853897094,
      "epoch": 1.7090115125049623,
      "grad_norm": 1.09375,
      "learning_rate": 0.000471278265852607,
      "loss": 5.142,
      "mean_token_accuracy": 0.1680700734257698,
      "num_tokens": 37733104.0,
      "step": 17220
    },
    {
      "entropy": 5.346209764480591,
      "epoch": 1.7095077411671298,
      "grad_norm": 1.0,
      "learning_rate": 0.0004712609414671245,
      "loss": 5.0725,
      "mean_token_accuracy": 0.16966979950666428,
      "num_tokens": 37742747.0,
      "step": 17225
    },
    {
      "entropy": 5.393371963500977,
      "epoch": 1.7100039698292973,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004712436122147935,
      "loss": 5.1353,
      "mean_token_accuracy": 0.16684239506721496,
      "num_tokens": 37753338.0,
      "step": 17230
    },
    {
      "entropy": 5.379377841949463,
      "epoch": 1.7105001984914647,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004712262780960438,
      "loss": 5.1159,
      "mean_token_accuracy": 0.1727169081568718,
      "num_tokens": 37763122.0,
      "step": 17235
    },
    {
      "entropy": 5.331936979293824,
      "epoch": 1.7109964271536324,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047120893911130525,
      "loss": 5.1311,
      "mean_token_accuracy": 0.16132695078849793,
      "num_tokens": 37774677.0,
      "step": 17240
    },
    {
      "entropy": 5.402272748947143,
      "epoch": 1.7114926558158,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047119159526100773,
      "loss": 5.1389,
      "mean_token_accuracy": 0.16163338273763656,
      "num_tokens": 37786186.0,
      "step": 17245
    },
    {
      "entropy": 5.371020269393921,
      "epoch": 1.7119888844779676,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047117424654558146,
      "loss": 5.1619,
      "mean_token_accuracy": 0.17053817361593246,
      "num_tokens": 37796837.0,
      "step": 17250
    },
    {
      "entropy": 5.376055669784546,
      "epoch": 1.712485113140135,
      "grad_norm": 1.0,
      "learning_rate": 0.0004711568929654566,
      "loss": 5.2167,
      "mean_token_accuracy": 0.16207532882690429,
      "num_tokens": 37808813.0,
      "step": 17255
    },
    {
      "entropy": 5.383073186874389,
      "epoch": 1.7129813418023025,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004711395345210635,
      "loss": 5.0867,
      "mean_token_accuracy": 0.16694195121526717,
      "num_tokens": 37819770.0,
      "step": 17260
    },
    {
      "entropy": 5.466442537307739,
      "epoch": 1.71347757046447,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004711221712128325,
      "loss": 5.1817,
      "mean_token_accuracy": 0.16511691808700563,
      "num_tokens": 37831831.0,
      "step": 17265
    },
    {
      "entropy": 5.417556858062744,
      "epoch": 1.7139737991266375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004711048030411943,
      "loss": 5.2086,
      "mean_token_accuracy": 0.165732179582119,
      "num_tokens": 37843092.0,
      "step": 17270
    },
    {
      "entropy": 5.406092548370362,
      "epoch": 1.714470027788805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047108743000657963,
      "loss": 5.1739,
      "mean_token_accuracy": 0.1743529409170151,
      "num_tokens": 37853192.0,
      "step": 17275
    },
    {
      "entropy": 5.491096353530883,
      "epoch": 1.7149662564509724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004710700521094192,
      "loss": 5.2015,
      "mean_token_accuracy": 0.16510847359895706,
      "num_tokens": 37863510.0,
      "step": 17280
    },
    {
      "entropy": 5.360181522369385,
      "epoch": 1.7154624851131401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047105266935014385,
      "loss": 5.1156,
      "mean_token_accuracy": 0.16883350312709808,
      "num_tokens": 37876156.0,
      "step": 17285
    },
    {
      "entropy": 5.399318075180053,
      "epoch": 1.7159587137753076,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047103528172918487,
      "loss": 5.175,
      "mean_token_accuracy": 0.16680110841989518,
      "num_tokens": 37887153.0,
      "step": 17290
    },
    {
      "entropy": 5.423823976516724,
      "epoch": 1.7164549424374753,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004710178892469733,
      "loss": 5.157,
      "mean_token_accuracy": 0.16966404616832734,
      "num_tokens": 37898820.0,
      "step": 17295
    },
    {
      "entropy": 5.3751383304595945,
      "epoch": 1.7169511710996428,
      "grad_norm": 1.0,
      "learning_rate": 0.00047100049190394044,
      "loss": 5.127,
      "mean_token_accuracy": 0.16493230313062668,
      "num_tokens": 37910257.0,
      "step": 17300
    },
    {
      "entropy": 5.418794298171997,
      "epoch": 1.7174473997618103,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004709830897005178,
      "loss": 5.257,
      "mean_token_accuracy": 0.16553752571344377,
      "num_tokens": 37921574.0,
      "step": 17305
    },
    {
      "entropy": 5.479655838012695,
      "epoch": 1.7179436284239777,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004709656826371367,
      "loss": 5.2093,
      "mean_token_accuracy": 0.1624714344739914,
      "num_tokens": 37932741.0,
      "step": 17310
    },
    {
      "entropy": 5.4389721870422365,
      "epoch": 1.7184398570861452,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004709482707142291,
      "loss": 5.1795,
      "mean_token_accuracy": 0.1667429268360138,
      "num_tokens": 37944592.0,
      "step": 17315
    },
    {
      "entropy": 5.383386325836182,
      "epoch": 1.7189360857483127,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047093085393222646,
      "loss": 5.1309,
      "mean_token_accuracy": 0.16854230612516402,
      "num_tokens": 37954983.0,
      "step": 17320
    },
    {
      "entropy": 5.396336650848388,
      "epoch": 1.7194323144104804,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047091343229156094,
      "loss": 5.2018,
      "mean_token_accuracy": 0.16473751664161682,
      "num_tokens": 37964491.0,
      "step": 17325
    },
    {
      "entropy": 5.400420904159546,
      "epoch": 1.7199285430726479,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004708960057926644,
      "loss": 5.0932,
      "mean_token_accuracy": 0.16618246883153914,
      "num_tokens": 37976140.0,
      "step": 17330
    },
    {
      "entropy": 5.381131362915039,
      "epoch": 1.7204247717348156,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047087857443596905,
      "loss": 5.1522,
      "mean_token_accuracy": 0.17310401201248168,
      "num_tokens": 37987776.0,
      "step": 17335
    },
    {
      "entropy": 5.408041381835938,
      "epoch": 1.720921000396983,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047086113822190716,
      "loss": 5.214,
      "mean_token_accuracy": 0.16242630928754806,
      "num_tokens": 37998088.0,
      "step": 17340
    },
    {
      "entropy": 5.376501035690308,
      "epoch": 1.7214172290591505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047084369715091103,
      "loss": 5.2009,
      "mean_token_accuracy": 0.16113226115703583,
      "num_tokens": 38009802.0,
      "step": 17345
    },
    {
      "entropy": 5.463438081741333,
      "epoch": 1.721913457721318,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004708262512234133,
      "loss": 5.2361,
      "mean_token_accuracy": 0.16917492747306823,
      "num_tokens": 38020797.0,
      "step": 17350
    },
    {
      "entropy": 5.483707475662231,
      "epoch": 1.7224096863834855,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004708088004398464,
      "loss": 5.2146,
      "mean_token_accuracy": 0.1652181327342987,
      "num_tokens": 38031816.0,
      "step": 17355
    },
    {
      "entropy": 5.380285453796387,
      "epoch": 1.722905915045653,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004707913448006433,
      "loss": 5.2203,
      "mean_token_accuracy": 0.155553138256073,
      "num_tokens": 38043953.0,
      "step": 17360
    },
    {
      "entropy": 5.321955060958862,
      "epoch": 1.7234021437078204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004707738843062367,
      "loss": 5.1067,
      "mean_token_accuracy": 0.17251148521900178,
      "num_tokens": 38053929.0,
      "step": 17365
    },
    {
      "entropy": 5.51962251663208,
      "epoch": 1.723898372369988,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004707564189570596,
      "loss": 5.3241,
      "mean_token_accuracy": 0.16040043234825135,
      "num_tokens": 38067091.0,
      "step": 17370
    },
    {
      "entropy": 5.486264276504516,
      "epoch": 1.7243946010321556,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004707389487535451,
      "loss": 5.1623,
      "mean_token_accuracy": 0.1650581657886505,
      "num_tokens": 38079002.0,
      "step": 17375
    },
    {
      "entropy": 5.4005943775177006,
      "epoch": 1.7248908296943233,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004707214736961265,
      "loss": 5.1749,
      "mean_token_accuracy": 0.17054188698530198,
      "num_tokens": 38089526.0,
      "step": 17380
    },
    {
      "entropy": 5.428851461410522,
      "epoch": 1.7253870583564908,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004707039937852371,
      "loss": 5.1575,
      "mean_token_accuracy": 0.16629530489444733,
      "num_tokens": 38099902.0,
      "step": 17385
    },
    {
      "entropy": 5.35589861869812,
      "epoch": 1.7258832870186582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047068650902131034,
      "loss": 5.1357,
      "mean_token_accuracy": 0.1737617015838623,
      "num_tokens": 38111725.0,
      "step": 17390
    },
    {
      "entropy": 5.382986164093017,
      "epoch": 1.7263795156808257,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004706690194047798,
      "loss": 5.1001,
      "mean_token_accuracy": 0.1668451339006424,
      "num_tokens": 38123522.0,
      "step": 17395
    },
    {
      "entropy": 5.4363665103912355,
      "epoch": 1.7268757443429932,
      "grad_norm": 1.125,
      "learning_rate": 0.0004706515249360792,
      "loss": 5.2317,
      "mean_token_accuracy": 0.16046728044748307,
      "num_tokens": 38133968.0,
      "step": 17400
    },
    {
      "entropy": 5.403831243515015,
      "epoch": 1.7273719730051607,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004706340256156425,
      "loss": 5.2734,
      "mean_token_accuracy": 0.15639201402664185,
      "num_tokens": 38145288.0,
      "step": 17405
    },
    {
      "entropy": 5.441918802261353,
      "epoch": 1.7278682016673284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004706165214439034,
      "loss": 5.1701,
      "mean_token_accuracy": 0.1650937333703041,
      "num_tokens": 38156141.0,
      "step": 17410
    },
    {
      "entropy": 5.450276517868042,
      "epoch": 1.7283644303294958,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004705990124212961,
      "loss": 5.1457,
      "mean_token_accuracy": 0.17493642568588258,
      "num_tokens": 38167211.0,
      "step": 17415
    },
    {
      "entropy": 5.361220264434815,
      "epoch": 1.7288606589916635,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004705814985482548,
      "loss": 5.2079,
      "mean_token_accuracy": 0.16297906786203384,
      "num_tokens": 38177360.0,
      "step": 17420
    },
    {
      "entropy": 5.374156522750854,
      "epoch": 1.729356887653831,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047056397982521366,
      "loss": 5.1118,
      "mean_token_accuracy": 0.17049003392457962,
      "num_tokens": 38187857.0,
      "step": 17425
    },
    {
      "entropy": 5.483945274353028,
      "epoch": 1.7298531163159985,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004705464562526074,
      "loss": 5.2181,
      "mean_token_accuracy": 0.15177076160907746,
      "num_tokens": 38198161.0,
      "step": 17430
    },
    {
      "entropy": 5.434214401245117,
      "epoch": 1.730349344978166,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004705289278308703,
      "loss": 5.1534,
      "mean_token_accuracy": 0.1715497627854347,
      "num_tokens": 38208612.0,
      "step": 17435
    },
    {
      "entropy": 5.400346326828003,
      "epoch": 1.7308455736403334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047051139456043707,
      "loss": 5.1953,
      "mean_token_accuracy": 0.16266478598117828,
      "num_tokens": 38220153.0,
      "step": 17440
    },
    {
      "entropy": 5.393369197845459,
      "epoch": 1.731341802302501,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004704938564417427,
      "loss": 5.1646,
      "mean_token_accuracy": 0.17188536375761032,
      "num_tokens": 38230790.0,
      "step": 17445
    },
    {
      "entropy": 5.420899343490601,
      "epoch": 1.7318380309646684,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047047631347522175,
      "loss": 5.1162,
      "mean_token_accuracy": 0.16906149089336395,
      "num_tokens": 38242585.0,
      "step": 17450
    },
    {
      "entropy": 5.4199622631072994,
      "epoch": 1.732334259626836,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047045876566130957,
      "loss": 5.1598,
      "mean_token_accuracy": 0.16582341641187667,
      "num_tokens": 38253316.0,
      "step": 17455
    },
    {
      "entropy": 5.388956499099732,
      "epoch": 1.7328304882890035,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004704412130004411,
      "loss": 5.1428,
      "mean_token_accuracy": 0.16868604868650436,
      "num_tokens": 38263230.0,
      "step": 17460
    },
    {
      "entropy": 5.453596210479736,
      "epoch": 1.7333267169511712,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047042365549305174,
      "loss": 5.2135,
      "mean_token_accuracy": 0.16263943463563918,
      "num_tokens": 38273416.0,
      "step": 17465
    },
    {
      "entropy": 5.360785293579101,
      "epoch": 1.7338229456133387,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004704060931395768,
      "loss": 5.1348,
      "mean_token_accuracy": 0.17258938550949096,
      "num_tokens": 38285764.0,
      "step": 17470
    },
    {
      "entropy": 5.441365146636963,
      "epoch": 1.7343191742755062,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047038852594045186,
      "loss": 5.2336,
      "mean_token_accuracy": 0.160908043384552,
      "num_tokens": 38296899.0,
      "step": 17475
    },
    {
      "entropy": 5.466432285308838,
      "epoch": 1.7348154029376737,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047037095389611245,
      "loss": 5.2315,
      "mean_token_accuracy": 0.16757635474205018,
      "num_tokens": 38309050.0,
      "step": 17480
    },
    {
      "entropy": 5.379732751846314,
      "epoch": 1.7353116315998411,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004703533770069944,
      "loss": 5.186,
      "mean_token_accuracy": 0.16513979882001878,
      "num_tokens": 38320831.0,
      "step": 17485
    },
    {
      "entropy": 5.412419557571411,
      "epoch": 1.7358078602620086,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047033579527353355,
      "loss": 5.1892,
      "mean_token_accuracy": 0.16275124847888947,
      "num_tokens": 38332096.0,
      "step": 17490
    },
    {
      "entropy": 5.465515899658203,
      "epoch": 1.7363040889241763,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004703182086961659,
      "loss": 5.2148,
      "mean_token_accuracy": 0.1628094032406807,
      "num_tokens": 38343110.0,
      "step": 17495
    },
    {
      "entropy": 5.46600832939148,
      "epoch": 1.7368003175863438,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004703006172753274,
      "loss": 5.2156,
      "mean_token_accuracy": 0.16733063161373138,
      "num_tokens": 38353971.0,
      "step": 17500
    },
    {
      "entropy": 5.34343409538269,
      "epoch": 1.7372965462485113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047028302101145467,
      "loss": 5.1291,
      "mean_token_accuracy": 0.17285202592611312,
      "num_tokens": 38365114.0,
      "step": 17505
    },
    {
      "entropy": 5.460682439804077,
      "epoch": 1.737792774910679,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004702654199049836,
      "loss": 5.1621,
      "mean_token_accuracy": 0.1618335947394371,
      "num_tokens": 38375784.0,
      "step": 17510
    },
    {
      "entropy": 5.414890718460083,
      "epoch": 1.7382890035728464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000470247813956351,
      "loss": 5.1393,
      "mean_token_accuracy": 0.1693231239914894,
      "num_tokens": 38386053.0,
      "step": 17515
    },
    {
      "entropy": 5.3241064071655275,
      "epoch": 1.738785232235014,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047023020316599334,
      "loss": 5.1385,
      "mean_token_accuracy": 0.16696538925170898,
      "num_tokens": 38396545.0,
      "step": 17520
    },
    {
      "entropy": 5.392079925537109,
      "epoch": 1.7392814608971814,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047021258753434725,
      "loss": 5.1168,
      "mean_token_accuracy": 0.1642117530107498,
      "num_tokens": 38406344.0,
      "step": 17525
    },
    {
      "entropy": 5.420540523529053,
      "epoch": 1.7397776895593489,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047019496706184966,
      "loss": 5.0921,
      "mean_token_accuracy": 0.17534428387880324,
      "num_tokens": 38416388.0,
      "step": 17530
    },
    {
      "entropy": 5.451218986511231,
      "epoch": 1.7402739182215163,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004701773417489375,
      "loss": 5.2481,
      "mean_token_accuracy": 0.1569109097123146,
      "num_tokens": 38427608.0,
      "step": 17535
    },
    {
      "entropy": 5.425534248352051,
      "epoch": 1.740770146883684,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047015971159604787,
      "loss": 5.1593,
      "mean_token_accuracy": 0.16459945887327193,
      "num_tokens": 38437921.0,
      "step": 17540
    },
    {
      "entropy": 5.40063738822937,
      "epoch": 1.7412663755458515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000470142076603618,
      "loss": 5.2107,
      "mean_token_accuracy": 0.16042704582214357,
      "num_tokens": 38448767.0,
      "step": 17545
    },
    {
      "entropy": 5.432316589355469,
      "epoch": 1.7417626042080192,
      "grad_norm": 1.0,
      "learning_rate": 0.00047012443677208496,
      "loss": 5.2812,
      "mean_token_accuracy": 0.16057053059339524,
      "num_tokens": 38460052.0,
      "step": 17550
    },
    {
      "entropy": 5.394101810455322,
      "epoch": 1.7422588328701867,
      "grad_norm": 1.125,
      "learning_rate": 0.0004701067921018864,
      "loss": 5.1329,
      "mean_token_accuracy": 0.16525890529155732,
      "num_tokens": 38470177.0,
      "step": 17555
    },
    {
      "entropy": 5.442356777191162,
      "epoch": 1.7427550615323542,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047008914259345986,
      "loss": 5.1589,
      "mean_token_accuracy": 0.15986270904541017,
      "num_tokens": 38481608.0,
      "step": 17560
    },
    {
      "entropy": 5.424878263473511,
      "epoch": 1.7432512901945216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047007148824724295,
      "loss": 5.2186,
      "mean_token_accuracy": 0.1631849601864815,
      "num_tokens": 38492534.0,
      "step": 17565
    },
    {
      "entropy": 5.4239466190338135,
      "epoch": 1.743747518856689,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004700538290636734,
      "loss": 5.2207,
      "mean_token_accuracy": 0.1638500511646271,
      "num_tokens": 38503629.0,
      "step": 17570
    },
    {
      "entropy": 5.379516983032227,
      "epoch": 1.7442437475188566,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047003616504318925,
      "loss": 5.1012,
      "mean_token_accuracy": 0.1672505721449852,
      "num_tokens": 38514428.0,
      "step": 17575
    },
    {
      "entropy": 5.4515495777130125,
      "epoch": 1.744739976181024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047001849618622844,
      "loss": 5.1768,
      "mean_token_accuracy": 0.16117069870233536,
      "num_tokens": 38525468.0,
      "step": 17580
    },
    {
      "entropy": 5.4076415538787845,
      "epoch": 1.7452362048431918,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047000082249322916,
      "loss": 5.1376,
      "mean_token_accuracy": 0.16432869881391526,
      "num_tokens": 38535095.0,
      "step": 17585
    },
    {
      "entropy": 5.4400410652160645,
      "epoch": 1.7457324335053592,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004699831439646297,
      "loss": 5.1534,
      "mean_token_accuracy": 0.16194335371255875,
      "num_tokens": 38546867.0,
      "step": 17590
    },
    {
      "entropy": 5.348616743087769,
      "epoch": 1.746228662167527,
      "grad_norm": 1.125,
      "learning_rate": 0.0004699654606008684,
      "loss": 5.0452,
      "mean_token_accuracy": 0.17517765909433364,
      "num_tokens": 38557231.0,
      "step": 17595
    },
    {
      "entropy": 5.456311321258545,
      "epoch": 1.7467248908296944,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004699477724023838,
      "loss": 5.1822,
      "mean_token_accuracy": 0.16690780073404313,
      "num_tokens": 38568857.0,
      "step": 17600
    },
    {
      "entropy": 5.445173692703247,
      "epoch": 1.7472211194918619,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004699300793696145,
      "loss": 5.1558,
      "mean_token_accuracy": 0.16256024092435836,
      "num_tokens": 38579177.0,
      "step": 17605
    },
    {
      "entropy": 5.351577949523926,
      "epoch": 1.7477173481540293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004699123815029993,
      "loss": 5.0397,
      "mean_token_accuracy": 0.17087241113185883,
      "num_tokens": 38588683.0,
      "step": 17610
    },
    {
      "entropy": 5.327899169921875,
      "epoch": 1.7482135768161968,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046989467880297704,
      "loss": 5.1347,
      "mean_token_accuracy": 0.16327082961797715,
      "num_tokens": 38599636.0,
      "step": 17615
    },
    {
      "entropy": 5.468163728713989,
      "epoch": 1.7487098054783643,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046987697126998673,
      "loss": 5.2619,
      "mean_token_accuracy": 0.15863164216279985,
      "num_tokens": 38609360.0,
      "step": 17620
    },
    {
      "entropy": 5.4700599193573,
      "epoch": 1.749206034140532,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004698592589044674,
      "loss": 5.1656,
      "mean_token_accuracy": 0.16418059915304184,
      "num_tokens": 38620996.0,
      "step": 17625
    },
    {
      "entropy": 5.46523118019104,
      "epoch": 1.7497022628026995,
      "grad_norm": 1.125,
      "learning_rate": 0.00046984154170685833,
      "loss": 5.2986,
      "mean_token_accuracy": 0.163292033970356,
      "num_tokens": 38632935.0,
      "step": 17630
    },
    {
      "entropy": 5.48703351020813,
      "epoch": 1.7501984914648672,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004698238196775989,
      "loss": 5.2475,
      "mean_token_accuracy": 0.16106752008199693,
      "num_tokens": 38644299.0,
      "step": 17635
    },
    {
      "entropy": 5.461573839187622,
      "epoch": 1.7506947201270346,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046980609281712857,
      "loss": 5.2203,
      "mean_token_accuracy": 0.16451325714588166,
      "num_tokens": 38654970.0,
      "step": 17640
    },
    {
      "entropy": 5.463474082946777,
      "epoch": 1.7511909487892021,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004697883611258868,
      "loss": 5.2555,
      "mean_token_accuracy": 0.1607454687356949,
      "num_tokens": 38666302.0,
      "step": 17645
    },
    {
      "entropy": 5.43006157875061,
      "epoch": 1.7516871774513696,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004697706246043136,
      "loss": 5.1615,
      "mean_token_accuracy": 0.16767675280570984,
      "num_tokens": 38678346.0,
      "step": 17650
    },
    {
      "entropy": 5.389701509475708,
      "epoch": 1.752183406113537,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046975288325284845,
      "loss": 5.0981,
      "mean_token_accuracy": 0.16961120218038558,
      "num_tokens": 38688242.0,
      "step": 17655
    },
    {
      "entropy": 5.452594518661499,
      "epoch": 1.7526796347757045,
      "grad_norm": 1.125,
      "learning_rate": 0.0004697351370719315,
      "loss": 5.2861,
      "mean_token_accuracy": 0.1577180117368698,
      "num_tokens": 38698756.0,
      "step": 17660
    },
    {
      "entropy": 5.4615724086761475,
      "epoch": 1.753175863437872,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004697173860620027,
      "loss": 5.2404,
      "mean_token_accuracy": 0.160337395966053,
      "num_tokens": 38710410.0,
      "step": 17665
    },
    {
      "entropy": 5.520569944381714,
      "epoch": 1.7536720921000397,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004696996302235024,
      "loss": 5.203,
      "mean_token_accuracy": 0.1587560847401619,
      "num_tokens": 38721917.0,
      "step": 17670
    },
    {
      "entropy": 5.4303652286529545,
      "epoch": 1.7541683207622072,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046968186955687074,
      "loss": 5.2093,
      "mean_token_accuracy": 0.16768753230571748,
      "num_tokens": 38733085.0,
      "step": 17675
    },
    {
      "entropy": 5.396076679229736,
      "epoch": 1.7546645494243749,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004696641040625483,
      "loss": 5.2262,
      "mean_token_accuracy": 0.1603509858250618,
      "num_tokens": 38744785.0,
      "step": 17680
    },
    {
      "entropy": 5.464212417602539,
      "epoch": 1.7551607780865424,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046964633374097544,
      "loss": 5.2132,
      "mean_token_accuracy": 0.1704317033290863,
      "num_tokens": 38755626.0,
      "step": 17685
    },
    {
      "entropy": 5.448471450805664,
      "epoch": 1.7556570067487098,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046962855859259306,
      "loss": 5.0838,
      "mean_token_accuracy": 0.16299442648887635,
      "num_tokens": 38766360.0,
      "step": 17690
    },
    {
      "entropy": 5.37502613067627,
      "epoch": 1.7561532354108773,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046961077861784167,
      "loss": 5.1727,
      "mean_token_accuracy": 0.16223552972078323,
      "num_tokens": 38776763.0,
      "step": 17695
    },
    {
      "entropy": 5.390544414520264,
      "epoch": 1.7566494640730448,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004695929938171624,
      "loss": 5.234,
      "mean_token_accuracy": 0.1652033194899559,
      "num_tokens": 38787664.0,
      "step": 17700
    },
    {
      "entropy": 5.41287784576416,
      "epoch": 1.7571456927352123,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004695752041909962,
      "loss": 5.145,
      "mean_token_accuracy": 0.16387696266174318,
      "num_tokens": 38798018.0,
      "step": 17705
    },
    {
      "entropy": 5.400228786468506,
      "epoch": 1.75764192139738,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004695574097397842,
      "loss": 5.0983,
      "mean_token_accuracy": 0.16528011113405228,
      "num_tokens": 38810381.0,
      "step": 17710
    },
    {
      "entropy": 5.331123638153076,
      "epoch": 1.7581381500595474,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046953961046396767,
      "loss": 5.1192,
      "mean_token_accuracy": 0.17389245927333832,
      "num_tokens": 38821042.0,
      "step": 17715
    },
    {
      "entropy": 5.370192527770996,
      "epoch": 1.7586343787217151,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046952180636398793,
      "loss": 5.0599,
      "mean_token_accuracy": 0.17194833159446715,
      "num_tokens": 38830890.0,
      "step": 17720
    },
    {
      "entropy": 5.407809495925903,
      "epoch": 1.7591306073838826,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004695039974402866,
      "loss": 5.1251,
      "mean_token_accuracy": 0.17146253883838652,
      "num_tokens": 38842648.0,
      "step": 17725
    },
    {
      "entropy": 5.396724367141724,
      "epoch": 1.75962683604605,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046948618369330533,
      "loss": 5.0821,
      "mean_token_accuracy": 0.17409319430589676,
      "num_tokens": 38852740.0,
      "step": 17730
    },
    {
      "entropy": 5.3786386013031,
      "epoch": 1.7601230647082176,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046946836512348566,
      "loss": 5.2037,
      "mean_token_accuracy": 0.15895124524831772,
      "num_tokens": 38863371.0,
      "step": 17735
    },
    {
      "entropy": 5.475363159179688,
      "epoch": 1.760619293370385,
      "grad_norm": 1.125,
      "learning_rate": 0.00046945054173126965,
      "loss": 5.1827,
      "mean_token_accuracy": 0.16238282471895218,
      "num_tokens": 38873893.0,
      "step": 17740
    },
    {
      "entropy": 5.42559404373169,
      "epoch": 1.7611155220325525,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046943271351709923,
      "loss": 5.2275,
      "mean_token_accuracy": 0.16175204664468765,
      "num_tokens": 38884778.0,
      "step": 17745
    },
    {
      "entropy": 5.351330804824829,
      "epoch": 1.76161175069472,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004694148804814164,
      "loss": 5.1673,
      "mean_token_accuracy": 0.1656200870871544,
      "num_tokens": 38896011.0,
      "step": 17750
    },
    {
      "entropy": 5.446309375762939,
      "epoch": 1.7621079793568877,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004693970426246635,
      "loss": 5.2032,
      "mean_token_accuracy": 0.16419016122817992,
      "num_tokens": 38907686.0,
      "step": 17755
    },
    {
      "entropy": 5.356894397735596,
      "epoch": 1.7626042080190552,
      "grad_norm": 1.125,
      "learning_rate": 0.00046937919994728287,
      "loss": 5.1709,
      "mean_token_accuracy": 0.167932191491127,
      "num_tokens": 38919180.0,
      "step": 17760
    },
    {
      "entropy": 5.386002159118652,
      "epoch": 1.7631004366812228,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004693613524497168,
      "loss": 5.1304,
      "mean_token_accuracy": 0.16746100187301635,
      "num_tokens": 38929600.0,
      "step": 17765
    },
    {
      "entropy": 5.399684810638428,
      "epoch": 1.7635966653433903,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046934350013240814,
      "loss": 5.0844,
      "mean_token_accuracy": 0.1765008509159088,
      "num_tokens": 38940404.0,
      "step": 17770
    },
    {
      "entropy": 5.374958181381226,
      "epoch": 1.7640928940055578,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004693256429957994,
      "loss": 5.1525,
      "mean_token_accuracy": 0.16899698078632355,
      "num_tokens": 38951929.0,
      "step": 17775
    },
    {
      "entropy": 5.382019472122193,
      "epoch": 1.7645891226677253,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004693077810403335,
      "loss": 5.1454,
      "mean_token_accuracy": 0.16965433210134506,
      "num_tokens": 38961954.0,
      "step": 17780
    },
    {
      "entropy": 5.344806718826294,
      "epoch": 1.7650853513298927,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046928991426645325,
      "loss": 5.1218,
      "mean_token_accuracy": 0.17320347875356673,
      "num_tokens": 38970974.0,
      "step": 17785
    },
    {
      "entropy": 5.424071598052978,
      "epoch": 1.7655815799920602,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046927204267460177,
      "loss": 5.209,
      "mean_token_accuracy": 0.16154689341783524,
      "num_tokens": 38982481.0,
      "step": 17790
    },
    {
      "entropy": 5.4869850158691404,
      "epoch": 1.766077808654228,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046925416626522235,
      "loss": 5.156,
      "mean_token_accuracy": 0.16865446418523788,
      "num_tokens": 38993279.0,
      "step": 17795
    },
    {
      "entropy": 5.388056516647339,
      "epoch": 1.7665740373163954,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004692362850387581,
      "loss": 5.1299,
      "mean_token_accuracy": 0.16876190751791,
      "num_tokens": 39003748.0,
      "step": 17800
    },
    {
      "entropy": 5.3686939716339115,
      "epoch": 1.7670702659785629,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046921839899565263,
      "loss": 5.1369,
      "mean_token_accuracy": 0.16673893928527833,
      "num_tokens": 39014902.0,
      "step": 17805
    },
    {
      "entropy": 5.411395645141601,
      "epoch": 1.7675664946407306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004692005081363493,
      "loss": 5.233,
      "mean_token_accuracy": 0.1608179196715355,
      "num_tokens": 39026484.0,
      "step": 17810
    },
    {
      "entropy": 5.3942607879638675,
      "epoch": 1.768062723302898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046918261246129176,
      "loss": 5.2113,
      "mean_token_accuracy": 0.1600827157497406,
      "num_tokens": 39036268.0,
      "step": 17815
    },
    {
      "entropy": 5.434029245376587,
      "epoch": 1.7685589519650655,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000469164711970924,
      "loss": 5.245,
      "mean_token_accuracy": 0.1639394074678421,
      "num_tokens": 39049092.0,
      "step": 17820
    },
    {
      "entropy": 5.504797744750976,
      "epoch": 1.769055180627233,
      "grad_norm": 1.125,
      "learning_rate": 0.0004691468066656896,
      "loss": 5.2018,
      "mean_token_accuracy": 0.16004678308963777,
      "num_tokens": 39059255.0,
      "step": 17825
    },
    {
      "entropy": 5.378424119949341,
      "epoch": 1.7695514092894005,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046912889654603286,
      "loss": 5.1063,
      "mean_token_accuracy": 0.17184173315763474,
      "num_tokens": 39069913.0,
      "step": 17830
    },
    {
      "entropy": 5.370528841018677,
      "epoch": 1.770047637951568,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004691109816123977,
      "loss": 5.1294,
      "mean_token_accuracy": 0.1700873479247093,
      "num_tokens": 39080666.0,
      "step": 17835
    },
    {
      "entropy": 5.459313869476318,
      "epoch": 1.7705438666137356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004690930618652286,
      "loss": 5.1966,
      "mean_token_accuracy": 0.1608204111456871,
      "num_tokens": 39091562.0,
      "step": 17840
    },
    {
      "entropy": 5.493925857543945,
      "epoch": 1.7710400952759031,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046907513730496975,
      "loss": 5.2769,
      "mean_token_accuracy": 0.1583893820643425,
      "num_tokens": 39102567.0,
      "step": 17845
    },
    {
      "entropy": 5.486578035354614,
      "epoch": 1.7715363239380708,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004690572079320657,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1733448475599289,
      "num_tokens": 39113260.0,
      "step": 17850
    },
    {
      "entropy": 5.405282258987427,
      "epoch": 1.7720325526002383,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046903927374696106,
      "loss": 5.1945,
      "mean_token_accuracy": 0.16605886369943618,
      "num_tokens": 39125790.0,
      "step": 17855
    },
    {
      "entropy": 5.411198139190674,
      "epoch": 1.7725287812624058,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004690213347501005,
      "loss": 5.2515,
      "mean_token_accuracy": 0.16036480218172072,
      "num_tokens": 39137220.0,
      "step": 17860
    },
    {
      "entropy": 5.416819334030151,
      "epoch": 1.7730250099245732,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004690033909419289,
      "loss": 5.1281,
      "mean_token_accuracy": 0.16820897310972213,
      "num_tokens": 39146362.0,
      "step": 17865
    },
    {
      "entropy": 5.463801908493042,
      "epoch": 1.7735212385867407,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046898544232289126,
      "loss": 5.2004,
      "mean_token_accuracy": 0.16304625570774078,
      "num_tokens": 39156822.0,
      "step": 17870
    },
    {
      "entropy": 5.4707084655761715,
      "epoch": 1.7740174672489082,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046896748889343267,
      "loss": 5.1765,
      "mean_token_accuracy": 0.16252601593732835,
      "num_tokens": 39166693.0,
      "step": 17875
    },
    {
      "entropy": 5.368331718444824,
      "epoch": 1.7745136959110757,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046894953065399836,
      "loss": 5.156,
      "mean_token_accuracy": 0.16609641760587693,
      "num_tokens": 39176889.0,
      "step": 17880
    },
    {
      "entropy": 5.413202905654908,
      "epoch": 1.7750099245732434,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046893156760503357,
      "loss": 5.1342,
      "mean_token_accuracy": 0.17340200543403625,
      "num_tokens": 39188077.0,
      "step": 17885
    },
    {
      "entropy": 5.383317470550537,
      "epoch": 1.7755061532354108,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004689135997469838,
      "loss": 5.1414,
      "mean_token_accuracy": 0.16335588842630386,
      "num_tokens": 39198182.0,
      "step": 17890
    },
    {
      "entropy": 5.465668487548828,
      "epoch": 1.7760023818975785,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004688956270802945,
      "loss": 5.1098,
      "mean_token_accuracy": 0.1735809251666069,
      "num_tokens": 39208801.0,
      "step": 17895
    },
    {
      "entropy": 5.331464099884033,
      "epoch": 1.776498610559746,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004688776496054117,
      "loss": 5.0064,
      "mean_token_accuracy": 0.17808210477232933,
      "num_tokens": 39220658.0,
      "step": 17900
    },
    {
      "entropy": 5.379657554626465,
      "epoch": 1.7769948392219135,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046885966732278077,
      "loss": 5.2327,
      "mean_token_accuracy": 0.1626279830932617,
      "num_tokens": 39231987.0,
      "step": 17905
    },
    {
      "entropy": 5.431660270690918,
      "epoch": 1.777491067884081,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046884168023284787,
      "loss": 5.1162,
      "mean_token_accuracy": 0.16810089498758315,
      "num_tokens": 39242797.0,
      "step": 17910
    },
    {
      "entropy": 5.425424528121948,
      "epoch": 1.7779872965462484,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046882368833605894,
      "loss": 5.1822,
      "mean_token_accuracy": 0.16733010858297348,
      "num_tokens": 39253621.0,
      "step": 17915
    },
    {
      "entropy": 5.348909521102906,
      "epoch": 1.778483525208416,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046880569163286023,
      "loss": 5.2039,
      "mean_token_accuracy": 0.16090297400951387,
      "num_tokens": 39264126.0,
      "step": 17920
    },
    {
      "entropy": 5.40479850769043,
      "epoch": 1.7789797538705836,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046878769012369807,
      "loss": 5.1216,
      "mean_token_accuracy": 0.1718825116753578,
      "num_tokens": 39275118.0,
      "step": 17925
    },
    {
      "entropy": 5.4156395435333256,
      "epoch": 1.779475982532751,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046876968380901864,
      "loss": 5.1488,
      "mean_token_accuracy": 0.163090281188488,
      "num_tokens": 39287237.0,
      "step": 17930
    },
    {
      "entropy": 5.464834451675415,
      "epoch": 1.7799722111949188,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046875167268926865,
      "loss": 5.2042,
      "mean_token_accuracy": 0.16385067403316497,
      "num_tokens": 39296995.0,
      "step": 17935
    },
    {
      "entropy": 5.372472095489502,
      "epoch": 1.7804684398570862,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004687336567648947,
      "loss": 5.0974,
      "mean_token_accuracy": 0.1673683136701584,
      "num_tokens": 39305980.0,
      "step": 17940
    },
    {
      "entropy": 5.424939250946045,
      "epoch": 1.7809646685192537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046871563603634347,
      "loss": 5.2179,
      "mean_token_accuracy": 0.1570023149251938,
      "num_tokens": 39316494.0,
      "step": 17945
    },
    {
      "entropy": 5.453716516494751,
      "epoch": 1.7814608971814212,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004686976105040619,
      "loss": 5.1991,
      "mean_token_accuracy": 0.16586970686912536,
      "num_tokens": 39327699.0,
      "step": 17950
    },
    {
      "entropy": 5.454422664642334,
      "epoch": 1.7819571258435887,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004686795801684969,
      "loss": 5.2034,
      "mean_token_accuracy": 0.16297926008701324,
      "num_tokens": 39338749.0,
      "step": 17955
    },
    {
      "entropy": 5.384872055053711,
      "epoch": 1.7824533545057561,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004686615450300958,
      "loss": 5.2037,
      "mean_token_accuracy": 0.16951082944869994,
      "num_tokens": 39349747.0,
      "step": 17960
    },
    {
      "entropy": 5.340085458755493,
      "epoch": 1.7829495831679236,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004686435050893056,
      "loss": 5.0444,
      "mean_token_accuracy": 0.1694420024752617,
      "num_tokens": 39360068.0,
      "step": 17965
    },
    {
      "entropy": 5.325376796722412,
      "epoch": 1.7834458118300913,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046862546034657376,
      "loss": 5.0869,
      "mean_token_accuracy": 0.1736004963517189,
      "num_tokens": 39369651.0,
      "step": 17970
    },
    {
      "entropy": 5.428344202041626,
      "epoch": 1.7839420404922588,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004686074108023477,
      "loss": 5.2795,
      "mean_token_accuracy": 0.16384112238883972,
      "num_tokens": 39381354.0,
      "step": 17975
    },
    {
      "entropy": 5.4691156387329105,
      "epoch": 1.7844382691544265,
      "grad_norm": 1.125,
      "learning_rate": 0.00046858935645707506,
      "loss": 5.1204,
      "mean_token_accuracy": 0.16415287405252457,
      "num_tokens": 39391793.0,
      "step": 17980
    },
    {
      "entropy": 5.406826305389404,
      "epoch": 1.784934497816594,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004685712973112035,
      "loss": 5.1511,
      "mean_token_accuracy": 0.1679369568824768,
      "num_tokens": 39402194.0,
      "step": 17985
    },
    {
      "entropy": 5.399519062042236,
      "epoch": 1.7854307264787614,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004685532333651809,
      "loss": 5.1644,
      "mean_token_accuracy": 0.16597305983304977,
      "num_tokens": 39413193.0,
      "step": 17990
    },
    {
      "entropy": 5.421874284744263,
      "epoch": 1.785926955140929,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046853516461945523,
      "loss": 5.1289,
      "mean_token_accuracy": 0.1749735713005066,
      "num_tokens": 39423072.0,
      "step": 17995
    },
    {
      "entropy": 5.453129434585572,
      "epoch": 1.7864231838030964,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046851709107447446,
      "loss": 5.2527,
      "mean_token_accuracy": 0.15968129336833953,
      "num_tokens": 39435088.0,
      "step": 18000
    },
    {
      "epoch": 1.7864231838030964,
      "eval_entropy": 5.294529446628831,
      "eval_loss": 5.300205707550049,
      "eval_mean_token_accuracy": 0.16411512008569223,
      "eval_num_tokens": 39435088.0,
      "eval_runtime": 26.718,
      "eval_samples_per_second": 1260.16,
      "eval_steps_per_second": 157.534,
      "step": 18000
    },
    {
      "entropy": 5.362164831161499,
      "epoch": 1.7869194124652639,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004684990127306868,
      "loss": 5.107,
      "mean_token_accuracy": 0.16722497940063477,
      "num_tokens": 39446129.0,
      "step": 18005
    },
    {
      "entropy": 5.406700229644775,
      "epoch": 1.7874156411274316,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046848092958854065,
      "loss": 5.0843,
      "mean_token_accuracy": 0.1828065484762192,
      "num_tokens": 39458127.0,
      "step": 18010
    },
    {
      "entropy": 5.4345787525177,
      "epoch": 1.787911869789599,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004684628416484843,
      "loss": 5.1274,
      "mean_token_accuracy": 0.174571593105793,
      "num_tokens": 39468204.0,
      "step": 18015
    },
    {
      "entropy": 5.395790052413941,
      "epoch": 1.7884080984517667,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004684447489109663,
      "loss": 5.2119,
      "mean_token_accuracy": 0.15702314674854279,
      "num_tokens": 39479114.0,
      "step": 18020
    },
    {
      "entropy": 5.4308747291564945,
      "epoch": 1.7889043271139342,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046842665137643546,
      "loss": 5.1524,
      "mean_token_accuracy": 0.17144710272550584,
      "num_tokens": 39489691.0,
      "step": 18025
    },
    {
      "entropy": 5.473092555999756,
      "epoch": 1.7894005557761017,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046840854904534046,
      "loss": 5.2812,
      "mean_token_accuracy": 0.15399994552135468,
      "num_tokens": 39501879.0,
      "step": 18030
    },
    {
      "entropy": 5.459778690338135,
      "epoch": 1.7898967844382692,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004683904419181302,
      "loss": 5.2092,
      "mean_token_accuracy": 0.16724476516246795,
      "num_tokens": 39513413.0,
      "step": 18035
    },
    {
      "entropy": 5.4510523796081545,
      "epoch": 1.7903930131004366,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046837232999525366,
      "loss": 5.1493,
      "mean_token_accuracy": 0.1658142700791359,
      "num_tokens": 39522719.0,
      "step": 18040
    },
    {
      "entropy": 5.444062995910644,
      "epoch": 1.7908892417626041,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046835421327716,
      "loss": 5.1892,
      "mean_token_accuracy": 0.1645371913909912,
      "num_tokens": 39532463.0,
      "step": 18045
    },
    {
      "entropy": 5.389033412933349,
      "epoch": 1.7913854704247716,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004683360917642986,
      "loss": 5.1583,
      "mean_token_accuracy": 0.15878596752882004,
      "num_tokens": 39543005.0,
      "step": 18050
    },
    {
      "entropy": 5.449963474273682,
      "epoch": 1.7918816990869393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046831796545711865,
      "loss": 5.2285,
      "mean_token_accuracy": 0.15875501781702042,
      "num_tokens": 39553468.0,
      "step": 18055
    },
    {
      "entropy": 5.526216602325439,
      "epoch": 1.7923779277491068,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046829983435606973,
      "loss": 5.2751,
      "mean_token_accuracy": 0.1619040846824646,
      "num_tokens": 39563539.0,
      "step": 18060
    },
    {
      "entropy": 5.4707708835601805,
      "epoch": 1.7928741564112745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004682816984616014,
      "loss": 5.2026,
      "mean_token_accuracy": 0.17047148644924165,
      "num_tokens": 39574835.0,
      "step": 18065
    },
    {
      "entropy": 5.428465938568115,
      "epoch": 1.793370385073442,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046826355777416346,
      "loss": 5.1627,
      "mean_token_accuracy": 0.17225721776485442,
      "num_tokens": 39586021.0,
      "step": 18070
    },
    {
      "entropy": 5.40920124053955,
      "epoch": 1.7938666137356094,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046824541229420583,
      "loss": 5.1206,
      "mean_token_accuracy": 0.1672231838107109,
      "num_tokens": 39596333.0,
      "step": 18075
    },
    {
      "entropy": 5.350096940994263,
      "epoch": 1.7943628423977769,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046822726202217825,
      "loss": 5.1419,
      "mean_token_accuracy": 0.17159661501646042,
      "num_tokens": 39607553.0,
      "step": 18080
    },
    {
      "entropy": 5.37385573387146,
      "epoch": 1.7948590710599444,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000468209106958531,
      "loss": 5.2137,
      "mean_token_accuracy": 0.16523204147815704,
      "num_tokens": 39619700.0,
      "step": 18085
    },
    {
      "entropy": 5.391466236114502,
      "epoch": 1.7953552997221118,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004681909471037142,
      "loss": 5.1249,
      "mean_token_accuracy": 0.16672007143497466,
      "num_tokens": 39630623.0,
      "step": 18090
    },
    {
      "entropy": 5.443663644790649,
      "epoch": 1.7958515283842795,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046817278245817815,
      "loss": 5.1278,
      "mean_token_accuracy": 0.16728514879941941,
      "num_tokens": 39641089.0,
      "step": 18095
    },
    {
      "entropy": 5.401769399642944,
      "epoch": 1.796347757046447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004681546130223735,
      "loss": 5.1789,
      "mean_token_accuracy": 0.16529291421175002,
      "num_tokens": 39652381.0,
      "step": 18100
    },
    {
      "entropy": 5.317348194122315,
      "epoch": 1.7968439857086145,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004681364387967505,
      "loss": 5.1028,
      "mean_token_accuracy": 0.16514031141996383,
      "num_tokens": 39662536.0,
      "step": 18105
    },
    {
      "entropy": 5.415900421142578,
      "epoch": 1.7973402143707822,
      "grad_norm": 1.125,
      "learning_rate": 0.0004681182597817601,
      "loss": 5.1532,
      "mean_token_accuracy": 0.16876110136508943,
      "num_tokens": 39673140.0,
      "step": 18110
    },
    {
      "entropy": 5.365043210983276,
      "epoch": 1.7978364430329496,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004681000759778529,
      "loss": 5.1557,
      "mean_token_accuracy": 0.16023166626691818,
      "num_tokens": 39683843.0,
      "step": 18115
    },
    {
      "entropy": 5.394741153717041,
      "epoch": 1.7983326716951171,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004680818873854799,
      "loss": 5.1168,
      "mean_token_accuracy": 0.1647314667701721,
      "num_tokens": 39694178.0,
      "step": 18120
    },
    {
      "entropy": 5.325719594955444,
      "epoch": 1.7988289003572846,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004680636940050922,
      "loss": 5.0975,
      "mean_token_accuracy": 0.173685322701931,
      "num_tokens": 39704818.0,
      "step": 18125
    },
    {
      "entropy": 5.368074655532837,
      "epoch": 1.799325129019452,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000468045495837141,
      "loss": 5.1185,
      "mean_token_accuracy": 0.16064599603414537,
      "num_tokens": 39715599.0,
      "step": 18130
    },
    {
      "entropy": 5.414115524291992,
      "epoch": 1.7998213576816195,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046802729288207735,
      "loss": 5.1291,
      "mean_token_accuracy": 0.16967869102954863,
      "num_tokens": 39725736.0,
      "step": 18135
    },
    {
      "entropy": 5.4021917343139645,
      "epoch": 1.8003175863437872,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046800908514035285,
      "loss": 5.183,
      "mean_token_accuracy": 0.1642035111784935,
      "num_tokens": 39735736.0,
      "step": 18140
    },
    {
      "entropy": 5.437336540222168,
      "epoch": 1.8008138150059547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046799087261241884,
      "loss": 5.2077,
      "mean_token_accuracy": 0.16382314562797545,
      "num_tokens": 39747443.0,
      "step": 18145
    },
    {
      "entropy": 5.442257595062256,
      "epoch": 1.8013100436681224,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004679726552987271,
      "loss": 5.1669,
      "mean_token_accuracy": 0.16545602530241013,
      "num_tokens": 39758128.0,
      "step": 18150
    },
    {
      "entropy": 5.401567220687866,
      "epoch": 1.80180627233029,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004679544331997294,
      "loss": 5.0665,
      "mean_token_accuracy": 0.17113794535398483,
      "num_tokens": 39769153.0,
      "step": 18155
    },
    {
      "entropy": 5.452924585342407,
      "epoch": 1.8023025009924574,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046793620631587743,
      "loss": 5.1666,
      "mean_token_accuracy": 0.1582968309521675,
      "num_tokens": 39779666.0,
      "step": 18160
    },
    {
      "entropy": 5.382218503952027,
      "epoch": 1.8027987296546248,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046791797464762334,
      "loss": 5.1985,
      "mean_token_accuracy": 0.15833325535058976,
      "num_tokens": 39790329.0,
      "step": 18165
    },
    {
      "entropy": 5.446792364120483,
      "epoch": 1.8032949583167923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046789973819541915,
      "loss": 5.189,
      "mean_token_accuracy": 0.1660532683134079,
      "num_tokens": 39801996.0,
      "step": 18170
    },
    {
      "entropy": 5.476498317718506,
      "epoch": 1.8037911869789598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004678814969597171,
      "loss": 5.1735,
      "mean_token_accuracy": 0.16243309825658797,
      "num_tokens": 39814064.0,
      "step": 18175
    },
    {
      "entropy": 5.406902027130127,
      "epoch": 1.8042874156411273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004678632509409696,
      "loss": 5.1081,
      "mean_token_accuracy": 0.1676637589931488,
      "num_tokens": 39825369.0,
      "step": 18180
    },
    {
      "entropy": 5.322072076797485,
      "epoch": 1.804783644303295,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000467845000139629,
      "loss": 5.1851,
      "mean_token_accuracy": 0.17084823697805404,
      "num_tokens": 39835218.0,
      "step": 18185
    },
    {
      "entropy": 5.460767459869385,
      "epoch": 1.8052798729654624,
      "grad_norm": 1.09375,
      "learning_rate": 0.000467826744556148,
      "loss": 5.2164,
      "mean_token_accuracy": 0.16122455894947052,
      "num_tokens": 39844706.0,
      "step": 18190
    },
    {
      "entropy": 5.368934154510498,
      "epoch": 1.8057761016276301,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004678084841909792,
      "loss": 5.1404,
      "mean_token_accuracy": 0.16879056245088578,
      "num_tokens": 39856348.0,
      "step": 18195
    },
    {
      "entropy": 5.360684204101562,
      "epoch": 1.8062723302897976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046779021904457555,
      "loss": 5.1887,
      "mean_token_accuracy": 0.16451582312583923,
      "num_tokens": 39866850.0,
      "step": 18200
    },
    {
      "entropy": 5.472676706314087,
      "epoch": 1.806768558951965,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004677719491173898,
      "loss": 5.2221,
      "mean_token_accuracy": 0.1683759406208992,
      "num_tokens": 39877647.0,
      "step": 18205
    },
    {
      "entropy": 5.432892465591431,
      "epoch": 1.8072647876141326,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046775367440987506,
      "loss": 5.249,
      "mean_token_accuracy": 0.16596645563840867,
      "num_tokens": 39889172.0,
      "step": 18210
    },
    {
      "entropy": 5.465551805496216,
      "epoch": 1.8077610162763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004677353949224846,
      "loss": 5.1645,
      "mean_token_accuracy": 0.1684268206357956,
      "num_tokens": 39900445.0,
      "step": 18215
    },
    {
      "entropy": 5.402951240539551,
      "epoch": 1.8082572449384675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046771711065567167,
      "loss": 5.1837,
      "mean_token_accuracy": 0.1618403598666191,
      "num_tokens": 39912825.0,
      "step": 18220
    },
    {
      "entropy": 5.328980350494385,
      "epoch": 1.8087534736006352,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004676988216098896,
      "loss": 5.1257,
      "mean_token_accuracy": 0.17009818702936172,
      "num_tokens": 39923370.0,
      "step": 18225
    },
    {
      "entropy": 5.466641283035278,
      "epoch": 1.8092497022628027,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004676805277855921,
      "loss": 5.1974,
      "mean_token_accuracy": 0.16684441715478898,
      "num_tokens": 39934724.0,
      "step": 18230
    },
    {
      "entropy": 5.419376564025879,
      "epoch": 1.8097459309249704,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004676622291832326,
      "loss": 5.1287,
      "mean_token_accuracy": 0.16061483174562455,
      "num_tokens": 39947064.0,
      "step": 18235
    },
    {
      "entropy": 5.3722724437713625,
      "epoch": 1.8102421595871379,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046764392580326506,
      "loss": 5.1146,
      "mean_token_accuracy": 0.1651623547077179,
      "num_tokens": 39957649.0,
      "step": 18240
    },
    {
      "entropy": 5.4084278583526615,
      "epoch": 1.8107383882493053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046762561764614316,
      "loss": 5.172,
      "mean_token_accuracy": 0.16282694935798644,
      "num_tokens": 39968086.0,
      "step": 18245
    },
    {
      "entropy": 5.362113285064697,
      "epoch": 1.8112346169114728,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000467607304712321,
      "loss": 5.1332,
      "mean_token_accuracy": 0.16226136684417725,
      "num_tokens": 39978794.0,
      "step": 18250
    },
    {
      "entropy": 5.4239089012146,
      "epoch": 1.8117308455736403,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046758898700225283,
      "loss": 5.1568,
      "mean_token_accuracy": 0.16683846861124038,
      "num_tokens": 39990383.0,
      "step": 18255
    },
    {
      "entropy": 5.435868167877198,
      "epoch": 1.8122270742358078,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004675706645163927,
      "loss": 5.217,
      "mean_token_accuracy": 0.15875897258520127,
      "num_tokens": 40001900.0,
      "step": 18260
    },
    {
      "entropy": 5.451136636734009,
      "epoch": 1.8127233028979752,
      "grad_norm": 1.015625,
      "learning_rate": 0.000467552337255195,
      "loss": 5.1347,
      "mean_token_accuracy": 0.15988864302635192,
      "num_tokens": 40012482.0,
      "step": 18265
    },
    {
      "entropy": 5.37092547416687,
      "epoch": 1.813219531560143,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004675340052191144,
      "loss": 5.1513,
      "mean_token_accuracy": 0.1659895345568657,
      "num_tokens": 40023264.0,
      "step": 18270
    },
    {
      "entropy": 5.425058841705322,
      "epoch": 1.8137157602223104,
      "grad_norm": 1.125,
      "learning_rate": 0.0004675156684086052,
      "loss": 5.2161,
      "mean_token_accuracy": 0.16233590096235276,
      "num_tokens": 40034754.0,
      "step": 18275
    },
    {
      "entropy": 5.376499652862549,
      "epoch": 1.814211988884478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004674973268241224,
      "loss": 5.1308,
      "mean_token_accuracy": 0.17033272236585617,
      "num_tokens": 40045781.0,
      "step": 18280
    },
    {
      "entropy": 5.455408191680908,
      "epoch": 1.8147082175466456,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004674789804661205,
      "loss": 5.2001,
      "mean_token_accuracy": 0.16479742974042894,
      "num_tokens": 40055700.0,
      "step": 18285
    },
    {
      "entropy": 5.430260467529297,
      "epoch": 1.815204446208813,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046746062933505477,
      "loss": 5.2333,
      "mean_token_accuracy": 0.16340002417564392,
      "num_tokens": 40066702.0,
      "step": 18290
    },
    {
      "entropy": 5.396597194671631,
      "epoch": 1.8157006748709805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004674422734313801,
      "loss": 5.227,
      "mean_token_accuracy": 0.16543692946434022,
      "num_tokens": 40077856.0,
      "step": 18295
    },
    {
      "entropy": 5.35215802192688,
      "epoch": 1.816196903533148,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046742391275555176,
      "loss": 5.1192,
      "mean_token_accuracy": 0.16846383213996888,
      "num_tokens": 40089205.0,
      "step": 18300
    },
    {
      "entropy": 5.462887859344482,
      "epoch": 1.8166931321953155,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000467405547308025,
      "loss": 5.1724,
      "mean_token_accuracy": 0.15907726287841797,
      "num_tokens": 40100028.0,
      "step": 18305
    },
    {
      "entropy": 5.508623123168945,
      "epoch": 1.8171893608574832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004673871770892553,
      "loss": 5.1788,
      "mean_token_accuracy": 0.1656948909163475,
      "num_tokens": 40109635.0,
      "step": 18310
    },
    {
      "entropy": 5.4259978294372555,
      "epoch": 1.8176855895196506,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046736880209969814,
      "loss": 5.1482,
      "mean_token_accuracy": 0.16405561417341233,
      "num_tokens": 40120539.0,
      "step": 18315
    },
    {
      "entropy": 5.390337514877319,
      "epoch": 1.8181818181818183,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046735042233980923,
      "loss": 5.2222,
      "mean_token_accuracy": 0.1536683276295662,
      "num_tokens": 40131738.0,
      "step": 18320
    },
    {
      "entropy": 5.497966861724853,
      "epoch": 1.8186780468439858,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046733203781004426,
      "loss": 5.2798,
      "mean_token_accuracy": 0.1619205579161644,
      "num_tokens": 40142800.0,
      "step": 18325
    },
    {
      "entropy": 5.501520586013794,
      "epoch": 1.8191742755061533,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046731364851085923,
      "loss": 5.2228,
      "mean_token_accuracy": 0.16291523724794388,
      "num_tokens": 40153992.0,
      "step": 18330
    },
    {
      "entropy": 5.427394104003906,
      "epoch": 1.8196705041683208,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004672952544427101,
      "loss": 5.1491,
      "mean_token_accuracy": 0.16663825809955596,
      "num_tokens": 40165213.0,
      "step": 18335
    },
    {
      "entropy": 5.3865800380706785,
      "epoch": 1.8201667328304882,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046727685560605305,
      "loss": 5.2202,
      "mean_token_accuracy": 0.1624307483434677,
      "num_tokens": 40176763.0,
      "step": 18340
    },
    {
      "entropy": 5.401355934143067,
      "epoch": 1.8206629614926557,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004672584520013443,
      "loss": 5.1134,
      "mean_token_accuracy": 0.16366345733404158,
      "num_tokens": 40187958.0,
      "step": 18345
    },
    {
      "entropy": 5.372914505004883,
      "epoch": 1.8211591901548232,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004672400436290402,
      "loss": 5.1148,
      "mean_token_accuracy": 0.17183471322059632,
      "num_tokens": 40197881.0,
      "step": 18350
    },
    {
      "entropy": 5.441558885574341,
      "epoch": 1.821655418816991,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046722163048959725,
      "loss": 5.1181,
      "mean_token_accuracy": 0.1720709189772606,
      "num_tokens": 40209322.0,
      "step": 18355
    },
    {
      "entropy": 5.389899826049804,
      "epoch": 1.8221516474791584,
      "grad_norm": 1.0,
      "learning_rate": 0.0004672032125834721,
      "loss": 5.1672,
      "mean_token_accuracy": 0.16839618533849715,
      "num_tokens": 40222186.0,
      "step": 18360
    },
    {
      "entropy": 5.452194786071777,
      "epoch": 1.822647876141326,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004671847899111214,
      "loss": 5.1764,
      "mean_token_accuracy": 0.16681119352579116,
      "num_tokens": 40232136.0,
      "step": 18365
    },
    {
      "entropy": 5.503888893127441,
      "epoch": 1.8231441048034935,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046716636247300206,
      "loss": 5.2337,
      "mean_token_accuracy": 0.16216229796409606,
      "num_tokens": 40244468.0,
      "step": 18370
    },
    {
      "entropy": 5.413626289367675,
      "epoch": 1.823640333465661,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046714793026957096,
      "loss": 5.1779,
      "mean_token_accuracy": 0.16283288300037385,
      "num_tokens": 40255441.0,
      "step": 18375
    },
    {
      "entropy": 5.363532400131225,
      "epoch": 1.8241365621278285,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004671294933012853,
      "loss": 5.1075,
      "mean_token_accuracy": 0.16430170834064484,
      "num_tokens": 40266109.0,
      "step": 18380
    },
    {
      "entropy": 5.375572061538696,
      "epoch": 1.824632790789996,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046711105156860223,
      "loss": 5.107,
      "mean_token_accuracy": 0.16833672523498536,
      "num_tokens": 40277635.0,
      "step": 18385
    },
    {
      "entropy": 5.430967712402344,
      "epoch": 1.8251290194521634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046709260507197896,
      "loss": 5.2692,
      "mean_token_accuracy": 0.16606980338692665,
      "num_tokens": 40288668.0,
      "step": 18390
    },
    {
      "entropy": 5.4592752933502195,
      "epoch": 1.8256252481143311,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046707415381187306,
      "loss": 5.1286,
      "mean_token_accuracy": 0.172028149664402,
      "num_tokens": 40300262.0,
      "step": 18395
    },
    {
      "entropy": 5.397436904907226,
      "epoch": 1.8261214767764986,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000467055697788742,
      "loss": 5.2042,
      "mean_token_accuracy": 0.1534188747406006,
      "num_tokens": 40310931.0,
      "step": 18400
    },
    {
      "entropy": 5.397539567947388,
      "epoch": 1.826617705438666,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046703723700304345,
      "loss": 5.1401,
      "mean_token_accuracy": 0.1639454498887062,
      "num_tokens": 40321623.0,
      "step": 18405
    },
    {
      "entropy": 5.36935453414917,
      "epoch": 1.8271139341008338,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004670187714552353,
      "loss": 5.0817,
      "mean_token_accuracy": 0.16577230393886566,
      "num_tokens": 40330951.0,
      "step": 18410
    },
    {
      "entropy": 5.422111320495605,
      "epoch": 1.8276101627630013,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046700030114577533,
      "loss": 5.1784,
      "mean_token_accuracy": 0.1704462319612503,
      "num_tokens": 40342971.0,
      "step": 18415
    },
    {
      "entropy": 5.415343284606934,
      "epoch": 1.8281063914251687,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004669818260751217,
      "loss": 5.1461,
      "mean_token_accuracy": 0.16760000288486482,
      "num_tokens": 40353837.0,
      "step": 18420
    },
    {
      "entropy": 5.436749982833862,
      "epoch": 1.8286026200873362,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004669633462437324,
      "loss": 5.2266,
      "mean_token_accuracy": 0.1674928456544876,
      "num_tokens": 40367389.0,
      "step": 18425
    },
    {
      "entropy": 5.409719514846802,
      "epoch": 1.8290988487495037,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046694486165206573,
      "loss": 5.0484,
      "mean_token_accuracy": 0.1721376210451126,
      "num_tokens": 40376793.0,
      "step": 18430
    },
    {
      "entropy": 5.404023933410644,
      "epoch": 1.8295950774116712,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004669263723005801,
      "loss": 5.1472,
      "mean_token_accuracy": 0.16912305802106858,
      "num_tokens": 40388025.0,
      "step": 18435
    },
    {
      "entropy": 5.413224840164185,
      "epoch": 1.8300913060738389,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046690787818973406,
      "loss": 5.2338,
      "mean_token_accuracy": 0.1639758750796318,
      "num_tokens": 40399731.0,
      "step": 18440
    },
    {
      "entropy": 5.462096214294434,
      "epoch": 1.8305875347360063,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046688937931998613,
      "loss": 5.2735,
      "mean_token_accuracy": 0.16211845874786376,
      "num_tokens": 40411078.0,
      "step": 18445
    },
    {
      "entropy": 5.500850009918213,
      "epoch": 1.831083763398174,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004668708756917951,
      "loss": 5.2256,
      "mean_token_accuracy": 0.16242025047540665,
      "num_tokens": 40422806.0,
      "step": 18450
    },
    {
      "entropy": 5.562418746948242,
      "epoch": 1.8315799920603415,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046685236730561975,
      "loss": 5.2878,
      "mean_token_accuracy": 0.16025808677077294,
      "num_tokens": 40434041.0,
      "step": 18455
    },
    {
      "entropy": 5.428767538070678,
      "epoch": 1.832076220722509,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046683385416191914,
      "loss": 5.1214,
      "mean_token_accuracy": 0.17167191505432128,
      "num_tokens": 40444658.0,
      "step": 18460
    },
    {
      "entropy": 5.391073513031006,
      "epoch": 1.8325724493846764,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046681533626115226,
      "loss": 5.1448,
      "mean_token_accuracy": 0.16665806323289872,
      "num_tokens": 40455745.0,
      "step": 18465
    },
    {
      "entropy": 5.405873584747314,
      "epoch": 1.833068678046844,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004667968136037784,
      "loss": 5.2208,
      "mean_token_accuracy": 0.16414207369089126,
      "num_tokens": 40468829.0,
      "step": 18470
    },
    {
      "entropy": 5.391037464141846,
      "epoch": 1.8335649067090114,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046677828619025687,
      "loss": 5.1378,
      "mean_token_accuracy": 0.16602183878421783,
      "num_tokens": 40479817.0,
      "step": 18475
    },
    {
      "entropy": 5.457935905456543,
      "epoch": 1.8340611353711789,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046675975402104706,
      "loss": 5.2333,
      "mean_token_accuracy": 0.16435811519622803,
      "num_tokens": 40490398.0,
      "step": 18480
    },
    {
      "entropy": 5.366128969192505,
      "epoch": 1.8345573640333466,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046674121709660855,
      "loss": 5.0585,
      "mean_token_accuracy": 0.17134886234998703,
      "num_tokens": 40500483.0,
      "step": 18485
    },
    {
      "entropy": 5.423874282836914,
      "epoch": 1.835053592695514,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000466722675417401,
      "loss": 5.2106,
      "mean_token_accuracy": 0.16198041737079621,
      "num_tokens": 40511494.0,
      "step": 18490
    },
    {
      "entropy": 5.412806367874145,
      "epoch": 1.8355498213576817,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046670412898388425,
      "loss": 5.1539,
      "mean_token_accuracy": 0.17454639673233033,
      "num_tokens": 40523573.0,
      "step": 18495
    },
    {
      "entropy": 5.35127592086792,
      "epoch": 1.8360460500198492,
      "grad_norm": 1.125,
      "learning_rate": 0.00046668557779651825,
      "loss": 5.2097,
      "mean_token_accuracy": 0.16208166182041167,
      "num_tokens": 40534904.0,
      "step": 18500
    },
    {
      "entropy": 5.408517599105835,
      "epoch": 1.8365422786820167,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046666702185576295,
      "loss": 5.1652,
      "mean_token_accuracy": 0.16412850320339203,
      "num_tokens": 40545053.0,
      "step": 18505
    },
    {
      "entropy": 5.440028047561645,
      "epoch": 1.8370385073441842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004666484611620785,
      "loss": 5.1005,
      "mean_token_accuracy": 0.169782792031765,
      "num_tokens": 40555305.0,
      "step": 18510
    },
    {
      "entropy": 5.357228136062622,
      "epoch": 1.8375347360063516,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004666298957159251,
      "loss": 5.1332,
      "mean_token_accuracy": 0.1635712891817093,
      "num_tokens": 40565611.0,
      "step": 18515
    },
    {
      "entropy": 5.425492429733277,
      "epoch": 1.8380309646685191,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046661132551776327,
      "loss": 5.2334,
      "mean_token_accuracy": 0.1632575049996376,
      "num_tokens": 40578321.0,
      "step": 18520
    },
    {
      "entropy": 5.4648350238800045,
      "epoch": 1.8385271933306868,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004665927505680535,
      "loss": 5.2177,
      "mean_token_accuracy": 0.17184462994337082,
      "num_tokens": 40588696.0,
      "step": 18525
    },
    {
      "entropy": 5.407544946670532,
      "epoch": 1.8390234219928543,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046657417086725625,
      "loss": 5.1595,
      "mean_token_accuracy": 0.16840251088142394,
      "num_tokens": 40599699.0,
      "step": 18530
    },
    {
      "entropy": 5.374041128158569,
      "epoch": 1.839519650655022,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004665555864158324,
      "loss": 5.1646,
      "mean_token_accuracy": 0.16536508351564408,
      "num_tokens": 40609706.0,
      "step": 18535
    },
    {
      "entropy": 5.3916419506072994,
      "epoch": 1.8400158793171895,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004665369972142428,
      "loss": 5.1949,
      "mean_token_accuracy": 0.17001994848251342,
      "num_tokens": 40619431.0,
      "step": 18540
    },
    {
      "entropy": 5.460467910766601,
      "epoch": 1.840512107979357,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046651840326294845,
      "loss": 5.1587,
      "mean_token_accuracy": 0.16686232537031173,
      "num_tokens": 40629904.0,
      "step": 18545
    },
    {
      "entropy": 5.377860736846924,
      "epoch": 1.8410083366415244,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004664998045624104,
      "loss": 5.213,
      "mean_token_accuracy": 0.16523574143648148,
      "num_tokens": 40641266.0,
      "step": 18550
    },
    {
      "entropy": 5.359386301040649,
      "epoch": 1.8415045653036919,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004664812011130897,
      "loss": 5.1457,
      "mean_token_accuracy": 0.16460828781127929,
      "num_tokens": 40652454.0,
      "step": 18555
    },
    {
      "entropy": 5.4140345573425295,
      "epoch": 1.8420007939658594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046646259291544787,
      "loss": 5.1307,
      "mean_token_accuracy": 0.16847693175077438,
      "num_tokens": 40664311.0,
      "step": 18560
    },
    {
      "entropy": 5.36181526184082,
      "epoch": 1.8424970226280268,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004664439799699463,
      "loss": 5.0691,
      "mean_token_accuracy": 0.17100065797567368,
      "num_tokens": 40675241.0,
      "step": 18565
    },
    {
      "entropy": 5.370873498916626,
      "epoch": 1.8429932512901945,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046642536227704665,
      "loss": 5.1409,
      "mean_token_accuracy": 0.16570040881633757,
      "num_tokens": 40685972.0,
      "step": 18570
    },
    {
      "entropy": 5.391624116897583,
      "epoch": 1.843489479952362,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046640673983721037,
      "loss": 5.1477,
      "mean_token_accuracy": 0.16798866391181946,
      "num_tokens": 40696967.0,
      "step": 18575
    },
    {
      "entropy": 5.42134919166565,
      "epoch": 1.8439857086145297,
      "grad_norm": 1.0,
      "learning_rate": 0.00046638811265089944,
      "loss": 5.1558,
      "mean_token_accuracy": 0.1621604859828949,
      "num_tokens": 40707710.0,
      "step": 18580
    },
    {
      "entropy": 5.490949726104736,
      "epoch": 1.8444819372766972,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004663694807185758,
      "loss": 5.3064,
      "mean_token_accuracy": 0.16239925920963288,
      "num_tokens": 40718586.0,
      "step": 18585
    },
    {
      "entropy": 5.468781280517578,
      "epoch": 1.8449781659388647,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046635084404070135,
      "loss": 5.1708,
      "mean_token_accuracy": 0.17167565673589708,
      "num_tokens": 40729792.0,
      "step": 18590
    },
    {
      "entropy": 5.496552848815918,
      "epoch": 1.8454743946010321,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004663322026177382,
      "loss": 5.2003,
      "mean_token_accuracy": 0.16058867126703263,
      "num_tokens": 40740974.0,
      "step": 18595
    },
    {
      "entropy": 5.343339681625366,
      "epoch": 1.8459706232631996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046631355645014875,
      "loss": 5.0152,
      "mean_token_accuracy": 0.17383771240711213,
      "num_tokens": 40750208.0,
      "step": 18600
    },
    {
      "entropy": 5.387039852142334,
      "epoch": 1.846466851925367,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046629490553839537,
      "loss": 5.1266,
      "mean_token_accuracy": 0.16721779108047485,
      "num_tokens": 40762163.0,
      "step": 18605
    },
    {
      "entropy": 5.415376281738281,
      "epoch": 1.8469630805875348,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004662762498829405,
      "loss": 5.2381,
      "mean_token_accuracy": 0.1660057410597801,
      "num_tokens": 40773028.0,
      "step": 18610
    },
    {
      "entropy": 5.480859661102295,
      "epoch": 1.8474593092497023,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004662575894842469,
      "loss": 5.2113,
      "mean_token_accuracy": 0.15883186906576158,
      "num_tokens": 40784939.0,
      "step": 18615
    },
    {
      "entropy": 5.451727342605591,
      "epoch": 1.84795553791187,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00046623892434277713,
      "loss": 5.1817,
      "mean_token_accuracy": 0.16468430161476136,
      "num_tokens": 40796170.0,
      "step": 18620
    },
    {
      "entropy": 5.394607257843018,
      "epoch": 1.8484517665740374,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004662202544589941,
      "loss": 5.1151,
      "mean_token_accuracy": 0.17095791101455687,
      "num_tokens": 40807184.0,
      "step": 18625
    },
    {
      "entropy": 5.382159185409546,
      "epoch": 1.848947995236205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046620157983336087,
      "loss": 5.1154,
      "mean_token_accuracy": 0.16693626642227172,
      "num_tokens": 40818474.0,
      "step": 18630
    },
    {
      "entropy": 5.420678567886353,
      "epoch": 1.8494442238983724,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004661829004663404,
      "loss": 5.1692,
      "mean_token_accuracy": 0.16465787440538407,
      "num_tokens": 40828789.0,
      "step": 18635
    },
    {
      "entropy": 5.387687540054321,
      "epoch": 1.8499404525605398,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004661642163583961,
      "loss": 5.1258,
      "mean_token_accuracy": 0.17276198416948318,
      "num_tokens": 40840564.0,
      "step": 18640
    },
    {
      "entropy": 5.347279262542725,
      "epoch": 1.8504366812227073,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046614552750999093,
      "loss": 5.1046,
      "mean_token_accuracy": 0.16814005821943284,
      "num_tokens": 40852528.0,
      "step": 18645
    },
    {
      "entropy": 5.399473237991333,
      "epoch": 1.8509329098848748,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004661268339215887,
      "loss": 5.1646,
      "mean_token_accuracy": 0.17033719420433044,
      "num_tokens": 40863242.0,
      "step": 18650
    },
    {
      "entropy": 5.374508619308472,
      "epoch": 1.8514291385470425,
      "grad_norm": 1.125,
      "learning_rate": 0.0004661081355936528,
      "loss": 5.0991,
      "mean_token_accuracy": 0.17537446767091752,
      "num_tokens": 40873721.0,
      "step": 18655
    },
    {
      "entropy": 5.450068187713623,
      "epoch": 1.85192536720921,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000466089432526647,
      "loss": 5.2023,
      "mean_token_accuracy": 0.16592039912939072,
      "num_tokens": 40884921.0,
      "step": 18660
    },
    {
      "entropy": 5.386263799667359,
      "epoch": 1.8524215958713777,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046607072472103493,
      "loss": 5.0759,
      "mean_token_accuracy": 0.16983017474412918,
      "num_tokens": 40896243.0,
      "step": 18665
    },
    {
      "entropy": 5.363033246994019,
      "epoch": 1.8529178245335451,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046605201217728067,
      "loss": 5.2107,
      "mean_token_accuracy": 0.16299648135900496,
      "num_tokens": 40907191.0,
      "step": 18670
    },
    {
      "entropy": 5.431312894821167,
      "epoch": 1.8534140531957126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046603329489584817,
      "loss": 5.1694,
      "mean_token_accuracy": 0.17209880352020263,
      "num_tokens": 40918737.0,
      "step": 18675
    },
    {
      "entropy": 5.4370763301849365,
      "epoch": 1.85391028185788,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004660145728772016,
      "loss": 5.166,
      "mean_token_accuracy": 0.16712798327207565,
      "num_tokens": 40930184.0,
      "step": 18680
    },
    {
      "entropy": 5.356495904922485,
      "epoch": 1.8544065105200476,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004659958461218052,
      "loss": 5.0881,
      "mean_token_accuracy": 0.1707771897315979,
      "num_tokens": 40941483.0,
      "step": 18685
    },
    {
      "entropy": 5.3664428234100345,
      "epoch": 1.854902739182215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046597711463012344,
      "loss": 5.1834,
      "mean_token_accuracy": 0.16627907156944274,
      "num_tokens": 40952392.0,
      "step": 18690
    },
    {
      "entropy": 5.350558614730835,
      "epoch": 1.8553989678443827,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004659583784026207,
      "loss": 5.0924,
      "mean_token_accuracy": 0.16871025860309602,
      "num_tokens": 40962539.0,
      "step": 18695
    },
    {
      "entropy": 5.34281005859375,
      "epoch": 1.8558951965065502,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046593963743976163,
      "loss": 5.0396,
      "mean_token_accuracy": 0.17704451084136963,
      "num_tokens": 40972594.0,
      "step": 18700
    },
    {
      "entropy": 5.451272678375244,
      "epoch": 1.8563914251687177,
      "grad_norm": 1.046875,
      "learning_rate": 0.000465920891742011,
      "loss": 5.1555,
      "mean_token_accuracy": 0.16926819384098052,
      "num_tokens": 40983460.0,
      "step": 18705
    },
    {
      "entropy": 5.415436124801635,
      "epoch": 1.8568876538308854,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004659021413098337,
      "loss": 5.1491,
      "mean_token_accuracy": 0.16559404134750366,
      "num_tokens": 40993945.0,
      "step": 18710
    },
    {
      "entropy": 5.327545976638794,
      "epoch": 1.8573838824930529,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046588338614369465,
      "loss": 5.1522,
      "mean_token_accuracy": 0.1681218445301056,
      "num_tokens": 41005781.0,
      "step": 18715
    },
    {
      "entropy": 5.448002290725708,
      "epoch": 1.8578801111552203,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046586462624405897,
      "loss": 5.2562,
      "mean_token_accuracy": 0.1657397210597992,
      "num_tokens": 41017650.0,
      "step": 18720
    },
    {
      "entropy": 5.451017570495606,
      "epoch": 1.8583763398173878,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004658458616113918,
      "loss": 5.1198,
      "mean_token_accuracy": 0.16796240210533142,
      "num_tokens": 41029196.0,
      "step": 18725
    },
    {
      "entropy": 5.407769012451172,
      "epoch": 1.8588725684795553,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004658270922461585,
      "loss": 5.1974,
      "mean_token_accuracy": 0.17067907452583314,
      "num_tokens": 41040204.0,
      "step": 18730
    },
    {
      "entropy": 5.3872322082519535,
      "epoch": 1.8593687971417228,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046580831814882446,
      "loss": 5.1762,
      "mean_token_accuracy": 0.1617768958210945,
      "num_tokens": 41050507.0,
      "step": 18735
    },
    {
      "entropy": 5.318296480178833,
      "epoch": 1.8598650258038905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046578953931985536,
      "loss": 5.1166,
      "mean_token_accuracy": 0.1652835875749588,
      "num_tokens": 41062001.0,
      "step": 18740
    },
    {
      "entropy": 5.426076459884643,
      "epoch": 1.860361254466058,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004657707557597167,
      "loss": 5.1648,
      "mean_token_accuracy": 0.16049816310405732,
      "num_tokens": 41073097.0,
      "step": 18745
    },
    {
      "entropy": 5.429980707168579,
      "epoch": 1.8608574831282256,
      "grad_norm": 1.125,
      "learning_rate": 0.00046575196746887447,
      "loss": 5.1746,
      "mean_token_accuracy": 0.16867667585611343,
      "num_tokens": 41083883.0,
      "step": 18750
    },
    {
      "entropy": 5.376496982574463,
      "epoch": 1.861353711790393,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046573317444779445,
      "loss": 5.1406,
      "mean_token_accuracy": 0.17066068947315216,
      "num_tokens": 41095506.0,
      "step": 18755
    },
    {
      "entropy": 5.421112060546875,
      "epoch": 1.8618499404525606,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004657143766969427,
      "loss": 5.1728,
      "mean_token_accuracy": 0.16679287403821946,
      "num_tokens": 41106568.0,
      "step": 18760
    },
    {
      "entropy": 5.383732891082763,
      "epoch": 1.862346169114728,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004656955742167853,
      "loss": 5.0748,
      "mean_token_accuracy": 0.16627056151628494,
      "num_tokens": 41117661.0,
      "step": 18765
    },
    {
      "entropy": 5.413411569595337,
      "epoch": 1.8628423977768955,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046567676700778865,
      "loss": 5.1694,
      "mean_token_accuracy": 0.1677068516612053,
      "num_tokens": 41129209.0,
      "step": 18770
    },
    {
      "entropy": 5.400642395019531,
      "epoch": 1.863338626439063,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046565795507041904,
      "loss": 5.2277,
      "mean_token_accuracy": 0.16497256606817245,
      "num_tokens": 41140715.0,
      "step": 18775
    },
    {
      "entropy": 5.447815322875977,
      "epoch": 1.8638348551012305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000465639138405143,
      "loss": 5.1901,
      "mean_token_accuracy": 0.16200560778379441,
      "num_tokens": 41151080.0,
      "step": 18780
    },
    {
      "entropy": 5.469663667678833,
      "epoch": 1.8643310837633982,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046562031701242703,
      "loss": 5.2028,
      "mean_token_accuracy": 0.16844283640384675,
      "num_tokens": 41162816.0,
      "step": 18785
    },
    {
      "entropy": 5.400076484680175,
      "epoch": 1.8648273124255657,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046560149089273803,
      "loss": 5.0932,
      "mean_token_accuracy": 0.17522107064723969,
      "num_tokens": 41173456.0,
      "step": 18790
    },
    {
      "entropy": 5.338046026229859,
      "epoch": 1.8653235410877333,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004655826600465427,
      "loss": 5.0537,
      "mean_token_accuracy": 0.16941849291324615,
      "num_tokens": 41184436.0,
      "step": 18795
    },
    {
      "entropy": 5.323607921600342,
      "epoch": 1.8658197697499008,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046556382447430816,
      "loss": 5.12,
      "mean_token_accuracy": 0.1718974933028221,
      "num_tokens": 41196417.0,
      "step": 18800
    },
    {
      "entropy": 5.390220594406128,
      "epoch": 1.8663159984120683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046554498417650134,
      "loss": 5.1418,
      "mean_token_accuracy": 0.1715352639555931,
      "num_tokens": 41206054.0,
      "step": 18805
    },
    {
      "entropy": 5.374078845977783,
      "epoch": 1.8668122270742358,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004655261391535895,
      "loss": 5.1413,
      "mean_token_accuracy": 0.1630064219236374,
      "num_tokens": 41217002.0,
      "step": 18810
    },
    {
      "entropy": 5.5307117938995365,
      "epoch": 1.8673084557364032,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004655072894060399,
      "loss": 5.2968,
      "mean_token_accuracy": 0.15711189806461334,
      "num_tokens": 41228485.0,
      "step": 18815
    },
    {
      "entropy": 5.4521646976470945,
      "epoch": 1.8678046843985707,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004654884349343201,
      "loss": 5.1899,
      "mean_token_accuracy": 0.1662985786795616,
      "num_tokens": 41240600.0,
      "step": 18820
    },
    {
      "entropy": 5.4902551651000975,
      "epoch": 1.8683009130607384,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046546957573889757,
      "loss": 5.2595,
      "mean_token_accuracy": 0.156369099766016,
      "num_tokens": 41251870.0,
      "step": 18825
    },
    {
      "entropy": 5.3625129699707035,
      "epoch": 1.868797141722906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046545071182024,
      "loss": 5.1677,
      "mean_token_accuracy": 0.16149144023656845,
      "num_tokens": 41264679.0,
      "step": 18830
    },
    {
      "entropy": 5.5253190994262695,
      "epoch": 1.8692933703850736,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046543184317881516,
      "loss": 5.2416,
      "mean_token_accuracy": 0.1671920031309128,
      "num_tokens": 41275690.0,
      "step": 18835
    },
    {
      "entropy": 5.437155055999756,
      "epoch": 1.869789599047241,
      "grad_norm": 1.203125,
      "learning_rate": 0.000465412969815091,
      "loss": 5.121,
      "mean_token_accuracy": 0.17339979410171508,
      "num_tokens": 41285118.0,
      "step": 18840
    },
    {
      "entropy": 5.349375247955322,
      "epoch": 1.8702858277094085,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004653940917295354,
      "loss": 5.1333,
      "mean_token_accuracy": 0.17048416882753373,
      "num_tokens": 41296330.0,
      "step": 18845
    },
    {
      "entropy": 5.339487552642822,
      "epoch": 1.870782056371576,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004653752089226166,
      "loss": 5.1567,
      "mean_token_accuracy": 0.16312770098447799,
      "num_tokens": 41307051.0,
      "step": 18850
    },
    {
      "entropy": 5.42032299041748,
      "epoch": 1.8712782850337435,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004653563213948029,
      "loss": 5.0881,
      "mean_token_accuracy": 0.17338867336511612,
      "num_tokens": 41317111.0,
      "step": 18855
    },
    {
      "entropy": 5.378041744232178,
      "epoch": 1.871774513695911,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046533742914656254,
      "loss": 5.0999,
      "mean_token_accuracy": 0.16604795902967454,
      "num_tokens": 41327509.0,
      "step": 18860
    },
    {
      "entropy": 5.3957679748535154,
      "epoch": 1.8722707423580784,
      "grad_norm": 1.125,
      "learning_rate": 0.00046531853217836414,
      "loss": 5.1832,
      "mean_token_accuracy": 0.16426987648010255,
      "num_tokens": 41338721.0,
      "step": 18865
    },
    {
      "entropy": 5.468838930130005,
      "epoch": 1.8727669710202461,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046529963049067623,
      "loss": 5.2114,
      "mean_token_accuracy": 0.16434373259544371,
      "num_tokens": 41349767.0,
      "step": 18870
    },
    {
      "entropy": 5.514481019973755,
      "epoch": 1.8732631996824136,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046528072408396747,
      "loss": 5.3259,
      "mean_token_accuracy": 0.15720009356737136,
      "num_tokens": 41361151.0,
      "step": 18875
    },
    {
      "entropy": 5.442857074737549,
      "epoch": 1.8737594283445813,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004652618129587069,
      "loss": 5.1178,
      "mean_token_accuracy": 0.1624073013663292,
      "num_tokens": 41372875.0,
      "step": 18880
    },
    {
      "entropy": 5.448298692703247,
      "epoch": 1.8742556570067488,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046524289711536327,
      "loss": 5.193,
      "mean_token_accuracy": 0.168617345392704,
      "num_tokens": 41383284.0,
      "step": 18885
    },
    {
      "entropy": 5.380942106246948,
      "epoch": 1.8747518856689163,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004652239765544057,
      "loss": 5.1604,
      "mean_token_accuracy": 0.1694989785552025,
      "num_tokens": 41395389.0,
      "step": 18890
    },
    {
      "entropy": 5.335462045669556,
      "epoch": 1.8752481143310837,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046520505127630346,
      "loss": 5.0995,
      "mean_token_accuracy": 0.16276436299085617,
      "num_tokens": 41406740.0,
      "step": 18895
    },
    {
      "entropy": 5.354051017761231,
      "epoch": 1.8757443429932512,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046518612128152576,
      "loss": 5.1126,
      "mean_token_accuracy": 0.16749123036861419,
      "num_tokens": 41417279.0,
      "step": 18900
    },
    {
      "entropy": 5.3802313804626465,
      "epoch": 1.8762405716554187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004651671865705421,
      "loss": 5.0879,
      "mean_token_accuracy": 0.16936087757349014,
      "num_tokens": 41428574.0,
      "step": 18905
    },
    {
      "entropy": 5.3551756858825685,
      "epoch": 1.8767368003175864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046514824714382204,
      "loss": 5.0749,
      "mean_token_accuracy": 0.16711199134588242,
      "num_tokens": 41438784.0,
      "step": 18910
    },
    {
      "entropy": 5.379175186157227,
      "epoch": 1.8772330289797539,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004651293030018351,
      "loss": 5.136,
      "mean_token_accuracy": 0.1645384058356285,
      "num_tokens": 41448481.0,
      "step": 18915
    },
    {
      "entropy": 5.453579616546631,
      "epoch": 1.8777292576419216,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046511035414505115,
      "loss": 5.1758,
      "mean_token_accuracy": 0.1668061301112175,
      "num_tokens": 41459097.0,
      "step": 18920
    },
    {
      "entropy": 5.493308258056641,
      "epoch": 1.878225486304089,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004650914005739401,
      "loss": 5.2134,
      "mean_token_accuracy": 0.16398849487304687,
      "num_tokens": 41469647.0,
      "step": 18925
    },
    {
      "entropy": 5.3861308097839355,
      "epoch": 1.8787217149662565,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004650724422889719,
      "loss": 5.1327,
      "mean_token_accuracy": 0.16541034877300262,
      "num_tokens": 41479826.0,
      "step": 18930
    },
    {
      "entropy": 5.427085256576538,
      "epoch": 1.879217943628424,
      "grad_norm": 1.125,
      "learning_rate": 0.0004650534792906167,
      "loss": 5.1143,
      "mean_token_accuracy": 0.16905913800001143,
      "num_tokens": 41490551.0,
      "step": 18935
    },
    {
      "entropy": 5.3681549549102785,
      "epoch": 1.8797141722905915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004650345115793447,
      "loss": 5.1446,
      "mean_token_accuracy": 0.17281076163053513,
      "num_tokens": 41500908.0,
      "step": 18940
    },
    {
      "entropy": 5.348565864562988,
      "epoch": 1.880210400952759,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046501553915562644,
      "loss": 5.112,
      "mean_token_accuracy": 0.16506529748439788,
      "num_tokens": 41511311.0,
      "step": 18945
    },
    {
      "entropy": 5.3872819423675535,
      "epoch": 1.8807066296149264,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046499656201993213,
      "loss": 5.1457,
      "mean_token_accuracy": 0.17088394314050676,
      "num_tokens": 41522068.0,
      "step": 18950
    },
    {
      "entropy": 5.410966157913208,
      "epoch": 1.881202858277094,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046497758017273255,
      "loss": 5.153,
      "mean_token_accuracy": 0.1620612159371376,
      "num_tokens": 41534099.0,
      "step": 18955
    },
    {
      "entropy": 5.447079133987427,
      "epoch": 1.8816990869392616,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046495859361449834,
      "loss": 5.2169,
      "mean_token_accuracy": 0.15949227809906005,
      "num_tokens": 41545870.0,
      "step": 18960
    },
    {
      "entropy": 5.4253558158874515,
      "epoch": 1.8821953156014293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004649396023457003,
      "loss": 5.1879,
      "mean_token_accuracy": 0.16722863018512726,
      "num_tokens": 41558175.0,
      "step": 18965
    },
    {
      "entropy": 5.38862829208374,
      "epoch": 1.8826915442635968,
      "grad_norm": 1.125,
      "learning_rate": 0.0004649206063668094,
      "loss": 5.1627,
      "mean_token_accuracy": 0.17412592470645905,
      "num_tokens": 41569326.0,
      "step": 18970
    },
    {
      "entropy": 5.401979541778564,
      "epoch": 1.8831877729257642,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004649016056782968,
      "loss": 5.1182,
      "mean_token_accuracy": 0.17110010236501694,
      "num_tokens": 41580065.0,
      "step": 18975
    },
    {
      "entropy": 5.479847145080567,
      "epoch": 1.8836840015879317,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004648826002806334,
      "loss": 5.2627,
      "mean_token_accuracy": 0.16142062097787857,
      "num_tokens": 41590915.0,
      "step": 18980
    },
    {
      "entropy": 5.34843373298645,
      "epoch": 1.8841802302500992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004648635901742909,
      "loss": 5.0971,
      "mean_token_accuracy": 0.16913120746612548,
      "num_tokens": 41602096.0,
      "step": 18985
    },
    {
      "entropy": 5.406638288497925,
      "epoch": 1.8846764589122667,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004648445753597403,
      "loss": 5.2049,
      "mean_token_accuracy": 0.1625268653035164,
      "num_tokens": 41612589.0,
      "step": 18990
    },
    {
      "entropy": 5.422667407989502,
      "epoch": 1.8851726875744343,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046482555583745336,
      "loss": 5.1015,
      "mean_token_accuracy": 0.16778650134801865,
      "num_tokens": 41624163.0,
      "step": 18995
    },
    {
      "entropy": 5.403867864608765,
      "epoch": 1.8856689162366018,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004648065316079017,
      "loss": 5.1,
      "mean_token_accuracy": 0.17351641952991487,
      "num_tokens": 41635919.0,
      "step": 19000
    },
    {
      "entropy": 5.382723999023438,
      "epoch": 1.8861651448987695,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046478750267155703,
      "loss": 5.1215,
      "mean_token_accuracy": 0.16541707664728164,
      "num_tokens": 41644723.0,
      "step": 19005
    },
    {
      "entropy": 5.374495887756348,
      "epoch": 1.886661373560937,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046476846902889125,
      "loss": 5.1865,
      "mean_token_accuracy": 0.16860900521278382,
      "num_tokens": 41657137.0,
      "step": 19010
    },
    {
      "entropy": 5.416308784484864,
      "epoch": 1.8871576022231045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004647494306803764,
      "loss": 5.1672,
      "mean_token_accuracy": 0.16580863744020463,
      "num_tokens": 41667773.0,
      "step": 19015
    },
    {
      "entropy": 5.5084234237670895,
      "epoch": 1.887653830885272,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046473038762648446,
      "loss": 5.2445,
      "mean_token_accuracy": 0.16199248433113098,
      "num_tokens": 41680549.0,
      "step": 19020
    },
    {
      "entropy": 5.430871057510376,
      "epoch": 1.8881500595474394,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046471133986768784,
      "loss": 5.1289,
      "mean_token_accuracy": 0.17128589749336243,
      "num_tokens": 41689700.0,
      "step": 19025
    },
    {
      "entropy": 5.431085634231567,
      "epoch": 1.888646288209607,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046469228740445874,
      "loss": 5.1798,
      "mean_token_accuracy": 0.1637709468603134,
      "num_tokens": 41699734.0,
      "step": 19030
    },
    {
      "entropy": 5.40373682975769,
      "epoch": 1.8891425168717744,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046467323023726965,
      "loss": 5.1804,
      "mean_token_accuracy": 0.16317180693149566,
      "num_tokens": 41710644.0,
      "step": 19035
    },
    {
      "entropy": 5.350886344909668,
      "epoch": 1.889638745533942,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046465416836659313,
      "loss": 5.0793,
      "mean_token_accuracy": 0.1752592921257019,
      "num_tokens": 41721416.0,
      "step": 19040
    },
    {
      "entropy": 5.401343917846679,
      "epoch": 1.8901349741961095,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000464635101792902,
      "loss": 5.0905,
      "mean_token_accuracy": 0.1672171950340271,
      "num_tokens": 41732139.0,
      "step": 19045
    },
    {
      "entropy": 5.393042182922363,
      "epoch": 1.8906312028582772,
      "grad_norm": 1.125,
      "learning_rate": 0.00046461603051666883,
      "loss": 5.1957,
      "mean_token_accuracy": 0.1647384822368622,
      "num_tokens": 41744233.0,
      "step": 19050
    },
    {
      "entropy": 5.398534440994263,
      "epoch": 1.8911274315204447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046459695453836676,
      "loss": 5.1147,
      "mean_token_accuracy": 0.17313210368156434,
      "num_tokens": 41755514.0,
      "step": 19055
    },
    {
      "entropy": 5.424861288070678,
      "epoch": 1.8916236601826122,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004645778738584688,
      "loss": 5.1361,
      "mean_token_accuracy": 0.17025094032287597,
      "num_tokens": 41765613.0,
      "step": 19060
    },
    {
      "entropy": 5.438427972793579,
      "epoch": 1.8921198888447797,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004645587884774481,
      "loss": 5.1826,
      "mean_token_accuracy": 0.1657859668135643,
      "num_tokens": 41777294.0,
      "step": 19065
    },
    {
      "entropy": 5.3604323387146,
      "epoch": 1.8926161175069471,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004645396983957778,
      "loss": 5.0856,
      "mean_token_accuracy": 0.17218597680330278,
      "num_tokens": 41788278.0,
      "step": 19070
    },
    {
      "entropy": 5.34559998512268,
      "epoch": 1.8931123461691146,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046452060361393147,
      "loss": 5.1517,
      "mean_token_accuracy": 0.16252499669790268,
      "num_tokens": 41798663.0,
      "step": 19075
    },
    {
      "entropy": 5.396477508544922,
      "epoch": 1.8936085748312823,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004645015041323825,
      "loss": 5.0993,
      "mean_token_accuracy": 0.17545172423124314,
      "num_tokens": 41808711.0,
      "step": 19080
    },
    {
      "entropy": 5.354438591003418,
      "epoch": 1.8941048034934498,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046448239995160455,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1623734638094902,
      "num_tokens": 41819679.0,
      "step": 19085
    },
    {
      "entropy": 5.410759019851684,
      "epoch": 1.8946010321556173,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004644632910720715,
      "loss": 5.1356,
      "mean_token_accuracy": 0.1658712774515152,
      "num_tokens": 41829649.0,
      "step": 19090
    },
    {
      "entropy": 5.4326952457427975,
      "epoch": 1.895097260817785,
      "grad_norm": 1.03125,
      "learning_rate": 0.000464444177494257,
      "loss": 5.191,
      "mean_token_accuracy": 0.16230482310056688,
      "num_tokens": 41842523.0,
      "step": 19095
    },
    {
      "entropy": 5.414219093322754,
      "epoch": 1.8955934894799524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004644250592186351,
      "loss": 5.1583,
      "mean_token_accuracy": 0.16640834957361222,
      "num_tokens": 41854514.0,
      "step": 19100
    },
    {
      "entropy": 5.432340860366821,
      "epoch": 1.89608971814212,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046440593624568,
      "loss": 5.1345,
      "mean_token_accuracy": 0.17103491723537445,
      "num_tokens": 41865166.0,
      "step": 19105
    },
    {
      "entropy": 5.428585720062256,
      "epoch": 1.8965859468042874,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004643868085758657,
      "loss": 5.0862,
      "mean_token_accuracy": 0.16951877623796463,
      "num_tokens": 41877264.0,
      "step": 19110
    },
    {
      "entropy": 5.357734727859497,
      "epoch": 1.8970821754664549,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046436767620966675,
      "loss": 5.1037,
      "mean_token_accuracy": 0.17052492499351501,
      "num_tokens": 41887345.0,
      "step": 19115
    },
    {
      "entropy": 5.376831340789795,
      "epoch": 1.8975784041286223,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046434853914755744,
      "loss": 5.1641,
      "mean_token_accuracy": 0.16190548837184907,
      "num_tokens": 41898877.0,
      "step": 19120
    },
    {
      "entropy": 5.341642236709594,
      "epoch": 1.89807463279079,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046432939739001237,
      "loss": 5.0793,
      "mean_token_accuracy": 0.17088921815156938,
      "num_tokens": 41910593.0,
      "step": 19125
    },
    {
      "entropy": 5.357475471496582,
      "epoch": 1.8985708614529575,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046431025093750613,
      "loss": 5.1237,
      "mean_token_accuracy": 0.17260655462741853,
      "num_tokens": 41920719.0,
      "step": 19130
    },
    {
      "entropy": 5.43325252532959,
      "epoch": 1.8990670901151252,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046429109979051367,
      "loss": 5.1534,
      "mean_token_accuracy": 0.16108745187520981,
      "num_tokens": 41931630.0,
      "step": 19135
    },
    {
      "entropy": 5.434892463684082,
      "epoch": 1.8995633187772927,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046427194394950985,
      "loss": 5.2107,
      "mean_token_accuracy": 0.16486406773328782,
      "num_tokens": 41943076.0,
      "step": 19140
    },
    {
      "entropy": 5.4527123928070065,
      "epoch": 1.9000595474394602,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004642527834149696,
      "loss": 5.0969,
      "mean_token_accuracy": 0.16936610788106918,
      "num_tokens": 41954294.0,
      "step": 19145
    },
    {
      "entropy": 5.3657495975494385,
      "epoch": 1.9005557761016276,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046423361818736814,
      "loss": 5.0812,
      "mean_token_accuracy": 0.1663456454873085,
      "num_tokens": 41964943.0,
      "step": 19150
    },
    {
      "entropy": 5.312002849578858,
      "epoch": 1.901052004763795,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004642144482671807,
      "loss": 5.1187,
      "mean_token_accuracy": 0.16977479457855224,
      "num_tokens": 41975211.0,
      "step": 19155
    },
    {
      "entropy": 5.419375419616699,
      "epoch": 1.9015482334259626,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046419527365488275,
      "loss": 5.1438,
      "mean_token_accuracy": 0.1657706692814827,
      "num_tokens": 41986492.0,
      "step": 19160
    },
    {
      "entropy": 5.407148790359497,
      "epoch": 1.90204446208813,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004641760943509496,
      "loss": 5.1195,
      "mean_token_accuracy": 0.17234733551740647,
      "num_tokens": 41997613.0,
      "step": 19165
    },
    {
      "entropy": 5.387767028808594,
      "epoch": 1.9025406907502977,
      "grad_norm": 1.0625,
      "learning_rate": 0.000464156910355857,
      "loss": 5.1907,
      "mean_token_accuracy": 0.15525337755680085,
      "num_tokens": 42008597.0,
      "step": 19170
    },
    {
      "entropy": 5.401942253112793,
      "epoch": 1.9030369194124652,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046413772167008057,
      "loss": 5.1618,
      "mean_token_accuracy": 0.1680848106741905,
      "num_tokens": 42019698.0,
      "step": 19175
    },
    {
      "entropy": 5.364930438995361,
      "epoch": 1.903533148074633,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046411852829409626,
      "loss": 5.0764,
      "mean_token_accuracy": 0.16643470376729966,
      "num_tokens": 42029473.0,
      "step": 19180
    },
    {
      "entropy": 5.381076669692993,
      "epoch": 1.9040293767368004,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046409933022837984,
      "loss": 5.1991,
      "mean_token_accuracy": 0.16547415554523467,
      "num_tokens": 42041494.0,
      "step": 19185
    },
    {
      "entropy": 5.451268863677979,
      "epoch": 1.9045256053989679,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004640801274734076,
      "loss": 5.1682,
      "mean_token_accuracy": 0.16671135574579238,
      "num_tokens": 42051798.0,
      "step": 19190
    },
    {
      "entropy": 5.400457382202148,
      "epoch": 1.9050218340611353,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004640609200296557,
      "loss": 5.0997,
      "mean_token_accuracy": 0.16751737296581268,
      "num_tokens": 42061802.0,
      "step": 19195
    },
    {
      "entropy": 5.406962060928345,
      "epoch": 1.9055180627233028,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004640417078976003,
      "loss": 5.2231,
      "mean_token_accuracy": 0.15990807563066484,
      "num_tokens": 42071503.0,
      "step": 19200
    },
    {
      "entropy": 5.283987617492675,
      "epoch": 1.9060142913854703,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004640224910777179,
      "loss": 5.0546,
      "mean_token_accuracy": 0.16925996840000151,
      "num_tokens": 42082104.0,
      "step": 19205
    },
    {
      "entropy": 5.399439811706543,
      "epoch": 1.906510520047638,
      "grad_norm": 0.984375,
      "learning_rate": 0.000464003269570485,
      "loss": 5.1192,
      "mean_token_accuracy": 0.17171578258275985,
      "num_tokens": 42092798.0,
      "step": 19210
    },
    {
      "entropy": 5.422141122817993,
      "epoch": 1.9070067487098055,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004639840433763783,
      "loss": 5.1769,
      "mean_token_accuracy": 0.16412162333726882,
      "num_tokens": 42103062.0,
      "step": 19215
    },
    {
      "entropy": 5.4248960494995115,
      "epoch": 1.9075029773719732,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046396481249587454,
      "loss": 5.2186,
      "mean_token_accuracy": 0.16517686396837233,
      "num_tokens": 42115062.0,
      "step": 19220
    },
    {
      "entropy": 5.45438232421875,
      "epoch": 1.9079992060341406,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004639455769294506,
      "loss": 5.1969,
      "mean_token_accuracy": 0.15957868993282318,
      "num_tokens": 42125937.0,
      "step": 19225
    },
    {
      "entropy": 5.432262134552002,
      "epoch": 1.9084954346963081,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004639263366775835,
      "loss": 5.1399,
      "mean_token_accuracy": 0.16809881180524827,
      "num_tokens": 42136916.0,
      "step": 19230
    },
    {
      "entropy": 5.4256120204925535,
      "epoch": 1.9089916633584756,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046390709174075036,
      "loss": 5.2076,
      "mean_token_accuracy": 0.16400429904460906,
      "num_tokens": 42148348.0,
      "step": 19235
    },
    {
      "entropy": 5.433365440368652,
      "epoch": 1.909487892020643,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004638878421194284,
      "loss": 5.1104,
      "mean_token_accuracy": 0.17426487505435945,
      "num_tokens": 42157586.0,
      "step": 19240
    },
    {
      "entropy": 5.400729990005493,
      "epoch": 1.9099841206828105,
      "grad_norm": 1.0,
      "learning_rate": 0.000463868587814095,
      "loss": 5.1607,
      "mean_token_accuracy": 0.16260536760091782,
      "num_tokens": 42169536.0,
      "step": 19245
    },
    {
      "entropy": 5.340735149383545,
      "epoch": 1.910480349344978,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004638493288252276,
      "loss": 5.0836,
      "mean_token_accuracy": 0.17488220632076262,
      "num_tokens": 42180024.0,
      "step": 19250
    },
    {
      "entropy": 5.358252668380738,
      "epoch": 1.9109765780071457,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046383006515330374,
      "loss": 5.0964,
      "mean_token_accuracy": 0.17272274643182756,
      "num_tokens": 42191070.0,
      "step": 19255
    },
    {
      "entropy": 5.389103937149048,
      "epoch": 1.9114728066693132,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004638107967988012,
      "loss": 5.1829,
      "mean_token_accuracy": 0.1671850323677063,
      "num_tokens": 42203643.0,
      "step": 19260
    },
    {
      "entropy": 5.457135629653931,
      "epoch": 1.9119690353314809,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046379152376219775,
      "loss": 5.2197,
      "mean_token_accuracy": 0.159285057336092,
      "num_tokens": 42214979.0,
      "step": 19265
    },
    {
      "entropy": 5.495288467407226,
      "epoch": 1.9124652639936484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046377224604397127,
      "loss": 5.1804,
      "mean_token_accuracy": 0.16360236704349518,
      "num_tokens": 42225152.0,
      "step": 19270
    },
    {
      "entropy": 5.432648038864135,
      "epoch": 1.9129614926558158,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004637529636445998,
      "loss": 5.0972,
      "mean_token_accuracy": 0.1692135289311409,
      "num_tokens": 42235151.0,
      "step": 19275
    },
    {
      "entropy": 5.356055545806885,
      "epoch": 1.9134577213179833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004637336765645617,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17489419877529144,
      "num_tokens": 42245419.0,
      "step": 19280
    },
    {
      "entropy": 5.343102550506591,
      "epoch": 1.9139539499801508,
      "grad_norm": 1.0,
      "learning_rate": 0.00046371438480433504,
      "loss": 5.0809,
      "mean_token_accuracy": 0.17302192151546478,
      "num_tokens": 42255836.0,
      "step": 19285
    },
    {
      "entropy": 5.346466112136841,
      "epoch": 1.9144501786423183,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004636950883643983,
      "loss": 5.1855,
      "mean_token_accuracy": 0.1661700949072838,
      "num_tokens": 42267490.0,
      "step": 19290
    },
    {
      "entropy": 5.414868974685669,
      "epoch": 1.914946407304486,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046367578724523,
      "loss": 5.1421,
      "mean_token_accuracy": 0.16436727046966554,
      "num_tokens": 42276957.0,
      "step": 19295
    },
    {
      "entropy": 5.412774658203125,
      "epoch": 1.9154426359666534,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004636564814473087,
      "loss": 5.0194,
      "mean_token_accuracy": 0.17648025155067443,
      "num_tokens": 42288469.0,
      "step": 19300
    },
    {
      "entropy": 5.3174210548400875,
      "epoch": 1.9159388646288211,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004636371709711131,
      "loss": 5.0771,
      "mean_token_accuracy": 0.1815100356936455,
      "num_tokens": 42299141.0,
      "step": 19305
    },
    {
      "entropy": 5.392801237106323,
      "epoch": 1.9164350932909886,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046361785581712224,
      "loss": 5.1896,
      "mean_token_accuracy": 0.17165447771549225,
      "num_tokens": 42310651.0,
      "step": 19310
    },
    {
      "entropy": 5.339021444320679,
      "epoch": 1.916931321953156,
      "grad_norm": 1.078125,
      "learning_rate": 0.000463598535985815,
      "loss": 5.0011,
      "mean_token_accuracy": 0.17837045341730118,
      "num_tokens": 42320529.0,
      "step": 19315
    },
    {
      "entropy": 5.342846918106079,
      "epoch": 1.9174275506153236,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004635792114776704,
      "loss": 5.0476,
      "mean_token_accuracy": 0.17140621542930604,
      "num_tokens": 42331216.0,
      "step": 19320
    },
    {
      "entropy": 5.28841199874878,
      "epoch": 1.917923779277491,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004635598822931677,
      "loss": 5.0538,
      "mean_token_accuracy": 0.1758050352334976,
      "num_tokens": 42341730.0,
      "step": 19325
    },
    {
      "entropy": 5.402301168441772,
      "epoch": 1.9184200079396585,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046354054843278624,
      "loss": 5.1779,
      "mean_token_accuracy": 0.16350951045751572,
      "num_tokens": 42352643.0,
      "step": 19330
    },
    {
      "entropy": 5.477546119689942,
      "epoch": 1.918916236601826,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046352120989700553,
      "loss": 5.2154,
      "mean_token_accuracy": 0.1599527895450592,
      "num_tokens": 42363979.0,
      "step": 19335
    },
    {
      "entropy": 5.4657790184021,
      "epoch": 1.9194124652639937,
      "grad_norm": 1.09375,
      "learning_rate": 0.000463501866686305,
      "loss": 5.1673,
      "mean_token_accuracy": 0.16432878077030183,
      "num_tokens": 42374104.0,
      "step": 19340
    },
    {
      "entropy": 5.442071771621704,
      "epoch": 1.9199086939261611,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046348251880116433,
      "loss": 5.1328,
      "mean_token_accuracy": 0.16844428479671478,
      "num_tokens": 42383576.0,
      "step": 19345
    },
    {
      "entropy": 5.36526403427124,
      "epoch": 1.9204049225883288,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046346316624206325,
      "loss": 5.1338,
      "mean_token_accuracy": 0.16680012494325638,
      "num_tokens": 42394625.0,
      "step": 19350
    },
    {
      "entropy": 5.374891376495361,
      "epoch": 1.9209011512504963,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004634438090094819,
      "loss": 5.0894,
      "mean_token_accuracy": 0.1756739467382431,
      "num_tokens": 42404764.0,
      "step": 19355
    },
    {
      "entropy": 5.411190986633301,
      "epoch": 1.9213973799126638,
      "grad_norm": 1.125,
      "learning_rate": 0.00046342444710390006,
      "loss": 5.0699,
      "mean_token_accuracy": 0.17365740984678268,
      "num_tokens": 42414386.0,
      "step": 19360
    },
    {
      "entropy": 5.394908237457275,
      "epoch": 1.9218936085748313,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000463405080525798,
      "loss": 5.1648,
      "mean_token_accuracy": 0.1679188385605812,
      "num_tokens": 42425671.0,
      "step": 19365
    },
    {
      "entropy": 5.412374973297119,
      "epoch": 1.9223898372369987,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004633857092756559,
      "loss": 5.1977,
      "mean_token_accuracy": 0.16295471638441086,
      "num_tokens": 42436365.0,
      "step": 19370
    },
    {
      "entropy": 5.409897708892823,
      "epoch": 1.9228860658991662,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004633663333539542,
      "loss": 5.0458,
      "mean_token_accuracy": 0.16757295727729798,
      "num_tokens": 42447624.0,
      "step": 19375
    },
    {
      "entropy": 5.411576938629151,
      "epoch": 1.923382294561334,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046334695276117324,
      "loss": 5.1425,
      "mean_token_accuracy": 0.16692752838134767,
      "num_tokens": 42458013.0,
      "step": 19380
    },
    {
      "entropy": 5.3967078685760494,
      "epoch": 1.9238785232235014,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004633275674977937,
      "loss": 5.1683,
      "mean_token_accuracy": 0.16139222234487532,
      "num_tokens": 42469807.0,
      "step": 19385
    },
    {
      "entropy": 5.3427817821502686,
      "epoch": 1.9243747518856689,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004633081775642964,
      "loss": 5.071,
      "mean_token_accuracy": 0.16809558421373366,
      "num_tokens": 42479240.0,
      "step": 19390
    },
    {
      "entropy": 5.375578689575195,
      "epoch": 1.9248709805478366,
      "grad_norm": 1.0,
      "learning_rate": 0.00046328878296116196,
      "loss": 5.1621,
      "mean_token_accuracy": 0.170993635058403,
      "num_tokens": 42489558.0,
      "step": 19395
    },
    {
      "entropy": 5.396966123580933,
      "epoch": 1.925367209210004,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046326938368887146,
      "loss": 5.1753,
      "mean_token_accuracy": 0.17026007026433945,
      "num_tokens": 42501252.0,
      "step": 19400
    },
    {
      "entropy": 5.457243299484253,
      "epoch": 1.9258634378721715,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046324997974790596,
      "loss": 5.2175,
      "mean_token_accuracy": 0.16711069792509078,
      "num_tokens": 42512330.0,
      "step": 19405
    },
    {
      "entropy": 5.472326564788818,
      "epoch": 1.926359666534339,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004632305711387466,
      "loss": 5.1947,
      "mean_token_accuracy": 0.16314466744661332,
      "num_tokens": 42525216.0,
      "step": 19410
    },
    {
      "entropy": 5.440891075134277,
      "epoch": 1.9268558951965065,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046321115786187466,
      "loss": 5.1277,
      "mean_token_accuracy": 0.16714057177305222,
      "num_tokens": 42537047.0,
      "step": 19415
    },
    {
      "entropy": 5.37639102935791,
      "epoch": 1.927352123858674,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046319173991777153,
      "loss": 5.1291,
      "mean_token_accuracy": 0.1663987949490547,
      "num_tokens": 42549784.0,
      "step": 19420
    },
    {
      "entropy": 5.402208137512207,
      "epoch": 1.9278483525208416,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004631723173069188,
      "loss": 5.1238,
      "mean_token_accuracy": 0.1665441706776619,
      "num_tokens": 42561079.0,
      "step": 19425
    },
    {
      "entropy": 5.392603588104248,
      "epoch": 1.928344581183009,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004631528900297981,
      "loss": 5.1753,
      "mean_token_accuracy": 0.16101975739002228,
      "num_tokens": 42572531.0,
      "step": 19430
    },
    {
      "entropy": 5.373914432525635,
      "epoch": 1.9288408098451768,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046313345808689113,
      "loss": 5.0898,
      "mean_token_accuracy": 0.16929156333208084,
      "num_tokens": 42583836.0,
      "step": 19435
    },
    {
      "entropy": 5.4412675380706785,
      "epoch": 1.9293370385073443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004631140214786798,
      "loss": 5.2259,
      "mean_token_accuracy": 0.16125105023384095,
      "num_tokens": 42595338.0,
      "step": 19440
    },
    {
      "entropy": 5.42824420928955,
      "epoch": 1.9298332671695118,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046309458020564606,
      "loss": 5.1523,
      "mean_token_accuracy": 0.16390303373336793,
      "num_tokens": 42606804.0,
      "step": 19445
    },
    {
      "entropy": 5.396226024627685,
      "epoch": 1.9303294958316792,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046307513426827214,
      "loss": 5.077,
      "mean_token_accuracy": 0.17317114472389222,
      "num_tokens": 42616987.0,
      "step": 19450
    },
    {
      "entropy": 5.378240537643433,
      "epoch": 1.9308257244938467,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046305568366704006,
      "loss": 5.1414,
      "mean_token_accuracy": 0.1670861840248108,
      "num_tokens": 42627633.0,
      "step": 19455
    },
    {
      "entropy": 5.4412554740905765,
      "epoch": 1.9313219531560142,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004630362284024322,
      "loss": 5.1147,
      "mean_token_accuracy": 0.16800030022859574,
      "num_tokens": 42638802.0,
      "step": 19460
    },
    {
      "entropy": 5.401471185684204,
      "epoch": 1.9318181818181817,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046301676847493116,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17713085561990738,
      "num_tokens": 42650802.0,
      "step": 19465
    },
    {
      "entropy": 5.323422622680664,
      "epoch": 1.9323144104803494,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046299730388501935,
      "loss": 5.0591,
      "mean_token_accuracy": 0.17239415645599365,
      "num_tokens": 42660290.0,
      "step": 19470
    },
    {
      "entropy": 5.376088619232178,
      "epoch": 1.9328106391425168,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046297783463317957,
      "loss": 5.1676,
      "mean_token_accuracy": 0.1705165311694145,
      "num_tokens": 42671182.0,
      "step": 19475
    },
    {
      "entropy": 5.449751186370849,
      "epoch": 1.9333068678046845,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004629583607198945,
      "loss": 5.1533,
      "mean_token_accuracy": 0.1716858461499214,
      "num_tokens": 42682499.0,
      "step": 19480
    },
    {
      "entropy": 5.374654245376587,
      "epoch": 1.933803096466852,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046293888214564704,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1666281044483185,
      "num_tokens": 42693728.0,
      "step": 19485
    },
    {
      "entropy": 5.422718095779419,
      "epoch": 1.9342993251290195,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046291939891092036,
      "loss": 5.163,
      "mean_token_accuracy": 0.16457717567682267,
      "num_tokens": 42704877.0,
      "step": 19490
    },
    {
      "entropy": 5.461507511138916,
      "epoch": 1.934795553791187,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004628999110161975,
      "loss": 5.1819,
      "mean_token_accuracy": 0.15986616611480714,
      "num_tokens": 42715787.0,
      "step": 19495
    },
    {
      "entropy": 5.426688098907471,
      "epoch": 1.9352917824533544,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046288041846196174,
      "loss": 5.1294,
      "mean_token_accuracy": 0.16849457174539567,
      "num_tokens": 42727557.0,
      "step": 19500
    },
    {
      "entropy": 5.4417274475097654,
      "epoch": 1.935788011115522,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046286092124869647,
      "loss": 5.1415,
      "mean_token_accuracy": 0.1655034676194191,
      "num_tokens": 42737540.0,
      "step": 19505
    },
    {
      "entropy": 5.471788740158081,
      "epoch": 1.9362842397776896,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046284141937688513,
      "loss": 5.1894,
      "mean_token_accuracy": 0.170176400244236,
      "num_tokens": 42748716.0,
      "step": 19510
    },
    {
      "entropy": 5.49242639541626,
      "epoch": 1.936780468439857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046282191284701123,
      "loss": 5.2124,
      "mean_token_accuracy": 0.16174225956201554,
      "num_tokens": 42758991.0,
      "step": 19515
    },
    {
      "entropy": 5.502219247817993,
      "epoch": 1.9372766971020248,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004628024016595587,
      "loss": 5.196,
      "mean_token_accuracy": 0.16290506720542908,
      "num_tokens": 42771105.0,
      "step": 19520
    },
    {
      "entropy": 5.4173332214355465,
      "epoch": 1.9377729257641922,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046278288581501136,
      "loss": 5.1425,
      "mean_token_accuracy": 0.16861090809106827,
      "num_tokens": 42783317.0,
      "step": 19525
    },
    {
      "entropy": 5.449121189117432,
      "epoch": 1.9382691544263597,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046276336531385305,
      "loss": 5.2786,
      "mean_token_accuracy": 0.16247311532497405,
      "num_tokens": 42794998.0,
      "step": 19530
    },
    {
      "entropy": 5.380791234970093,
      "epoch": 1.9387653830885272,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046274384015656785,
      "loss": 5.1365,
      "mean_token_accuracy": 0.17198197394609452,
      "num_tokens": 42806197.0,
      "step": 19535
    },
    {
      "entropy": 5.33711051940918,
      "epoch": 1.9392616117506947,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046272431034363997,
      "loss": 5.0406,
      "mean_token_accuracy": 0.1665802627801895,
      "num_tokens": 42816522.0,
      "step": 19540
    },
    {
      "entropy": 5.319559907913208,
      "epoch": 1.9397578404128621,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004627047758755538,
      "loss": 5.0547,
      "mean_token_accuracy": 0.17180028557777405,
      "num_tokens": 42826003.0,
      "step": 19545
    },
    {
      "entropy": 5.407004880905151,
      "epoch": 1.9402540690750296,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004626852367527935,
      "loss": 5.1501,
      "mean_token_accuracy": 0.16102501302957534,
      "num_tokens": 42836346.0,
      "step": 19550
    },
    {
      "entropy": 5.393400382995606,
      "epoch": 1.9407502977371973,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004626656929758438,
      "loss": 5.185,
      "mean_token_accuracy": 0.16816026866436004,
      "num_tokens": 42846296.0,
      "step": 19555
    },
    {
      "entropy": 5.444653797149658,
      "epoch": 1.9412465263993648,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004626461445451894,
      "loss": 5.2118,
      "mean_token_accuracy": 0.1660264953970909,
      "num_tokens": 42856323.0,
      "step": 19560
    },
    {
      "entropy": 5.484603071212769,
      "epoch": 1.9417427550615325,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004626265914613149,
      "loss": 5.1981,
      "mean_token_accuracy": 0.16627684384584426,
      "num_tokens": 42866937.0,
      "step": 19565
    },
    {
      "entropy": 5.449637126922608,
      "epoch": 1.9422389837237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004626070337247052,
      "loss": 5.1671,
      "mean_token_accuracy": 0.15977009981870652,
      "num_tokens": 42877194.0,
      "step": 19570
    },
    {
      "entropy": 5.408547830581665,
      "epoch": 1.9427352123858674,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004625874713358453,
      "loss": 5.1804,
      "mean_token_accuracy": 0.1644926518201828,
      "num_tokens": 42887741.0,
      "step": 19575
    },
    {
      "entropy": 5.39282193183899,
      "epoch": 1.943231441048035,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046256790429522035,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1772836208343506,
      "num_tokens": 42898195.0,
      "step": 19580
    },
    {
      "entropy": 5.466822767257691,
      "epoch": 1.9437276697102024,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004625483326033155,
      "loss": 5.1581,
      "mean_token_accuracy": 0.16825207322835922,
      "num_tokens": 42908852.0,
      "step": 19585
    },
    {
      "entropy": 5.360301733016968,
      "epoch": 1.9442238983723699,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004625287562606162,
      "loss": 5.0923,
      "mean_token_accuracy": 0.17008597850799562,
      "num_tokens": 42920166.0,
      "step": 19590
    },
    {
      "entropy": 5.426490736007691,
      "epoch": 1.9447201270345376,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004625091752676078,
      "loss": 5.1636,
      "mean_token_accuracy": 0.16595747619867324,
      "num_tokens": 42931677.0,
      "step": 19595
    },
    {
      "entropy": 5.403318166732788,
      "epoch": 1.945216355696705,
      "grad_norm": 1.03125,
      "learning_rate": 0.000462489589624776,
      "loss": 5.1537,
      "mean_token_accuracy": 0.15975116789340973,
      "num_tokens": 42942346.0,
      "step": 19600
    },
    {
      "entropy": 5.328712272644043,
      "epoch": 1.9457125843588727,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004624699993326063,
      "loss": 5.0987,
      "mean_token_accuracy": 0.17088836431503296,
      "num_tokens": 42953821.0,
      "step": 19605
    },
    {
      "entropy": 5.394847393035889,
      "epoch": 1.9462088130210402,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046245040439158456,
      "loss": 5.1067,
      "mean_token_accuracy": 0.16864792257547379,
      "num_tokens": 42964358.0,
      "step": 19610
    },
    {
      "entropy": 5.406155395507812,
      "epoch": 1.9467050416832077,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046243080480219673,
      "loss": 5.1038,
      "mean_token_accuracy": 0.16838450282812117,
      "num_tokens": 42976459.0,
      "step": 19615
    },
    {
      "entropy": 5.462235736846924,
      "epoch": 1.9472012703453752,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004624112005649288,
      "loss": 5.2465,
      "mean_token_accuracy": 0.16167334616184234,
      "num_tokens": 42988603.0,
      "step": 19620
    },
    {
      "entropy": 5.371326494216919,
      "epoch": 1.9476974990075426,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046239159168026694,
      "loss": 5.1558,
      "mean_token_accuracy": 0.16654073148965837,
      "num_tokens": 43000262.0,
      "step": 19625
    },
    {
      "entropy": 5.328092432022094,
      "epoch": 1.94819372766971,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004623719781486974,
      "loss": 5.1038,
      "mean_token_accuracy": 0.17301279902458191,
      "num_tokens": 43009363.0,
      "step": 19630
    },
    {
      "entropy": 5.432528257369995,
      "epoch": 1.9486899563318776,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046235235997070664,
      "loss": 5.172,
      "mean_token_accuracy": 0.16934423595666886,
      "num_tokens": 43019893.0,
      "step": 19635
    },
    {
      "entropy": 5.333480787277222,
      "epoch": 1.9491861849940453,
      "grad_norm": 1.03125,
      "learning_rate": 0.000462332737146781,
      "loss": 5.0729,
      "mean_token_accuracy": 0.17247299253940582,
      "num_tokens": 43031152.0,
      "step": 19640
    },
    {
      "entropy": 5.397505426406861,
      "epoch": 1.9496824136562128,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004623131096774071,
      "loss": 5.2165,
      "mean_token_accuracy": 0.16457413285970687,
      "num_tokens": 43041247.0,
      "step": 19645
    },
    {
      "entropy": 5.416206932067871,
      "epoch": 1.9501786423183805,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004622934775630717,
      "loss": 5.134,
      "mean_token_accuracy": 0.1624265879392624,
      "num_tokens": 43052065.0,
      "step": 19650
    },
    {
      "entropy": 5.38266544342041,
      "epoch": 1.950674870980548,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046227384080426163,
      "loss": 5.1637,
      "mean_token_accuracy": 0.16128694340586663,
      "num_tokens": 43063217.0,
      "step": 19655
    },
    {
      "entropy": 5.40373044013977,
      "epoch": 1.9511710996427154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004622541994014639,
      "loss": 5.167,
      "mean_token_accuracy": 0.17088363766670228,
      "num_tokens": 43074463.0,
      "step": 19660
    },
    {
      "entropy": 5.5168437480926515,
      "epoch": 1.9516673283048829,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004622345533551655,
      "loss": 5.1957,
      "mean_token_accuracy": 0.16045141071081162,
      "num_tokens": 43086004.0,
      "step": 19665
    },
    {
      "entropy": 5.423062419891357,
      "epoch": 1.9521635569670504,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004622149026658537,
      "loss": 5.1181,
      "mean_token_accuracy": 0.17026802897453308,
      "num_tokens": 43096973.0,
      "step": 19670
    },
    {
      "entropy": 5.4513226509094235,
      "epoch": 1.9526597856292178,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004621952473340156,
      "loss": 5.2103,
      "mean_token_accuracy": 0.1655504286289215,
      "num_tokens": 43107337.0,
      "step": 19675
    },
    {
      "entropy": 5.414171028137207,
      "epoch": 1.9531560142913855,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004621755873601388,
      "loss": 5.1074,
      "mean_token_accuracy": 0.1751821294426918,
      "num_tokens": 43118668.0,
      "step": 19680
    },
    {
      "entropy": 5.437110137939453,
      "epoch": 1.953652242953553,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004621559227447108,
      "loss": 5.2044,
      "mean_token_accuracy": 0.1641070544719696,
      "num_tokens": 43129030.0,
      "step": 19685
    },
    {
      "entropy": 5.311437749862671,
      "epoch": 1.9541484716157205,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004621362534882192,
      "loss": 5.0793,
      "mean_token_accuracy": 0.17224566638469696,
      "num_tokens": 43139120.0,
      "step": 19690
    },
    {
      "entropy": 5.396817207336426,
      "epoch": 1.9546447002778882,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004621165795911517,
      "loss": 5.177,
      "mean_token_accuracy": 0.1647260919213295,
      "num_tokens": 43152390.0,
      "step": 19695
    },
    {
      "entropy": 5.376423788070679,
      "epoch": 1.9551409289400556,
      "grad_norm": 1.125,
      "learning_rate": 0.00046209690105399623,
      "loss": 4.9836,
      "mean_token_accuracy": 0.17670640349388123,
      "num_tokens": 43162230.0,
      "step": 19700
    },
    {
      "entropy": 5.303509092330932,
      "epoch": 1.9556371576022231,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046207721787724085,
      "loss": 5.1057,
      "mean_token_accuracy": 0.17058635801076888,
      "num_tokens": 43172751.0,
      "step": 19705
    },
    {
      "entropy": 5.453035879135132,
      "epoch": 1.9561333862643906,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046205753006137347,
      "loss": 5.2047,
      "mean_token_accuracy": 0.17228127270936966,
      "num_tokens": 43183835.0,
      "step": 19710
    },
    {
      "entropy": 5.471931791305542,
      "epoch": 1.956629614926558,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046203783760688245,
      "loss": 5.0754,
      "mean_token_accuracy": 0.1673237904906273,
      "num_tokens": 43193178.0,
      "step": 19715
    },
    {
      "entropy": 5.38984899520874,
      "epoch": 1.9571258435887255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004620181405142561,
      "loss": 5.241,
      "mean_token_accuracy": 0.1613100916147232,
      "num_tokens": 43204968.0,
      "step": 19720
    },
    {
      "entropy": 5.357039165496826,
      "epoch": 1.9576220722508932,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004619984387839829,
      "loss": 5.256,
      "mean_token_accuracy": 0.17014575749635696,
      "num_tokens": 43214916.0,
      "step": 19725
    },
    {
      "entropy": 5.503975486755371,
      "epoch": 1.9581183009130607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046197873241655133,
      "loss": 5.2417,
      "mean_token_accuracy": 0.1624678537249565,
      "num_tokens": 43226374.0,
      "step": 19730
    },
    {
      "entropy": 5.39903073310852,
      "epoch": 1.9586145295752284,
      "grad_norm": 1.0,
      "learning_rate": 0.0004619590214124501,
      "loss": 5.1193,
      "mean_token_accuracy": 0.16355736851692199,
      "num_tokens": 43237296.0,
      "step": 19735
    },
    {
      "entropy": 5.5230895519256595,
      "epoch": 1.9591107582373959,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046193930577216797,
      "loss": 5.2923,
      "mean_token_accuracy": 0.15185001194477082,
      "num_tokens": 43248220.0,
      "step": 19740
    },
    {
      "entropy": 5.385967254638672,
      "epoch": 1.9596069868995634,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004619195854961939,
      "loss": 5.0801,
      "mean_token_accuracy": 0.17384767681360244,
      "num_tokens": 43259913.0,
      "step": 19745
    },
    {
      "entropy": 5.413989782333374,
      "epoch": 1.9601032155617308,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046189986058501687,
      "loss": 5.2036,
      "mean_token_accuracy": 0.15791984647512436,
      "num_tokens": 43272317.0,
      "step": 19750
    },
    {
      "entropy": 5.310960960388184,
      "epoch": 1.9605994442238983,
      "grad_norm": 1.09375,
      "learning_rate": 0.000461880131039126,
      "loss": 5.0566,
      "mean_token_accuracy": 0.17387327700853347,
      "num_tokens": 43283170.0,
      "step": 19755
    },
    {
      "entropy": 5.363578987121582,
      "epoch": 1.9610956728860658,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004618603968590107,
      "loss": 5.1373,
      "mean_token_accuracy": 0.17014632150530815,
      "num_tokens": 43293899.0,
      "step": 19760
    },
    {
      "entropy": 5.449765205383301,
      "epoch": 1.9615919015482333,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046184065804516007,
      "loss": 5.1962,
      "mean_token_accuracy": 0.165854811668396,
      "num_tokens": 43305638.0,
      "step": 19765
    },
    {
      "entropy": 5.451696729660034,
      "epoch": 1.962088130210401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004618209145980638,
      "loss": 5.2518,
      "mean_token_accuracy": 0.16418867111206054,
      "num_tokens": 43315900.0,
      "step": 19770
    },
    {
      "entropy": 5.523223638534546,
      "epoch": 1.9625843588725684,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004618011665182114,
      "loss": 5.3105,
      "mean_token_accuracy": 0.16042123287916182,
      "num_tokens": 43328464.0,
      "step": 19775
    },
    {
      "entropy": 5.347068929672242,
      "epoch": 1.9630805875347361,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004617814138060926,
      "loss": 5.0003,
      "mean_token_accuracy": 0.17659809291362763,
      "num_tokens": 43339303.0,
      "step": 19780
    },
    {
      "entropy": 5.432402515411377,
      "epoch": 1.9635768161969036,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004617616564621972,
      "loss": 5.2196,
      "mean_token_accuracy": 0.16343667954206467,
      "num_tokens": 43349269.0,
      "step": 19785
    },
    {
      "entropy": 5.406648969650268,
      "epoch": 1.964073044859071,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046174189448701516,
      "loss": 5.0834,
      "mean_token_accuracy": 0.17063161134719848,
      "num_tokens": 43361090.0,
      "step": 19790
    },
    {
      "entropy": 5.34601526260376,
      "epoch": 1.9645692735212386,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004617221278810365,
      "loss": 5.0752,
      "mean_token_accuracy": 0.16701819449663163,
      "num_tokens": 43373448.0,
      "step": 19795
    },
    {
      "entropy": 5.44362998008728,
      "epoch": 1.965065502183406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046170235664475147,
      "loss": 5.2285,
      "mean_token_accuracy": 0.16370176672935485,
      "num_tokens": 43386202.0,
      "step": 19800
    },
    {
      "entropy": 5.401737928390503,
      "epoch": 1.9655617308455735,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046168258077865026,
      "loss": 5.1341,
      "mean_token_accuracy": 0.17246710658073425,
      "num_tokens": 43396363.0,
      "step": 19805
    },
    {
      "entropy": 5.3829151630401615,
      "epoch": 1.9660579595077412,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004616628002832233,
      "loss": 5.0266,
      "mean_token_accuracy": 0.17456901967525482,
      "num_tokens": 43406076.0,
      "step": 19810
    },
    {
      "entropy": 5.40518102645874,
      "epoch": 1.9665541881699087,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046164301515896124,
      "loss": 5.1958,
      "mean_token_accuracy": 0.16759460270404816,
      "num_tokens": 43416226.0,
      "step": 19815
    },
    {
      "entropy": 5.344092464447021,
      "epoch": 1.9670504168320764,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004616232254063544,
      "loss": 5.0238,
      "mean_token_accuracy": 0.1706991448998451,
      "num_tokens": 43427364.0,
      "step": 19820
    },
    {
      "entropy": 5.369702100753784,
      "epoch": 1.9675466454942439,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046160343102589384,
      "loss": 5.1567,
      "mean_token_accuracy": 0.17060515880584717,
      "num_tokens": 43437854.0,
      "step": 19825
    },
    {
      "entropy": 5.433363437652588,
      "epoch": 1.9680428741564113,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004615836320180702,
      "loss": 5.1887,
      "mean_token_accuracy": 0.1696033075451851,
      "num_tokens": 43449370.0,
      "step": 19830
    },
    {
      "entropy": 5.365792274475098,
      "epoch": 1.9685391028185788,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046156382838337455,
      "loss": 5.137,
      "mean_token_accuracy": 0.17582536935806276,
      "num_tokens": 43461631.0,
      "step": 19835
    },
    {
      "entropy": 5.437546634674073,
      "epoch": 1.9690353314807463,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046154402012229807,
      "loss": 5.1215,
      "mean_token_accuracy": 0.165701462328434,
      "num_tokens": 43472357.0,
      "step": 19840
    },
    {
      "entropy": 5.363189363479615,
      "epoch": 1.9695315601429138,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046152420723533175,
      "loss": 5.0803,
      "mean_token_accuracy": 0.17316187024116517,
      "num_tokens": 43483642.0,
      "step": 19845
    },
    {
      "entropy": 5.430753326416015,
      "epoch": 1.9700277888050812,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046150438972296706,
      "loss": 5.1771,
      "mean_token_accuracy": 0.16682811677455903,
      "num_tokens": 43495136.0,
      "step": 19850
    },
    {
      "entropy": 5.358109712600708,
      "epoch": 1.970524017467249,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046148456758569536,
      "loss": 5.1418,
      "mean_token_accuracy": 0.16962178200483322,
      "num_tokens": 43506376.0,
      "step": 19855
    },
    {
      "entropy": 5.394447994232178,
      "epoch": 1.9710202461294164,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046146474082400823,
      "loss": 5.1232,
      "mean_token_accuracy": 0.17019163817167282,
      "num_tokens": 43516658.0,
      "step": 19860
    },
    {
      "entropy": 5.386396837234497,
      "epoch": 1.971516474791584,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046144490943839735,
      "loss": 5.0741,
      "mean_token_accuracy": 0.1706658586859703,
      "num_tokens": 43526907.0,
      "step": 19865
    },
    {
      "entropy": 5.448320388793945,
      "epoch": 1.9720127034537516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004614250734293544,
      "loss": 5.1784,
      "mean_token_accuracy": 0.16702838987112045,
      "num_tokens": 43537739.0,
      "step": 19870
    },
    {
      "entropy": 5.339805364608765,
      "epoch": 1.972508932115919,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046140523279737146,
      "loss": 5.1401,
      "mean_token_accuracy": 0.17053551375865936,
      "num_tokens": 43549240.0,
      "step": 19875
    },
    {
      "entropy": 5.395406818389892,
      "epoch": 1.9730051607780865,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004613853875429403,
      "loss": 5.0954,
      "mean_token_accuracy": 0.16741308718919753,
      "num_tokens": 43559857.0,
      "step": 19880
    },
    {
      "entropy": 5.347815418243409,
      "epoch": 1.973501389440254,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004613655376665532,
      "loss": 5.0766,
      "mean_token_accuracy": 0.16969573497772217,
      "num_tokens": 43570262.0,
      "step": 19885
    },
    {
      "entropy": 5.43888487815857,
      "epoch": 1.9739976181024215,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004613456831687024,
      "loss": 5.1568,
      "mean_token_accuracy": 0.16406352669000626,
      "num_tokens": 43581965.0,
      "step": 19890
    },
    {
      "entropy": 5.429673337936402,
      "epoch": 1.9744938467645892,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046132582404988004,
      "loss": 5.1197,
      "mean_token_accuracy": 0.168502838909626,
      "num_tokens": 43592728.0,
      "step": 19895
    },
    {
      "entropy": 5.461760807037353,
      "epoch": 1.9749900754267566,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046130596031057885,
      "loss": 5.0812,
      "mean_token_accuracy": 0.16552177369594573,
      "num_tokens": 43602437.0,
      "step": 19900
    },
    {
      "entropy": 5.392862939834595,
      "epoch": 1.9754863040889243,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046128609195129135,
      "loss": 5.2097,
      "mean_token_accuracy": 0.16556576639413834,
      "num_tokens": 43613115.0,
      "step": 19905
    },
    {
      "entropy": 5.439435338973999,
      "epoch": 1.9759825327510918,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004612662189725101,
      "loss": 5.2001,
      "mean_token_accuracy": 0.16500074714422225,
      "num_tokens": 43623700.0,
      "step": 19910
    },
    {
      "entropy": 5.480802202224732,
      "epoch": 1.9764787614132593,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000461246341374728,
      "loss": 5.1947,
      "mean_token_accuracy": 0.1657433457672596,
      "num_tokens": 43633712.0,
      "step": 19915
    },
    {
      "entropy": 5.428330802917481,
      "epoch": 1.9769749900754268,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046122645915843796,
      "loss": 5.2384,
      "mean_token_accuracy": 0.16406229734420777,
      "num_tokens": 43643705.0,
      "step": 19920
    },
    {
      "entropy": 5.438689374923706,
      "epoch": 1.9774712187375942,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046120657232413305,
      "loss": 5.1867,
      "mean_token_accuracy": 0.17139729857444763,
      "num_tokens": 43655376.0,
      "step": 19925
    },
    {
      "entropy": 5.417524528503418,
      "epoch": 1.9779674473997617,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004611866808723064,
      "loss": 5.0809,
      "mean_token_accuracy": 0.17353419065475464,
      "num_tokens": 43666329.0,
      "step": 19930
    },
    {
      "entropy": 5.360719633102417,
      "epoch": 1.9784636760619292,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046116678480345117,
      "loss": 5.102,
      "mean_token_accuracy": 0.1671719178557396,
      "num_tokens": 43676415.0,
      "step": 19935
    },
    {
      "entropy": 5.341747665405274,
      "epoch": 1.9789599047240969,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046114688411806083,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17169731855392456,
      "num_tokens": 43686997.0,
      "step": 19940
    },
    {
      "entropy": 5.364024686813354,
      "epoch": 1.9794561333862644,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046112697881662895,
      "loss": 5.0796,
      "mean_token_accuracy": 0.17265159785747528,
      "num_tokens": 43697342.0,
      "step": 19945
    },
    {
      "entropy": 5.4830575466156,
      "epoch": 1.979952362048432,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046110706889964905,
      "loss": 5.1841,
      "mean_token_accuracy": 0.16631000787019729,
      "num_tokens": 43707876.0,
      "step": 19950
    },
    {
      "entropy": 5.391189193725586,
      "epoch": 1.9804485907105995,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004610871543676149,
      "loss": 5.0554,
      "mean_token_accuracy": 0.17583166807889938,
      "num_tokens": 43718268.0,
      "step": 19955
    },
    {
      "entropy": 5.3913685321807865,
      "epoch": 1.980944819372767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046106723522102024,
      "loss": 5.2043,
      "mean_token_accuracy": 0.15802356004714965,
      "num_tokens": 43729853.0,
      "step": 19960
    },
    {
      "entropy": 5.434724950790406,
      "epoch": 1.9814410480349345,
      "grad_norm": 1.125,
      "learning_rate": 0.00046104731146035904,
      "loss": 5.1646,
      "mean_token_accuracy": 0.16843919456005096,
      "num_tokens": 43739886.0,
      "step": 19965
    },
    {
      "entropy": 5.446300220489502,
      "epoch": 1.981937276697102,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004610273830861255,
      "loss": 5.1586,
      "mean_token_accuracy": 0.16468456089496614,
      "num_tokens": 43750536.0,
      "step": 19970
    },
    {
      "entropy": 5.438199996948242,
      "epoch": 1.9824335053592694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004610074500988137,
      "loss": 5.1369,
      "mean_token_accuracy": 0.1684786856174469,
      "num_tokens": 43763192.0,
      "step": 19975
    },
    {
      "entropy": 5.448578929901123,
      "epoch": 1.9829297340214371,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004609875124989179,
      "loss": 5.1733,
      "mean_token_accuracy": 0.16483449041843415,
      "num_tokens": 43774117.0,
      "step": 19980
    },
    {
      "entropy": 5.3603393077850345,
      "epoch": 1.9834259626836046,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004609675702869326,
      "loss": 5.0993,
      "mean_token_accuracy": 0.16941318958997725,
      "num_tokens": 43784708.0,
      "step": 19985
    },
    {
      "entropy": 5.425611257553101,
      "epoch": 1.983922191345772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004609476234633523,
      "loss": 5.1848,
      "mean_token_accuracy": 0.1687060996890068,
      "num_tokens": 43796767.0,
      "step": 19990
    },
    {
      "entropy": 5.455774450302124,
      "epoch": 1.9844184200079398,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046092767202867156,
      "loss": 5.2643,
      "mean_token_accuracy": 0.15858353823423385,
      "num_tokens": 43807203.0,
      "step": 19995
    },
    {
      "entropy": 5.395251178741455,
      "epoch": 1.9849146486701073,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046090771598338525,
      "loss": 5.1631,
      "mean_token_accuracy": 0.16980056017637252,
      "num_tokens": 43817123.0,
      "step": 20000
    },
    {
      "entropy": 5.4092512130737305,
      "epoch": 1.9854108773322747,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004608877553279881,
      "loss": 5.124,
      "mean_token_accuracy": 0.17157810032367707,
      "num_tokens": 43828625.0,
      "step": 20005
    },
    {
      "entropy": 5.384819841384887,
      "epoch": 1.9859071059944422,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004608677900629753,
      "loss": 5.0934,
      "mean_token_accuracy": 0.172351698577404,
      "num_tokens": 43840012.0,
      "step": 20010
    },
    {
      "entropy": 5.3264271259307865,
      "epoch": 1.9864033346566097,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046084782018884177,
      "loss": 5.0907,
      "mean_token_accuracy": 0.17100428640842438,
      "num_tokens": 43851253.0,
      "step": 20015
    },
    {
      "entropy": 5.477815437316894,
      "epoch": 1.9868995633187772,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046082784570608266,
      "loss": 5.1984,
      "mean_token_accuracy": 0.16761325001716615,
      "num_tokens": 43864389.0,
      "step": 20020
    },
    {
      "entropy": 5.445384836196899,
      "epoch": 1.9873957919809448,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046080786661519354,
      "loss": 5.2036,
      "mean_token_accuracy": 0.16528184562921525,
      "num_tokens": 43876446.0,
      "step": 20025
    },
    {
      "entropy": 5.383210229873657,
      "epoch": 1.9878920206431123,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004607878829166696,
      "loss": 5.1426,
      "mean_token_accuracy": 0.16674996465444564,
      "num_tokens": 43888318.0,
      "step": 20030
    },
    {
      "entropy": 5.360523366928101,
      "epoch": 1.98838824930528,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046076789461100653,
      "loss": 5.0819,
      "mean_token_accuracy": 0.177898570895195,
      "num_tokens": 43898358.0,
      "step": 20035
    },
    {
      "entropy": 5.438930654525757,
      "epoch": 1.9888844779674475,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004607479016987001,
      "loss": 5.1501,
      "mean_token_accuracy": 0.16930540651082993,
      "num_tokens": 43909831.0,
      "step": 20040
    },
    {
      "entropy": 5.439978742599488,
      "epoch": 1.989380706629615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004607279041802458,
      "loss": 5.1005,
      "mean_token_accuracy": 0.16887544393539428,
      "num_tokens": 43919648.0,
      "step": 20045
    },
    {
      "entropy": 5.367334175109863,
      "epoch": 1.9898769352917824,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046070790205613967,
      "loss": 5.1806,
      "mean_token_accuracy": 0.16503848135471344,
      "num_tokens": 43930563.0,
      "step": 20050
    },
    {
      "entropy": 5.447083759307861,
      "epoch": 1.99037316395395,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046068789532687784,
      "loss": 5.2359,
      "mean_token_accuracy": 0.15888031274080278,
      "num_tokens": 43942490.0,
      "step": 20055
    },
    {
      "entropy": 5.3544398784637455,
      "epoch": 1.9908693926161174,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004606678839929563,
      "loss": 4.9918,
      "mean_token_accuracy": 0.18316900730133057,
      "num_tokens": 43952187.0,
      "step": 20060
    },
    {
      "entropy": 5.31886134147644,
      "epoch": 1.9913656212782849,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004606478680548713,
      "loss": 5.0797,
      "mean_token_accuracy": 0.17065695077180862,
      "num_tokens": 43964203.0,
      "step": 20065
    },
    {
      "entropy": 5.369805860519409,
      "epoch": 1.9918618499404526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046062784751311917,
      "loss": 5.1371,
      "mean_token_accuracy": 0.16739059537649154,
      "num_tokens": 43974048.0,
      "step": 20070
    },
    {
      "entropy": 5.437464237213135,
      "epoch": 1.99235807860262,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046060782236819647,
      "loss": 5.154,
      "mean_token_accuracy": 0.1665111869573593,
      "num_tokens": 43984469.0,
      "step": 20075
    },
    {
      "entropy": 5.396588659286499,
      "epoch": 1.9928543072647877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004605877926205996,
      "loss": 5.1637,
      "mean_token_accuracy": 0.17018131166696548,
      "num_tokens": 43995494.0,
      "step": 20080
    },
    {
      "entropy": 5.414907741546631,
      "epoch": 1.9933505359269552,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004605677582708256,
      "loss": 5.075,
      "mean_token_accuracy": 0.1694796234369278,
      "num_tokens": 44005869.0,
      "step": 20085
    },
    {
      "entropy": 5.442898941040039,
      "epoch": 1.9938467645891227,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004605477193193709,
      "loss": 5.1473,
      "mean_token_accuracy": 0.1660605326294899,
      "num_tokens": 44016956.0,
      "step": 20090
    },
    {
      "entropy": 5.422897148132324,
      "epoch": 1.9943429932512902,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004605276757667326,
      "loss": 5.2224,
      "mean_token_accuracy": 0.15790706872940063,
      "num_tokens": 44029524.0,
      "step": 20095
    },
    {
      "entropy": 5.411297464370728,
      "epoch": 1.9948392219134576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004605076276134078,
      "loss": 5.0919,
      "mean_token_accuracy": 0.17263091802597047,
      "num_tokens": 44039523.0,
      "step": 20100
    },
    {
      "entropy": 5.4490697383880615,
      "epoch": 1.9953354505756251,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046048757485989345,
      "loss": 5.2376,
      "mean_token_accuracy": 0.157802115380764,
      "num_tokens": 44050538.0,
      "step": 20105
    },
    {
      "entropy": 5.465253210067749,
      "epoch": 1.9958316792377928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000460467517506687,
      "loss": 5.1987,
      "mean_token_accuracy": 0.16041300892829896,
      "num_tokens": 44061108.0,
      "step": 20110
    },
    {
      "entropy": 5.3616455554962155,
      "epoch": 1.9963279078999603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004604474555542858,
      "loss": 5.1251,
      "mean_token_accuracy": 0.17224353402853013,
      "num_tokens": 44072706.0,
      "step": 20115
    },
    {
      "entropy": 5.395077180862427,
      "epoch": 1.996824136562128,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004604273890031872,
      "loss": 5.1197,
      "mean_token_accuracy": 0.16386050134897232,
      "num_tokens": 44082697.0,
      "step": 20120
    },
    {
      "entropy": 5.376400423049927,
      "epoch": 1.9973203652242955,
      "grad_norm": 1.078125,
      "learning_rate": 0.000460407317853889,
      "loss": 5.0984,
      "mean_token_accuracy": 0.17212993949651717,
      "num_tokens": 44093265.0,
      "step": 20125
    },
    {
      "entropy": 5.35819993019104,
      "epoch": 1.997816593886463,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004603872421068888,
      "loss": 5.099,
      "mean_token_accuracy": 0.16832070648670197,
      "num_tokens": 44103535.0,
      "step": 20130
    },
    {
      "entropy": 5.380383825302124,
      "epoch": 1.9983128225486304,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004603671617626845,
      "loss": 5.0698,
      "mean_token_accuracy": 0.1685510903596878,
      "num_tokens": 44113903.0,
      "step": 20135
    },
    {
      "entropy": 5.405369663238526,
      "epoch": 1.9988090512107979,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046034707682177393,
      "loss": 5.1218,
      "mean_token_accuracy": 0.17167123407125473,
      "num_tokens": 44125148.0,
      "step": 20140
    },
    {
      "entropy": 5.48974347114563,
      "epoch": 1.9993052798729654,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004603269872846554,
      "loss": 5.2313,
      "mean_token_accuracy": 0.15785697251558303,
      "num_tokens": 44136563.0,
      "step": 20145
    },
    {
      "entropy": 5.500528955459595,
      "epoch": 1.9998015085351328,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046030689315182676,
      "loss": 5.1941,
      "mean_token_accuracy": 0.16652860790491103,
      "num_tokens": 44147521.0,
      "step": 20150
    },
    {
      "entropy": 5.442811059951782,
      "epoch": 2.0002977371973003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004602867944237866,
      "loss": 5.1803,
      "mean_token_accuracy": 0.16941146552562714,
      "num_tokens": 44158977.0,
      "step": 20155
    },
    {
      "entropy": 5.398957586288452,
      "epoch": 2.0007939658594682,
      "grad_norm": 1.0,
      "learning_rate": 0.00046026669110103315,
      "loss": 5.1283,
      "mean_token_accuracy": 0.1716260552406311,
      "num_tokens": 44170986.0,
      "step": 20160
    },
    {
      "entropy": 5.498660230636597,
      "epoch": 2.0012901945216357,
      "grad_norm": 1.125,
      "learning_rate": 0.00046024658318406493,
      "loss": 5.1698,
      "mean_token_accuracy": 0.16800285130739212,
      "num_tokens": 44180936.0,
      "step": 20165
    },
    {
      "entropy": 5.493064069747925,
      "epoch": 2.001786423183803,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004602264706733807,
      "loss": 5.1948,
      "mean_token_accuracy": 0.16763851940631866,
      "num_tokens": 44191741.0,
      "step": 20170
    },
    {
      "entropy": 5.409128046035766,
      "epoch": 2.0022826518459707,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004602063535694791,
      "loss": 5.0494,
      "mean_token_accuracy": 0.1663500189781189,
      "num_tokens": 44202310.0,
      "step": 20175
    },
    {
      "entropy": 5.4170574188232425,
      "epoch": 2.002778880508138,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000460186231872859,
      "loss": 5.1008,
      "mean_token_accuracy": 0.1694093883037567,
      "num_tokens": 44213514.0,
      "step": 20180
    },
    {
      "entropy": 5.4764519214630125,
      "epoch": 2.0032751091703056,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046016610558401936,
      "loss": 5.1315,
      "mean_token_accuracy": 0.16340166330337524,
      "num_tokens": 44224718.0,
      "step": 20185
    },
    {
      "entropy": 5.4162029266357425,
      "epoch": 2.003771337832473,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046014597470345935,
      "loss": 5.159,
      "mean_token_accuracy": 0.17170678079128265,
      "num_tokens": 44235999.0,
      "step": 20190
    },
    {
      "entropy": 5.387201738357544,
      "epoch": 2.0042675664946406,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004601258392316781,
      "loss": 5.0106,
      "mean_token_accuracy": 0.17216521203517915,
      "num_tokens": 44246323.0,
      "step": 20195
    },
    {
      "entropy": 5.3632759094238285,
      "epoch": 2.0047637951568085,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046010569916917496,
      "loss": 5.0303,
      "mean_token_accuracy": 0.17731910198926926,
      "num_tokens": 44256323.0,
      "step": 20200
    },
    {
      "entropy": 5.427643251419068,
      "epoch": 2.005260023818976,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004600855545164493,
      "loss": 5.04,
      "mean_token_accuracy": 0.16897486448287963,
      "num_tokens": 44266845.0,
      "step": 20205
    },
    {
      "entropy": 5.383930969238281,
      "epoch": 2.0057562524811434,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004600654052740008,
      "loss": 5.1389,
      "mean_token_accuracy": 0.17412062734365463,
      "num_tokens": 44277861.0,
      "step": 20210
    },
    {
      "entropy": 5.366725826263428,
      "epoch": 2.006252481143311,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004600452514423289,
      "loss": 5.0802,
      "mean_token_accuracy": 0.17173012495040893,
      "num_tokens": 44288511.0,
      "step": 20215
    },
    {
      "entropy": 5.411542558670044,
      "epoch": 2.0067487098054784,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004600250930219336,
      "loss": 5.1692,
      "mean_token_accuracy": 0.15995140224695206,
      "num_tokens": 44299752.0,
      "step": 20220
    },
    {
      "entropy": 5.394157886505127,
      "epoch": 2.007244938467646,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046000493001331465,
      "loss": 5.0229,
      "mean_token_accuracy": 0.17186353504657745,
      "num_tokens": 44309443.0,
      "step": 20225
    },
    {
      "entropy": 5.434767389297486,
      "epoch": 2.0077411671298133,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045998476241697214,
      "loss": 5.0709,
      "mean_token_accuracy": 0.1736485704779625,
      "num_tokens": 44320030.0,
      "step": 20230
    },
    {
      "entropy": 5.3130804061889645,
      "epoch": 2.008237395791981,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004599645902334061,
      "loss": 4.9611,
      "mean_token_accuracy": 0.1789971262216568,
      "num_tokens": 44330077.0,
      "step": 20235
    },
    {
      "entropy": 5.291054534912109,
      "epoch": 2.0087336244541483,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045994441346311677,
      "loss": 5.0483,
      "mean_token_accuracy": 0.17425603717565535,
      "num_tokens": 44340336.0,
      "step": 20240
    },
    {
      "entropy": 5.436720371246338,
      "epoch": 2.009229853116316,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004599242321066045,
      "loss": 5.0367,
      "mean_token_accuracy": 0.17278434038162233,
      "num_tokens": 44350704.0,
      "step": 20245
    },
    {
      "entropy": 5.451545572280883,
      "epoch": 2.0097260817784837,
      "grad_norm": 1.125,
      "learning_rate": 0.00045990404616436976,
      "loss": 5.0729,
      "mean_token_accuracy": 0.17385764718055724,
      "num_tokens": 44360541.0,
      "step": 20250
    },
    {
      "entropy": 5.3082390308380125,
      "epoch": 2.010222310440651,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004598838556369131,
      "loss": 4.9773,
      "mean_token_accuracy": 0.1796424627304077,
      "num_tokens": 44372132.0,
      "step": 20255
    },
    {
      "entropy": 5.312396383285522,
      "epoch": 2.0107185391028186,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045986366052473525,
      "loss": 4.9739,
      "mean_token_accuracy": 0.1804016724228859,
      "num_tokens": 44383594.0,
      "step": 20260
    },
    {
      "entropy": 5.306493854522705,
      "epoch": 2.011214767764986,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004598434608283369,
      "loss": 5.0044,
      "mean_token_accuracy": 0.17730974555015563,
      "num_tokens": 44394101.0,
      "step": 20265
    },
    {
      "entropy": 5.407181596755981,
      "epoch": 2.0117109964271536,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004598232565482191,
      "loss": 5.1186,
      "mean_token_accuracy": 0.17162234634160994,
      "num_tokens": 44405670.0,
      "step": 20270
    },
    {
      "entropy": 5.43667368888855,
      "epoch": 2.012207225089321,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00045980304768488274,
      "loss": 5.0761,
      "mean_token_accuracy": 0.17441981434822082,
      "num_tokens": 44417748.0,
      "step": 20275
    },
    {
      "entropy": 5.437514877319336,
      "epoch": 2.0127034537514885,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045978283423882906,
      "loss": 5.0536,
      "mean_token_accuracy": 0.170981727540493,
      "num_tokens": 44429648.0,
      "step": 20280
    },
    {
      "entropy": 5.3349833488464355,
      "epoch": 2.0131996824136564,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045976261621055915,
      "loss": 5.0706,
      "mean_token_accuracy": 0.17335824370384217,
      "num_tokens": 44440867.0,
      "step": 20285
    },
    {
      "entropy": 5.421066093444824,
      "epoch": 2.013695911075824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045974239360057457,
      "loss": 5.0891,
      "mean_token_accuracy": 0.1703145235776901,
      "num_tokens": 44452981.0,
      "step": 20290
    },
    {
      "entropy": 5.364318943023681,
      "epoch": 2.0141921397379914,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045972216640937676,
      "loss": 5.022,
      "mean_token_accuracy": 0.17762356847524643,
      "num_tokens": 44464243.0,
      "step": 20295
    },
    {
      "entropy": 5.398188829421997,
      "epoch": 2.014688368400159,
      "grad_norm": 1.125,
      "learning_rate": 0.0004597019346374672,
      "loss": 5.1101,
      "mean_token_accuracy": 0.17090481668710708,
      "num_tokens": 44474376.0,
      "step": 20300
    },
    {
      "entropy": 5.388177585601807,
      "epoch": 2.0151845970623263,
      "grad_norm": 1.0,
      "learning_rate": 0.0004596816982853477,
      "loss": 4.9953,
      "mean_token_accuracy": 0.18342289328575134,
      "num_tokens": 44485686.0,
      "step": 20305
    },
    {
      "entropy": 5.309461164474487,
      "epoch": 2.015680825724494,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045966145735352,
      "loss": 5.0115,
      "mean_token_accuracy": 0.180811570584774,
      "num_tokens": 44496096.0,
      "step": 20310
    },
    {
      "entropy": 5.457325077056884,
      "epoch": 2.0161770543866613,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045964121184248604,
      "loss": 5.1213,
      "mean_token_accuracy": 0.16254040896892546,
      "num_tokens": 44507555.0,
      "step": 20315
    },
    {
      "entropy": 5.4204014301300045,
      "epoch": 2.0166732830488288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004596209617527479,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1708361580967903,
      "num_tokens": 44519288.0,
      "step": 20320
    },
    {
      "entropy": 5.415294361114502,
      "epoch": 2.0171695117109962,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004596007070848078,
      "loss": 5.1198,
      "mean_token_accuracy": 0.15992119908332825,
      "num_tokens": 44530146.0,
      "step": 20325
    },
    {
      "entropy": 5.420668268203736,
      "epoch": 2.017665740373164,
      "grad_norm": 1.125,
      "learning_rate": 0.00045958044783916795,
      "loss": 5.1051,
      "mean_token_accuracy": 0.17169228196144104,
      "num_tokens": 44541411.0,
      "step": 20330
    },
    {
      "entropy": 5.359437561035156,
      "epoch": 2.0181619690353316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004595601840163306,
      "loss": 5.0778,
      "mean_token_accuracy": 0.16564609110355377,
      "num_tokens": 44552556.0,
      "step": 20335
    },
    {
      "entropy": 5.38335075378418,
      "epoch": 2.018658197697499,
      "grad_norm": 1.125,
      "learning_rate": 0.00045953991561679855,
      "loss": 5.0874,
      "mean_token_accuracy": 0.16990085691213608,
      "num_tokens": 44563052.0,
      "step": 20340
    },
    {
      "entropy": 5.403846883773804,
      "epoch": 2.0191544263596666,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004595196426410742,
      "loss": 5.0319,
      "mean_token_accuracy": 0.17331536412239074,
      "num_tokens": 44574020.0,
      "step": 20345
    },
    {
      "entropy": 5.361953020095825,
      "epoch": 2.019650655021834,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045949936508966015,
      "loss": 5.0003,
      "mean_token_accuracy": 0.17732471376657485,
      "num_tokens": 44586261.0,
      "step": 20350
    },
    {
      "entropy": 5.4013043403625485,
      "epoch": 2.0201468836840015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004594790829630596,
      "loss": 5.0418,
      "mean_token_accuracy": 0.16869607120752333,
      "num_tokens": 44596391.0,
      "step": 20355
    },
    {
      "entropy": 5.324212503433228,
      "epoch": 2.020643112346169,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004594587962617752,
      "loss": 4.9917,
      "mean_token_accuracy": 0.18051545321941376,
      "num_tokens": 44606263.0,
      "step": 20360
    },
    {
      "entropy": 5.353090238571167,
      "epoch": 2.0211393410083365,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004594385049863102,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1675088569521904,
      "num_tokens": 44617332.0,
      "step": 20365
    },
    {
      "entropy": 5.417057275772095,
      "epoch": 2.021635569670504,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045941820913716766,
      "loss": 5.0908,
      "mean_token_accuracy": 0.1756642460823059,
      "num_tokens": 44629784.0,
      "step": 20370
    },
    {
      "entropy": 5.492676877975464,
      "epoch": 2.022131798332672,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045939790871485093,
      "loss": 5.1862,
      "mean_token_accuracy": 0.16359304040670394,
      "num_tokens": 44641398.0,
      "step": 20375
    },
    {
      "entropy": 5.325419282913208,
      "epoch": 2.0226280269948393,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045937760371986334,
      "loss": 4.9487,
      "mean_token_accuracy": 0.18271668255329132,
      "num_tokens": 44652592.0,
      "step": 20380
    },
    {
      "entropy": 5.412125158309936,
      "epoch": 2.023124255657007,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00045935729415270857,
      "loss": 5.0404,
      "mean_token_accuracy": 0.17700771391391754,
      "num_tokens": 44663896.0,
      "step": 20385
    },
    {
      "entropy": 5.466985702514648,
      "epoch": 2.0236204843191743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004593369800138901,
      "loss": 5.1713,
      "mean_token_accuracy": 0.16308438330888747,
      "num_tokens": 44675004.0,
      "step": 20390
    },
    {
      "entropy": 5.425266218185425,
      "epoch": 2.0241167129813418,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004593166613039118,
      "loss": 5.1052,
      "mean_token_accuracy": 0.17657176852226258,
      "num_tokens": 44686778.0,
      "step": 20395
    },
    {
      "entropy": 5.368587684631348,
      "epoch": 2.0246129416435092,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045929633802327743,
      "loss": 5.0444,
      "mean_token_accuracy": 0.17847374379634856,
      "num_tokens": 44697600.0,
      "step": 20400
    },
    {
      "entropy": 5.360335063934326,
      "epoch": 2.0251091703056767,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000459276010172491,
      "loss": 5.1183,
      "mean_token_accuracy": 0.16835459023714067,
      "num_tokens": 44707453.0,
      "step": 20405
    },
    {
      "entropy": 5.39288444519043,
      "epoch": 2.025605398967844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045925567775205656,
      "loss": 5.1685,
      "mean_token_accuracy": 0.1636221796274185,
      "num_tokens": 44718351.0,
      "step": 20410
    },
    {
      "entropy": 5.43981556892395,
      "epoch": 2.026101627630012,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045923534076247846,
      "loss": 5.013,
      "mean_token_accuracy": 0.17710710763931276,
      "num_tokens": 44728884.0,
      "step": 20415
    },
    {
      "entropy": 5.394426059722901,
      "epoch": 2.0265978562921796,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045921499920426083,
      "loss": 5.041,
      "mean_token_accuracy": 0.17243741303682328,
      "num_tokens": 44740769.0,
      "step": 20420
    },
    {
      "entropy": 5.414039611816406,
      "epoch": 2.027094084954347,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045919465307790817,
      "loss": 5.1738,
      "mean_token_accuracy": 0.16495579481124878,
      "num_tokens": 44753094.0,
      "step": 20425
    },
    {
      "entropy": 5.306338739395142,
      "epoch": 2.0275903136165145,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045917430238392505,
      "loss": 4.9278,
      "mean_token_accuracy": 0.177229605615139,
      "num_tokens": 44762996.0,
      "step": 20430
    },
    {
      "entropy": 5.35363917350769,
      "epoch": 2.028086542278682,
      "grad_norm": 0.859375,
      "learning_rate": 0.000459153947122816,
      "loss": 5.1115,
      "mean_token_accuracy": 0.171942338347435,
      "num_tokens": 44775343.0,
      "step": 20435
    },
    {
      "entropy": 5.31217212677002,
      "epoch": 2.0285827709408495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004591335872950859,
      "loss": 4.9533,
      "mean_token_accuracy": 0.17918279618024827,
      "num_tokens": 44785383.0,
      "step": 20440
    },
    {
      "entropy": 5.322952461242676,
      "epoch": 2.029078999603017,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004591132229012397,
      "loss": 5.0181,
      "mean_token_accuracy": 0.17781300246715545,
      "num_tokens": 44796805.0,
      "step": 20445
    },
    {
      "entropy": 5.391804599761963,
      "epoch": 2.0295752282651844,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045909285394178217,
      "loss": 5.0254,
      "mean_token_accuracy": 0.17650724053382874,
      "num_tokens": 44806800.0,
      "step": 20450
    },
    {
      "entropy": 5.434646320343018,
      "epoch": 2.030071456927352,
      "grad_norm": 1.0,
      "learning_rate": 0.0004590724804172187,
      "loss": 5.1638,
      "mean_token_accuracy": 0.16380602419376372,
      "num_tokens": 44818745.0,
      "step": 20455
    },
    {
      "entropy": 5.312346792221069,
      "epoch": 2.03056768558952,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004590521023280543,
      "loss": 5.0121,
      "mean_token_accuracy": 0.17351637035608292,
      "num_tokens": 44829825.0,
      "step": 20460
    },
    {
      "entropy": 5.326686859130859,
      "epoch": 2.0310639142516873,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045903171967479425,
      "loss": 5.0102,
      "mean_token_accuracy": 0.18013980984687805,
      "num_tokens": 44841742.0,
      "step": 20465
    },
    {
      "entropy": 5.413286113739014,
      "epoch": 2.031560142913855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045901133245794426,
      "loss": 5.1369,
      "mean_token_accuracy": 0.1746071234345436,
      "num_tokens": 44853019.0,
      "step": 20470
    },
    {
      "entropy": 5.388130235671997,
      "epoch": 2.0320563715760223,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004589909406780096,
      "loss": 4.9888,
      "mean_token_accuracy": 0.17771071344614028,
      "num_tokens": 44863147.0,
      "step": 20475
    },
    {
      "entropy": 5.398185300827026,
      "epoch": 2.0325526002381897,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045897054433549617,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1790820464491844,
      "num_tokens": 44873638.0,
      "step": 20480
    },
    {
      "entropy": 5.326075839996338,
      "epoch": 2.033048828900357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045895014343090963,
      "loss": 5.0701,
      "mean_token_accuracy": 0.1677207112312317,
      "num_tokens": 44885752.0,
      "step": 20485
    },
    {
      "entropy": 5.347051286697388,
      "epoch": 2.0335450575625247,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045892973796475594,
      "loss": 5.0179,
      "mean_token_accuracy": 0.17252234667539595,
      "num_tokens": 44897471.0,
      "step": 20490
    },
    {
      "entropy": 5.4466126441955565,
      "epoch": 2.034041286224692,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000458909327937541,
      "loss": 5.0931,
      "mean_token_accuracy": 0.16723982840776444,
      "num_tokens": 44907236.0,
      "step": 20495
    },
    {
      "entropy": 5.338018178939819,
      "epoch": 2.03453751488686,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045888891334977105,
      "loss": 5.0092,
      "mean_token_accuracy": 0.18057094216346742,
      "num_tokens": 44918103.0,
      "step": 20500
    },
    {
      "entropy": 5.323795127868652,
      "epoch": 2.0350337435490276,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004588684942019524,
      "loss": 5.0841,
      "mean_token_accuracy": 0.17172406911849974,
      "num_tokens": 44928263.0,
      "step": 20505
    },
    {
      "entropy": 5.378190040588379,
      "epoch": 2.035529972211195,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045884807049459116,
      "loss": 5.0535,
      "mean_token_accuracy": 0.17366544306278228,
      "num_tokens": 44939487.0,
      "step": 20510
    },
    {
      "entropy": 5.393610858917237,
      "epoch": 2.0360262008733625,
      "grad_norm": 1.046875,
      "learning_rate": 0.000458827642228194,
      "loss": 5.1546,
      "mean_token_accuracy": 0.1663413390517235,
      "num_tokens": 44949749.0,
      "step": 20515
    },
    {
      "entropy": 5.39757924079895,
      "epoch": 2.03652242953553,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045880720940326744,
      "loss": 5.054,
      "mean_token_accuracy": 0.17095716893672944,
      "num_tokens": 44960347.0,
      "step": 20520
    },
    {
      "entropy": 5.458985662460327,
      "epoch": 2.0370186581976975,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045878677202031804,
      "loss": 5.1364,
      "mean_token_accuracy": 0.16703088730573654,
      "num_tokens": 44970420.0,
      "step": 20525
    },
    {
      "entropy": 5.360573196411133,
      "epoch": 2.037514886859865,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004587663300798529,
      "loss": 5.0499,
      "mean_token_accuracy": 0.17774952650070192,
      "num_tokens": 44980888.0,
      "step": 20530
    },
    {
      "entropy": 5.386511278152466,
      "epoch": 2.0380111155220324,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004587458835823786,
      "loss": 5.0586,
      "mean_token_accuracy": 0.17238570898771285,
      "num_tokens": 44992721.0,
      "step": 20535
    },
    {
      "entropy": 5.314067697525024,
      "epoch": 2.0385073441842,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045872543252840236,
      "loss": 4.996,
      "mean_token_accuracy": 0.17551333606243133,
      "num_tokens": 45003603.0,
      "step": 20540
    },
    {
      "entropy": 5.35052981376648,
      "epoch": 2.039003572846368,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045870497691843134,
      "loss": 5.054,
      "mean_token_accuracy": 0.1809846892952919,
      "num_tokens": 45014864.0,
      "step": 20545
    },
    {
      "entropy": 5.342767715454102,
      "epoch": 2.0394998015085353,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004586845167529727,
      "loss": 5.0966,
      "mean_token_accuracy": 0.1709667131304741,
      "num_tokens": 45027766.0,
      "step": 20550
    },
    {
      "entropy": 5.359860420227051,
      "epoch": 2.0399960301707027,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00045866405203253384,
      "loss": 5.0456,
      "mean_token_accuracy": 0.17250837832689286,
      "num_tokens": 45039524.0,
      "step": 20555
    },
    {
      "entropy": 5.409212446212768,
      "epoch": 2.04049225883287,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004586435827576223,
      "loss": 5.0936,
      "mean_token_accuracy": 0.16889739036560059,
      "num_tokens": 45050848.0,
      "step": 20560
    },
    {
      "entropy": 5.3685081005096436,
      "epoch": 2.0409884874950377,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004586231089287456,
      "loss": 5.0222,
      "mean_token_accuracy": 0.17166829854249954,
      "num_tokens": 45061229.0,
      "step": 20565
    },
    {
      "entropy": 5.329183864593506,
      "epoch": 2.041484716157205,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004586026305464115,
      "loss": 4.9682,
      "mean_token_accuracy": 0.17832460552453994,
      "num_tokens": 45070466.0,
      "step": 20570
    },
    {
      "entropy": 5.37541880607605,
      "epoch": 2.0419809448193726,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045858214761112777,
      "loss": 5.0682,
      "mean_token_accuracy": 0.17065610587596894,
      "num_tokens": 45081839.0,
      "step": 20575
    },
    {
      "entropy": 5.432844638824463,
      "epoch": 2.04247717348154,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045856166012340235,
      "loss": 5.1237,
      "mean_token_accuracy": 0.17451781630516053,
      "num_tokens": 45092458.0,
      "step": 20580
    },
    {
      "entropy": 5.434857749938965,
      "epoch": 2.0429734021437076,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004585411680837433,
      "loss": 5.1115,
      "mean_token_accuracy": 0.16583824008703232,
      "num_tokens": 45104629.0,
      "step": 20585
    },
    {
      "entropy": 5.356353378295898,
      "epoch": 2.0434696308058755,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045852067149265875,
      "loss": 4.9517,
      "mean_token_accuracy": 0.17414254546165467,
      "num_tokens": 45115556.0,
      "step": 20590
    },
    {
      "entropy": 5.311757898330688,
      "epoch": 2.043965859468043,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000458500170350657,
      "loss": 5.0322,
      "mean_token_accuracy": 0.17146580368280412,
      "num_tokens": 45126489.0,
      "step": 20595
    },
    {
      "entropy": 5.396221351623535,
      "epoch": 2.0444620881302105,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045847966465824654,
      "loss": 5.1504,
      "mean_token_accuracy": 0.1669980302453041,
      "num_tokens": 45138882.0,
      "step": 20600
    },
    {
      "entropy": 5.457683277130127,
      "epoch": 2.044958316792378,
      "grad_norm": 1.125,
      "learning_rate": 0.0004584591544159356,
      "loss": 5.0841,
      "mean_token_accuracy": 0.171192866563797,
      "num_tokens": 45149941.0,
      "step": 20605
    },
    {
      "entropy": 5.458730316162109,
      "epoch": 2.0454545454545454,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004584386396242331,
      "loss": 5.1125,
      "mean_token_accuracy": 0.16964179575443267,
      "num_tokens": 45162267.0,
      "step": 20610
    },
    {
      "entropy": 5.386421537399292,
      "epoch": 2.045950774116713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045841812028364756,
      "loss": 5.188,
      "mean_token_accuracy": 0.16593748331069946,
      "num_tokens": 45174444.0,
      "step": 20615
    },
    {
      "entropy": 5.386126470565796,
      "epoch": 2.0464470027788804,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045839759639468774,
      "loss": 5.0454,
      "mean_token_accuracy": 0.17360633611679077,
      "num_tokens": 45183957.0,
      "step": 20620
    },
    {
      "entropy": 5.441244983673096,
      "epoch": 2.046943231441048,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004583770679578628,
      "loss": 5.1348,
      "mean_token_accuracy": 0.17370228469371796,
      "num_tokens": 45195607.0,
      "step": 20625
    },
    {
      "entropy": 5.375789213180542,
      "epoch": 2.0474394601032158,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004583565349736818,
      "loss": 5.0327,
      "mean_token_accuracy": 0.1719804435968399,
      "num_tokens": 45205627.0,
      "step": 20630
    },
    {
      "entropy": 5.3298228740692135,
      "epoch": 2.0479356887653832,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004583359974426538,
      "loss": 5.0701,
      "mean_token_accuracy": 0.167018760740757,
      "num_tokens": 45216331.0,
      "step": 20635
    },
    {
      "entropy": 5.3914906024932865,
      "epoch": 2.0484319174275507,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045831545536528806,
      "loss": 5.1055,
      "mean_token_accuracy": 0.1661143943667412,
      "num_tokens": 45227826.0,
      "step": 20640
    },
    {
      "entropy": 5.360878896713257,
      "epoch": 2.048928146089718,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045829490874209404,
      "loss": 5.046,
      "mean_token_accuracy": 0.17078365087509156,
      "num_tokens": 45238093.0,
      "step": 20645
    },
    {
      "entropy": 5.334298849105835,
      "epoch": 2.0494243747518857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045827435757358136,
      "loss": 5.061,
      "mean_token_accuracy": 0.17784079015254975,
      "num_tokens": 45248660.0,
      "step": 20650
    },
    {
      "entropy": 5.4594604015350345,
      "epoch": 2.049920603414053,
      "grad_norm": 1.0,
      "learning_rate": 0.0004582538018602594,
      "loss": 5.1437,
      "mean_token_accuracy": 0.16608968675136565,
      "num_tokens": 45261059.0,
      "step": 20655
    },
    {
      "entropy": 5.396690130233765,
      "epoch": 2.0504168320762206,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004582332416026381,
      "loss": 5.0334,
      "mean_token_accuracy": 0.17018125355243682,
      "num_tokens": 45272210.0,
      "step": 20660
    },
    {
      "entropy": 5.38276424407959,
      "epoch": 2.050913060738388,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045821267680122725,
      "loss": 5.1233,
      "mean_token_accuracy": 0.17033070474863052,
      "num_tokens": 45283488.0,
      "step": 20665
    },
    {
      "entropy": 5.341908121109009,
      "epoch": 2.0514092894005556,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004581921074565368,
      "loss": 4.9912,
      "mean_token_accuracy": 0.17574659287929534,
      "num_tokens": 45292824.0,
      "step": 20670
    },
    {
      "entropy": 5.348388957977295,
      "epoch": 2.0519055180627235,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004581715335690768,
      "loss": 5.0519,
      "mean_token_accuracy": 0.17487603425979614,
      "num_tokens": 45303102.0,
      "step": 20675
    },
    {
      "entropy": 5.374080038070678,
      "epoch": 2.052401746724891,
      "grad_norm": 1.0,
      "learning_rate": 0.00045815095513935747,
      "loss": 4.9811,
      "mean_token_accuracy": 0.16990201771259308,
      "num_tokens": 45313825.0,
      "step": 20680
    },
    {
      "entropy": 5.362378549575806,
      "epoch": 2.0528979753870584,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004581303721678891,
      "loss": 5.0156,
      "mean_token_accuracy": 0.1811266139149666,
      "num_tokens": 45325789.0,
      "step": 20685
    },
    {
      "entropy": 5.3462786197662355,
      "epoch": 2.053394204049226,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004581097846551821,
      "loss": 5.1056,
      "mean_token_accuracy": 0.16642937958240508,
      "num_tokens": 45337630.0,
      "step": 20690
    },
    {
      "entropy": 5.447678661346435,
      "epoch": 2.0538904327113934,
      "grad_norm": 1.0625,
      "learning_rate": 0.000458089192601747,
      "loss": 5.1375,
      "mean_token_accuracy": 0.16379803270101548,
      "num_tokens": 45349815.0,
      "step": 20695
    },
    {
      "entropy": 5.360766983032226,
      "epoch": 2.054386661373561,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045806859600809455,
      "loss": 5.0645,
      "mean_token_accuracy": 0.17003174424171447,
      "num_tokens": 45360741.0,
      "step": 20700
    },
    {
      "entropy": 5.310501909255981,
      "epoch": 2.0548828900357283,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004580479948747353,
      "loss": 4.9645,
      "mean_token_accuracy": 0.1794826626777649,
      "num_tokens": 45371258.0,
      "step": 20705
    },
    {
      "entropy": 5.376096105575561,
      "epoch": 2.055379118697896,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045802738920218015,
      "loss": 5.068,
      "mean_token_accuracy": 0.17134176790714264,
      "num_tokens": 45381914.0,
      "step": 20710
    },
    {
      "entropy": 5.318615484237671,
      "epoch": 2.0558753473600637,
      "grad_norm": 1.125,
      "learning_rate": 0.0004580067789909403,
      "loss": 5.0783,
      "mean_token_accuracy": 0.1762777239084244,
      "num_tokens": 45393159.0,
      "step": 20715
    },
    {
      "entropy": 5.333478498458862,
      "epoch": 2.056371576022231,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045798616424152654,
      "loss": 4.9464,
      "mean_token_accuracy": 0.17810809463262559,
      "num_tokens": 45403076.0,
      "step": 20720
    },
    {
      "entropy": 5.370905017852783,
      "epoch": 2.0568678046843987,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045796554495445023,
      "loss": 5.099,
      "mean_token_accuracy": 0.1692257195711136,
      "num_tokens": 45412962.0,
      "step": 20725
    },
    {
      "entropy": 5.348581266403198,
      "epoch": 2.057364033346566,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004579449211302228,
      "loss": 5.0199,
      "mean_token_accuracy": 0.17381089627742768,
      "num_tokens": 45423620.0,
      "step": 20730
    },
    {
      "entropy": 5.363291883468628,
      "epoch": 2.0578602620087336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004579242927693554,
      "loss": 5.0632,
      "mean_token_accuracy": 0.16720576584339142,
      "num_tokens": 45435901.0,
      "step": 20735
    },
    {
      "entropy": 5.365499448776245,
      "epoch": 2.058356490670901,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004579036598723598,
      "loss": 4.9757,
      "mean_token_accuracy": 0.18095103800296783,
      "num_tokens": 45445409.0,
      "step": 20740
    },
    {
      "entropy": 5.456441593170166,
      "epoch": 2.0588527193330686,
      "grad_norm": 1.671875,
      "learning_rate": 0.00045788302243974745,
      "loss": 5.1409,
      "mean_token_accuracy": 0.17394131273031235,
      "num_tokens": 45455794.0,
      "step": 20745
    },
    {
      "entropy": 5.401806783676148,
      "epoch": 2.059348947995236,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004578623804720302,
      "loss": 5.0086,
      "mean_token_accuracy": 0.1749672457575798,
      "num_tokens": 45465922.0,
      "step": 20750
    },
    {
      "entropy": 5.3354902267456055,
      "epoch": 2.0598451766574035,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004578417339697201,
      "loss": 5.1148,
      "mean_token_accuracy": 0.16516130268573762,
      "num_tokens": 45476314.0,
      "step": 20755
    },
    {
      "entropy": 5.38496961593628,
      "epoch": 2.0603414053195714,
      "grad_norm": 1.125,
      "learning_rate": 0.00045782108293332884,
      "loss": 5.0668,
      "mean_token_accuracy": 0.17044861763715743,
      "num_tokens": 45487232.0,
      "step": 20760
    },
    {
      "entropy": 5.459442758560181,
      "epoch": 2.060837633981739,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004578004273633687,
      "loss": 5.119,
      "mean_token_accuracy": 0.1739198759198189,
      "num_tokens": 45498368.0,
      "step": 20765
    },
    {
      "entropy": 5.3398278713226315,
      "epoch": 2.0613338626439064,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045777976726035194,
      "loss": 5.0567,
      "mean_token_accuracy": 0.17328181117773056,
      "num_tokens": 45507600.0,
      "step": 20770
    },
    {
      "entropy": 5.399073362350464,
      "epoch": 2.061830091306074,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004577591026247907,
      "loss": 5.136,
      "mean_token_accuracy": 0.16608935743570327,
      "num_tokens": 45518594.0,
      "step": 20775
    },
    {
      "entropy": 5.376103162765503,
      "epoch": 2.0623263199682413,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004577384334571976,
      "loss": 5.0252,
      "mean_token_accuracy": 0.1730777531862259,
      "num_tokens": 45529929.0,
      "step": 20780
    },
    {
      "entropy": 5.373984241485596,
      "epoch": 2.062822548630409,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004577177597580851,
      "loss": 5.0359,
      "mean_token_accuracy": 0.17800827771425248,
      "num_tokens": 45540228.0,
      "step": 20785
    },
    {
      "entropy": 5.296068382263184,
      "epoch": 2.0633187772925763,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045769708152796593,
      "loss": 4.9897,
      "mean_token_accuracy": 0.176832914352417,
      "num_tokens": 45552026.0,
      "step": 20790
    },
    {
      "entropy": 5.315315628051758,
      "epoch": 2.0638150059547438,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045767639876735276,
      "loss": 5.0982,
      "mean_token_accuracy": 0.1734484165906906,
      "num_tokens": 45563579.0,
      "step": 20795
    },
    {
      "entropy": 5.4521242618560795,
      "epoch": 2.0643112346169117,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004576557114767585,
      "loss": 5.1519,
      "mean_token_accuracy": 0.16390457451343537,
      "num_tokens": 45573800.0,
      "step": 20800
    },
    {
      "entropy": 5.363963556289673,
      "epoch": 2.064807463279079,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045763501965669627,
      "loss": 5.0243,
      "mean_token_accuracy": 0.17761609852313995,
      "num_tokens": 45583758.0,
      "step": 20805
    },
    {
      "entropy": 5.37861123085022,
      "epoch": 2.0653036919412466,
      "grad_norm": 0.984375,
      "learning_rate": 0.000457614323307679,
      "loss": 5.0229,
      "mean_token_accuracy": 0.17040371596813203,
      "num_tokens": 45595154.0,
      "step": 20810
    },
    {
      "entropy": 5.43823390007019,
      "epoch": 2.065799920603414,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045759362243022,
      "loss": 5.1435,
      "mean_token_accuracy": 0.16895062178373338,
      "num_tokens": 45606557.0,
      "step": 20815
    },
    {
      "entropy": 5.3312615871429445,
      "epoch": 2.0662961492655816,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004575729170248327,
      "loss": 4.9848,
      "mean_token_accuracy": 0.1815398469567299,
      "num_tokens": 45615531.0,
      "step": 20820
    },
    {
      "entropy": 5.281439208984375,
      "epoch": 2.066792377927749,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004575522070920305,
      "loss": 4.9847,
      "mean_token_accuracy": 0.17401143461465834,
      "num_tokens": 45626512.0,
      "step": 20825
    },
    {
      "entropy": 5.43540244102478,
      "epoch": 2.0672886065899165,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045753149263232687,
      "loss": 5.1206,
      "mean_token_accuracy": 0.16561357378959657,
      "num_tokens": 45638005.0,
      "step": 20830
    },
    {
      "entropy": 5.385746145248413,
      "epoch": 2.067784835252084,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004575107736462355,
      "loss": 5.1048,
      "mean_token_accuracy": 0.16425144523382187,
      "num_tokens": 45647740.0,
      "step": 20835
    },
    {
      "entropy": 5.337829208374023,
      "epoch": 2.0682810639142515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004574900501342703,
      "loss": 5.0189,
      "mean_token_accuracy": 0.17689135670661926,
      "num_tokens": 45659021.0,
      "step": 20840
    },
    {
      "entropy": 5.379702568054199,
      "epoch": 2.0687772925764194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004574693220969451,
      "loss": 5.0981,
      "mean_token_accuracy": 0.1678236484527588,
      "num_tokens": 45669413.0,
      "step": 20845
    },
    {
      "entropy": 5.454207515716552,
      "epoch": 2.069273521238587,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004574485895347738,
      "loss": 5.1654,
      "mean_token_accuracy": 0.1715831220149994,
      "num_tokens": 45680719.0,
      "step": 20850
    },
    {
      "entropy": 5.325667095184326,
      "epoch": 2.0697697499007544,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045742785244827066,
      "loss": 4.9891,
      "mean_token_accuracy": 0.17648353725671767,
      "num_tokens": 45690393.0,
      "step": 20855
    },
    {
      "entropy": 5.375811576843262,
      "epoch": 2.070265978562922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004574071108379499,
      "loss": 5.1134,
      "mean_token_accuracy": 0.16967443972826005,
      "num_tokens": 45701304.0,
      "step": 20860
    },
    {
      "entropy": 5.399863433837891,
      "epoch": 2.0707622072250893,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004573863647043259,
      "loss": 4.9464,
      "mean_token_accuracy": 0.17845842391252517,
      "num_tokens": 45711892.0,
      "step": 20865
    },
    {
      "entropy": 5.295188283920288,
      "epoch": 2.0712584358872568,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004573656140479129,
      "loss": 5.051,
      "mean_token_accuracy": 0.1720167800784111,
      "num_tokens": 45723693.0,
      "step": 20870
    },
    {
      "entropy": 5.3117334842681885,
      "epoch": 2.0717546645494243,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004573448588692258,
      "loss": 5.0293,
      "mean_token_accuracy": 0.17116363048553468,
      "num_tokens": 45734392.0,
      "step": 20875
    },
    {
      "entropy": 5.314636850357056,
      "epoch": 2.0722508932115917,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000457324099168779,
      "loss": 4.9525,
      "mean_token_accuracy": 0.18275747448205948,
      "num_tokens": 45745732.0,
      "step": 20880
    },
    {
      "entropy": 5.308532333374023,
      "epoch": 2.0727471218737596,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004573033349470874,
      "loss": 5.0289,
      "mean_token_accuracy": 0.1781419634819031,
      "num_tokens": 45756671.0,
      "step": 20885
    },
    {
      "entropy": 5.433636617660523,
      "epoch": 2.073243350535927,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004572825662046661,
      "loss": 5.1118,
      "mean_token_accuracy": 0.16690876483917236,
      "num_tokens": 45767685.0,
      "step": 20890
    },
    {
      "entropy": 5.409076499938965,
      "epoch": 2.0737395791980946,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004572617929420298,
      "loss": 5.0321,
      "mean_token_accuracy": 0.17290572375059127,
      "num_tokens": 45777251.0,
      "step": 20895
    },
    {
      "entropy": 5.390455865859986,
      "epoch": 2.074235807860262,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045724101515969375,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17058699801564217,
      "num_tokens": 45790037.0,
      "step": 20900
    },
    {
      "entropy": 5.411048173904419,
      "epoch": 2.0747320365224295,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045722023285817327,
      "loss": 5.1495,
      "mean_token_accuracy": 0.164389406144619,
      "num_tokens": 45800803.0,
      "step": 20905
    },
    {
      "entropy": 5.441731977462768,
      "epoch": 2.075228265184597,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004571994460379837,
      "loss": 5.1038,
      "mean_token_accuracy": 0.1636028692126274,
      "num_tokens": 45812447.0,
      "step": 20910
    },
    {
      "entropy": 5.388584566116333,
      "epoch": 2.0757244938467645,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045717865469964045,
      "loss": 5.0496,
      "mean_token_accuracy": 0.16763579696416855,
      "num_tokens": 45822172.0,
      "step": 20915
    },
    {
      "entropy": 5.452678108215332,
      "epoch": 2.076220722508932,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045715785884365914,
      "loss": 5.1699,
      "mean_token_accuracy": 0.17074971795082092,
      "num_tokens": 45832583.0,
      "step": 20920
    },
    {
      "entropy": 5.449528360366822,
      "epoch": 2.0767169511710994,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045713705847055537,
      "loss": 5.1852,
      "mean_token_accuracy": 0.1714400753378868,
      "num_tokens": 45842640.0,
      "step": 20925
    },
    {
      "entropy": 5.467917728424072,
      "epoch": 2.0772131798332674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004571162535808451,
      "loss": 5.1728,
      "mean_token_accuracy": 0.16633205860853195,
      "num_tokens": 45854618.0,
      "step": 20930
    },
    {
      "entropy": 5.395715808868408,
      "epoch": 2.077709408495435,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004570954441750441,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1727788120508194,
      "num_tokens": 45866130.0,
      "step": 20935
    },
    {
      "entropy": 5.397499132156372,
      "epoch": 2.0782056371576023,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004570746302536685,
      "loss": 5.0797,
      "mean_token_accuracy": 0.1726262927055359,
      "num_tokens": 45877187.0,
      "step": 20940
    },
    {
      "entropy": 5.355258417129517,
      "epoch": 2.07870186581977,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045705381181723444,
      "loss": 5.1088,
      "mean_token_accuracy": 0.17266033291816713,
      "num_tokens": 45889557.0,
      "step": 20945
    },
    {
      "entropy": 5.379432487487793,
      "epoch": 2.0791980944819373,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045703298886625804,
      "loss": 5.088,
      "mean_token_accuracy": 0.16821659356355667,
      "num_tokens": 45900860.0,
      "step": 20950
    },
    {
      "entropy": 5.37594575881958,
      "epoch": 2.0796943231441047,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004570121614012558,
      "loss": 5.1089,
      "mean_token_accuracy": 0.16755294650793076,
      "num_tokens": 45912231.0,
      "step": 20955
    },
    {
      "entropy": 5.443831491470337,
      "epoch": 2.080190551806272,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045699132942274425,
      "loss": 5.1593,
      "mean_token_accuracy": 0.16480531990528108,
      "num_tokens": 45922722.0,
      "step": 20960
    },
    {
      "entropy": 5.3428138256072994,
      "epoch": 2.0806867804684397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045697049293123984,
      "loss": 5.0469,
      "mean_token_accuracy": 0.17374687641859055,
      "num_tokens": 45932444.0,
      "step": 20965
    },
    {
      "entropy": 5.415097284317016,
      "epoch": 2.0811830091306076,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004569496519272593,
      "loss": 5.0691,
      "mean_token_accuracy": 0.1689396843314171,
      "num_tokens": 45942826.0,
      "step": 20970
    },
    {
      "entropy": 5.381955528259278,
      "epoch": 2.081679237792775,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045692880641131945,
      "loss": 5.067,
      "mean_token_accuracy": 0.17469535171985626,
      "num_tokens": 45953935.0,
      "step": 20975
    },
    {
      "entropy": 5.419738626480102,
      "epoch": 2.0821754664549426,
      "grad_norm": 1.0,
      "learning_rate": 0.0004569079563839372,
      "loss": 5.1686,
      "mean_token_accuracy": 0.16183257102966309,
      "num_tokens": 45964693.0,
      "step": 20980
    },
    {
      "entropy": 5.474231386184693,
      "epoch": 2.08267169511711,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045688710184562965,
      "loss": 5.0797,
      "mean_token_accuracy": 0.17011580020189285,
      "num_tokens": 45976481.0,
      "step": 20985
    },
    {
      "entropy": 5.441715860366822,
      "epoch": 2.0831679237792775,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045686624279691387,
      "loss": 5.1586,
      "mean_token_accuracy": 0.16925472617149354,
      "num_tokens": 45988879.0,
      "step": 20990
    },
    {
      "entropy": 5.345638370513916,
      "epoch": 2.083664152441445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004568453792383072,
      "loss": 5.0284,
      "mean_token_accuracy": 0.17838707864284514,
      "num_tokens": 45999332.0,
      "step": 20995
    },
    {
      "entropy": 5.414608955383301,
      "epoch": 2.0841603811036125,
      "grad_norm": 1.03125,
      "learning_rate": 0.000456824511170327,
      "loss": 5.0549,
      "mean_token_accuracy": 0.1745280608534813,
      "num_tokens": 46009841.0,
      "step": 21000
    },
    {
      "epoch": 2.0841603811036125,
      "eval_entropy": 5.256798637142712,
      "eval_loss": 5.2463603019714355,
      "eval_mean_token_accuracy": 0.1691076588686613,
      "eval_num_tokens": 46009841.0,
      "eval_runtime": 26.7133,
      "eval_samples_per_second": 1260.384,
      "eval_steps_per_second": 157.562,
      "step": 21000
    },
    {
      "entropy": 5.337066984176635,
      "epoch": 2.08465660976578,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004568036385934906,
      "loss": 4.9976,
      "mean_token_accuracy": 0.1724056273698807,
      "num_tokens": 46020678.0,
      "step": 21005
    },
    {
      "entropy": 5.38824872970581,
      "epoch": 2.0851528384279474,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045678276150831586,
      "loss": 5.0764,
      "mean_token_accuracy": 0.17251751571893692,
      "num_tokens": 46032036.0,
      "step": 21010
    },
    {
      "entropy": 5.442033290863037,
      "epoch": 2.0856490670901153,
      "grad_norm": 1.0,
      "learning_rate": 0.0004567618799153202,
      "loss": 5.0679,
      "mean_token_accuracy": 0.17296954095363617,
      "num_tokens": 46042675.0,
      "step": 21015
    },
    {
      "entropy": 5.299561882019043,
      "epoch": 2.086145295752283,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045674099381502167,
      "loss": 5.0694,
      "mean_token_accuracy": 0.17213140726089476,
      "num_tokens": 46053616.0,
      "step": 21020
    },
    {
      "entropy": 5.4043221950531,
      "epoch": 2.0866415244144503,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004567201032079381,
      "loss": 5.0914,
      "mean_token_accuracy": 0.17337000221014023,
      "num_tokens": 46063755.0,
      "step": 21025
    },
    {
      "entropy": 5.4606891632080075,
      "epoch": 2.0871377530766178,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045669920809458754,
      "loss": 5.0818,
      "mean_token_accuracy": 0.1752253755927086,
      "num_tokens": 46074459.0,
      "step": 21030
    },
    {
      "entropy": 5.501385307312011,
      "epoch": 2.0876339817387852,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004566783084754881,
      "loss": 5.1623,
      "mean_token_accuracy": 0.16770317554473876,
      "num_tokens": 46085053.0,
      "step": 21035
    },
    {
      "entropy": 5.320533275604248,
      "epoch": 2.0881302104009527,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045665740435115815,
      "loss": 4.9688,
      "mean_token_accuracy": 0.18281051218509675,
      "num_tokens": 46094697.0,
      "step": 21040
    },
    {
      "entropy": 5.330243158340454,
      "epoch": 2.08862643906312,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000456636495722116,
      "loss": 5.1015,
      "mean_token_accuracy": 0.1692603498697281,
      "num_tokens": 46105617.0,
      "step": 21045
    },
    {
      "entropy": 5.438247776031494,
      "epoch": 2.0891226677252877,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004566155825888801,
      "loss": 5.0597,
      "mean_token_accuracy": 0.1721243292093277,
      "num_tokens": 46115450.0,
      "step": 21050
    },
    {
      "entropy": 5.397001457214356,
      "epoch": 2.089618896387455,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004565946649519691,
      "loss": 5.134,
      "mean_token_accuracy": 0.17068720757961273,
      "num_tokens": 46126511.0,
      "step": 21055
    },
    {
      "entropy": 5.337567234039307,
      "epoch": 2.090115125049623,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045657374281190177,
      "loss": 5.0194,
      "mean_token_accuracy": 0.17652539908885956,
      "num_tokens": 46136294.0,
      "step": 21060
    },
    {
      "entropy": 5.34055905342102,
      "epoch": 2.0906113537117905,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004565528161691969,
      "loss": 5.0945,
      "mean_token_accuracy": 0.16248857975006104,
      "num_tokens": 46148926.0,
      "step": 21065
    },
    {
      "entropy": 5.417394161224365,
      "epoch": 2.091107582373958,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004565318850243733,
      "loss": 5.1027,
      "mean_token_accuracy": 0.17379750907421113,
      "num_tokens": 46160130.0,
      "step": 21070
    },
    {
      "entropy": 5.337324810028076,
      "epoch": 2.0916038110361255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004565109493779502,
      "loss": 5.0355,
      "mean_token_accuracy": 0.17481701523065568,
      "num_tokens": 46169919.0,
      "step": 21075
    },
    {
      "entropy": 5.383826208114624,
      "epoch": 2.092100039698293,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045649000923044664,
      "loss": 5.1456,
      "mean_token_accuracy": 0.16622894555330275,
      "num_tokens": 46181186.0,
      "step": 21080
    },
    {
      "entropy": 5.385074281692505,
      "epoch": 2.0925962683604604,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045646906458238196,
      "loss": 5.0962,
      "mean_token_accuracy": 0.16944107562303543,
      "num_tokens": 46192172.0,
      "step": 21085
    },
    {
      "entropy": 5.372131586074829,
      "epoch": 2.093092497022628,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004564481154342754,
      "loss": 5.0865,
      "mean_token_accuracy": 0.16634039133787154,
      "num_tokens": 46202678.0,
      "step": 21090
    },
    {
      "entropy": 5.40507869720459,
      "epoch": 2.0935887256847954,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004564271617866467,
      "loss": 5.0789,
      "mean_token_accuracy": 0.1733351990580559,
      "num_tokens": 46212213.0,
      "step": 21095
    },
    {
      "entropy": 5.4373393058776855,
      "epoch": 2.0940849543469633,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045640620364001524,
      "loss": 5.1214,
      "mean_token_accuracy": 0.16820697486400604,
      "num_tokens": 46223044.0,
      "step": 21100
    },
    {
      "entropy": 5.427464532852173,
      "epoch": 2.0945811830091308,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045638524099490086,
      "loss": 5.1061,
      "mean_token_accuracy": 0.16967812180519104,
      "num_tokens": 46236008.0,
      "step": 21105
    },
    {
      "entropy": 5.384048748016357,
      "epoch": 2.0950774116712982,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045636427385182334,
      "loss": 5.0643,
      "mean_token_accuracy": 0.17106099277734757,
      "num_tokens": 46247631.0,
      "step": 21110
    },
    {
      "entropy": 5.401796054840088,
      "epoch": 2.0955736403334657,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045634330221130265,
      "loss": 5.0327,
      "mean_token_accuracy": 0.17042427510023117,
      "num_tokens": 46258294.0,
      "step": 21115
    },
    {
      "entropy": 5.3495698928833,
      "epoch": 2.096069868995633,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004563223260738589,
      "loss": 5.0408,
      "mean_token_accuracy": 0.16641611307859422,
      "num_tokens": 46269611.0,
      "step": 21120
    },
    {
      "entropy": 5.35426778793335,
      "epoch": 2.0965660976578007,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000456301345440012,
      "loss": 5.1292,
      "mean_token_accuracy": 0.17540946900844573,
      "num_tokens": 46280731.0,
      "step": 21125
    },
    {
      "entropy": 5.38690938949585,
      "epoch": 2.097062326319968,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004562803603102825,
      "loss": 5.0447,
      "mean_token_accuracy": 0.1663181334733963,
      "num_tokens": 46290226.0,
      "step": 21130
    },
    {
      "entropy": 5.439906215667724,
      "epoch": 2.0975585549821356,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045625937068519075,
      "loss": 5.0636,
      "mean_token_accuracy": 0.1747654989361763,
      "num_tokens": 46300388.0,
      "step": 21135
    },
    {
      "entropy": 5.390469646453857,
      "epoch": 2.098054783644303,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004562383765652571,
      "loss": 5.1228,
      "mean_token_accuracy": 0.16994300484657288,
      "num_tokens": 46312214.0,
      "step": 21140
    },
    {
      "entropy": 5.348788690567017,
      "epoch": 2.098551012306471,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004562173779510023,
      "loss": 5.1366,
      "mean_token_accuracy": 0.1650004044175148,
      "num_tokens": 46322728.0,
      "step": 21145
    },
    {
      "entropy": 5.358194351196289,
      "epoch": 2.0990472409686385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000456196374842947,
      "loss": 5.0317,
      "mean_token_accuracy": 0.17852895110845565,
      "num_tokens": 46334169.0,
      "step": 21150
    },
    {
      "entropy": 5.466904020309448,
      "epoch": 2.099543469630806,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045617536724161203,
      "loss": 5.0117,
      "mean_token_accuracy": 0.1685807392001152,
      "num_tokens": 46344601.0,
      "step": 21155
    },
    {
      "entropy": 5.383177423477173,
      "epoch": 2.1000396982929734,
      "grad_norm": 1.125,
      "learning_rate": 0.00045615435514751834,
      "loss": 5.1093,
      "mean_token_accuracy": 0.16585493534803392,
      "num_tokens": 46355633.0,
      "step": 21160
    },
    {
      "entropy": 5.365744256973267,
      "epoch": 2.100535926955141,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000456133338561187,
      "loss": 5.136,
      "mean_token_accuracy": 0.16406747996807097,
      "num_tokens": 46365981.0,
      "step": 21165
    },
    {
      "entropy": 5.412968730926513,
      "epoch": 2.1010321556173084,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004561123174831391,
      "loss": 5.1097,
      "mean_token_accuracy": 0.1749818041920662,
      "num_tokens": 46376664.0,
      "step": 21170
    },
    {
      "entropy": 5.378501224517822,
      "epoch": 2.101528384279476,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004560912919138962,
      "loss": 5.024,
      "mean_token_accuracy": 0.17553543001413346,
      "num_tokens": 46387918.0,
      "step": 21175
    },
    {
      "entropy": 5.4328282356262205,
      "epoch": 2.1020246129416433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045607026185397924,
      "loss": 5.0641,
      "mean_token_accuracy": 0.1731535464525223,
      "num_tokens": 46398660.0,
      "step": 21180
    },
    {
      "entropy": 5.343570327758789,
      "epoch": 2.102520841603811,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004560492273039101,
      "loss": 5.0125,
      "mean_token_accuracy": 0.17687406092882157,
      "num_tokens": 46409365.0,
      "step": 21185
    },
    {
      "entropy": 5.338657093048096,
      "epoch": 2.1030170702659787,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045602818826421024,
      "loss": 5.0046,
      "mean_token_accuracy": 0.17878929227590562,
      "num_tokens": 46419457.0,
      "step": 21190
    },
    {
      "entropy": 5.386030912399292,
      "epoch": 2.103513298928146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004560071447354013,
      "loss": 5.0977,
      "mean_token_accuracy": 0.16960126161575317,
      "num_tokens": 46430189.0,
      "step": 21195
    },
    {
      "entropy": 5.306434106826782,
      "epoch": 2.1040095275903137,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004559860967180053,
      "loss": 5.0481,
      "mean_token_accuracy": 0.17510274201631545,
      "num_tokens": 46440555.0,
      "step": 21200
    },
    {
      "entropy": 5.44195032119751,
      "epoch": 2.104505756252481,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004559650442125441,
      "loss": 5.0917,
      "mean_token_accuracy": 0.16554640829563141,
      "num_tokens": 46451582.0,
      "step": 21205
    },
    {
      "entropy": 5.494320058822632,
      "epoch": 2.1050019849146486,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004559439872195397,
      "loss": 5.097,
      "mean_token_accuracy": 0.1722371146082878,
      "num_tokens": 46463614.0,
      "step": 21210
    },
    {
      "entropy": 5.386815690994263,
      "epoch": 2.105498213576816,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004559229257395144,
      "loss": 5.1069,
      "mean_token_accuracy": 0.16551014482975007,
      "num_tokens": 46474279.0,
      "step": 21215
    },
    {
      "entropy": 5.428035402297974,
      "epoch": 2.1059944422389836,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045590185977299025,
      "loss": 5.1988,
      "mean_token_accuracy": 0.166376930475235,
      "num_tokens": 46485531.0,
      "step": 21220
    },
    {
      "entropy": 5.374498414993286,
      "epoch": 2.106490670901151,
      "grad_norm": 1.0,
      "learning_rate": 0.0004558807893204899,
      "loss": 5.0085,
      "mean_token_accuracy": 0.17767874747514725,
      "num_tokens": 46496197.0,
      "step": 21225
    },
    {
      "entropy": 5.389449214935302,
      "epoch": 2.106986899563319,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004558597143825357,
      "loss": 5.0931,
      "mean_token_accuracy": 0.16941645741462708,
      "num_tokens": 46506759.0,
      "step": 21230
    },
    {
      "entropy": 5.394826221466064,
      "epoch": 2.1074831282254864,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004558386349596503,
      "loss": 5.0812,
      "mean_token_accuracy": 0.17250393629074096,
      "num_tokens": 46518956.0,
      "step": 21235
    },
    {
      "entropy": 5.484100294113159,
      "epoch": 2.107979356887654,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004558175510523564,
      "loss": 5.1613,
      "mean_token_accuracy": 0.17146418988704681,
      "num_tokens": 46530109.0,
      "step": 21240
    },
    {
      "entropy": 5.388852071762085,
      "epoch": 2.1084755855498214,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045579646266117696,
      "loss": 5.0219,
      "mean_token_accuracy": 0.17176612764596938,
      "num_tokens": 46542206.0,
      "step": 21245
    },
    {
      "entropy": 5.432830572128296,
      "epoch": 2.108971814211989,
      "grad_norm": 1.25,
      "learning_rate": 0.00045577536978663475,
      "loss": 5.1951,
      "mean_token_accuracy": 0.17071758061647416,
      "num_tokens": 46554143.0,
      "step": 21250
    },
    {
      "entropy": 5.337934780120849,
      "epoch": 2.1094680428741563,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004557542724292529,
      "loss": 5.0038,
      "mean_token_accuracy": 0.17401469349861146,
      "num_tokens": 46565697.0,
      "step": 21255
    },
    {
      "entropy": 5.404529571533203,
      "epoch": 2.109964271536324,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004557331705895545,
      "loss": 5.0662,
      "mean_token_accuracy": 0.17238716781139374,
      "num_tokens": 46576358.0,
      "step": 21260
    },
    {
      "entropy": 5.42969012260437,
      "epoch": 2.1104605001984913,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004557120642680631,
      "loss": 5.1111,
      "mean_token_accuracy": 0.1651165470480919,
      "num_tokens": 46589422.0,
      "step": 21265
    },
    {
      "entropy": 5.371573209762573,
      "epoch": 2.1109567288606588,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004556909534653018,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17668355405330657,
      "num_tokens": 46599623.0,
      "step": 21270
    },
    {
      "entropy": 5.347667074203491,
      "epoch": 2.1114529575228267,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004556698381817942,
      "loss": 5.0103,
      "mean_token_accuracy": 0.17923943251371383,
      "num_tokens": 46609321.0,
      "step": 21275
    },
    {
      "entropy": 5.296820497512817,
      "epoch": 2.111949186184994,
      "grad_norm": 1.015625,
      "learning_rate": 0.000455648718418064,
      "loss": 5.0911,
      "mean_token_accuracy": 0.17002615779638292,
      "num_tokens": 46620265.0,
      "step": 21280
    },
    {
      "entropy": 5.400065803527832,
      "epoch": 2.1124454148471616,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004556275941746348,
      "loss": 5.1097,
      "mean_token_accuracy": 0.17392515689134597,
      "num_tokens": 46630776.0,
      "step": 21285
    },
    {
      "entropy": 5.336551284790039,
      "epoch": 2.112941643509329,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045560646545203044,
      "loss": 5.0092,
      "mean_token_accuracy": 0.17930993735790252,
      "num_tokens": 46641382.0,
      "step": 21290
    },
    {
      "entropy": 5.328059911727905,
      "epoch": 2.1134378721714966,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045558533225077496,
      "loss": 4.9959,
      "mean_token_accuracy": 0.1776877298951149,
      "num_tokens": 46653589.0,
      "step": 21295
    },
    {
      "entropy": 5.3539965629577635,
      "epoch": 2.113934100833664,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045556419457139235,
      "loss": 5.0461,
      "mean_token_accuracy": 0.17149013876914979,
      "num_tokens": 46664595.0,
      "step": 21300
    },
    {
      "entropy": 5.420930814743042,
      "epoch": 2.1144303294958315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045554305241440684,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17100189626216888,
      "num_tokens": 46675453.0,
      "step": 21305
    },
    {
      "entropy": 5.3386719703674315,
      "epoch": 2.114926558157999,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004555219057803426,
      "loss": 4.9804,
      "mean_token_accuracy": 0.17547924667596818,
      "num_tokens": 46685668.0,
      "step": 21310
    },
    {
      "entropy": 5.372794342041016,
      "epoch": 2.115422786820167,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004555007546697242,
      "loss": 5.099,
      "mean_token_accuracy": 0.16909148246049882,
      "num_tokens": 46696144.0,
      "step": 21315
    },
    {
      "entropy": 5.388325595855713,
      "epoch": 2.1159190154823344,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045547959908307585,
      "loss": 5.0656,
      "mean_token_accuracy": 0.16820987612009047,
      "num_tokens": 46707892.0,
      "step": 21320
    },
    {
      "entropy": 5.410978984832764,
      "epoch": 2.116415244144502,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004554584390209225,
      "loss": 5.0741,
      "mean_token_accuracy": 0.17511778473854064,
      "num_tokens": 46721099.0,
      "step": 21325
    },
    {
      "entropy": 5.408205175399781,
      "epoch": 2.1169114728066694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004554372744837886,
      "loss": 5.106,
      "mean_token_accuracy": 0.16851598769426346,
      "num_tokens": 46732974.0,
      "step": 21330
    },
    {
      "entropy": 5.386367607116699,
      "epoch": 2.117407701468837,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004554161054721992,
      "loss": 5.1487,
      "mean_token_accuracy": 0.17031199783086776,
      "num_tokens": 46743148.0,
      "step": 21335
    },
    {
      "entropy": 5.360859107971192,
      "epoch": 2.1179039301310043,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004553949319866791,
      "loss": 5.0265,
      "mean_token_accuracy": 0.17943513691425322,
      "num_tokens": 46754214.0,
      "step": 21340
    },
    {
      "entropy": 5.396155548095703,
      "epoch": 2.118400158793172,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045537375402775334,
      "loss": 5.0712,
      "mean_token_accuracy": 0.1717638850212097,
      "num_tokens": 46764112.0,
      "step": 21345
    },
    {
      "entropy": 5.314063453674317,
      "epoch": 2.1188963874553393,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045535257159594724,
      "loss": 5.0167,
      "mean_token_accuracy": 0.17950788736343384,
      "num_tokens": 46775202.0,
      "step": 21350
    },
    {
      "entropy": 5.486456871032715,
      "epoch": 2.1193926161175067,
      "grad_norm": 1.15625,
      "learning_rate": 0.000455331384691786,
      "loss": 5.0755,
      "mean_token_accuracy": 0.17167954593896867,
      "num_tokens": 46786557.0,
      "step": 21355
    },
    {
      "entropy": 5.3479883670806885,
      "epoch": 2.1198888447796747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004553101933157949,
      "loss": 5.0382,
      "mean_token_accuracy": 0.16374576836824417,
      "num_tokens": 46797082.0,
      "step": 21360
    },
    {
      "entropy": 5.33760404586792,
      "epoch": 2.120385073441842,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045528899746849965,
      "loss": 5.0627,
      "mean_token_accuracy": 0.17596685141324997,
      "num_tokens": 46806679.0,
      "step": 21365
    },
    {
      "entropy": 5.497006368637085,
      "epoch": 2.1208813021040096,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004552677971504257,
      "loss": 5.1983,
      "mean_token_accuracy": 0.16106846183538437,
      "num_tokens": 46818169.0,
      "step": 21370
    },
    {
      "entropy": 5.501601123809815,
      "epoch": 2.121377530766177,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004552465923620988,
      "loss": 5.224,
      "mean_token_accuracy": 0.16556275337934495,
      "num_tokens": 46829552.0,
      "step": 21375
    },
    {
      "entropy": 5.4649327278137205,
      "epoch": 2.1218737594283446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004552253831040448,
      "loss": 5.1417,
      "mean_token_accuracy": 0.17154500335454942,
      "num_tokens": 46840808.0,
      "step": 21380
    },
    {
      "entropy": 5.4244156837463375,
      "epoch": 2.122369988090512,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045520416937678973,
      "loss": 5.0817,
      "mean_token_accuracy": 0.16452176868915558,
      "num_tokens": 46852066.0,
      "step": 21385
    },
    {
      "entropy": 5.300200510025024,
      "epoch": 2.1228662167526795,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045518295118085955,
      "loss": 5.0199,
      "mean_token_accuracy": 0.17257951945066452,
      "num_tokens": 46863163.0,
      "step": 21390
    },
    {
      "entropy": 5.365149927139282,
      "epoch": 2.123362445414847,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045516172851678036,
      "loss": 5.0674,
      "mean_token_accuracy": 0.17179541438817977,
      "num_tokens": 46873246.0,
      "step": 21395
    },
    {
      "entropy": 5.361225986480713,
      "epoch": 2.123858674077015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045514050138507857,
      "loss": 5.088,
      "mean_token_accuracy": 0.17186149656772615,
      "num_tokens": 46883881.0,
      "step": 21400
    },
    {
      "entropy": 5.392772483825683,
      "epoch": 2.1243549027391824,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004551192697862805,
      "loss": 5.1097,
      "mean_token_accuracy": 0.1722362831234932,
      "num_tokens": 46894907.0,
      "step": 21405
    },
    {
      "entropy": 5.36645393371582,
      "epoch": 2.12485113140135,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004550980337209127,
      "loss": 5.0181,
      "mean_token_accuracy": 0.16860586553812026,
      "num_tokens": 46905220.0,
      "step": 21410
    },
    {
      "entropy": 5.340509939193725,
      "epoch": 2.1253473600635173,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045507679318950166,
      "loss": 5.0384,
      "mean_token_accuracy": 0.1768256515264511,
      "num_tokens": 46916328.0,
      "step": 21415
    },
    {
      "entropy": 5.553907346725464,
      "epoch": 2.125843588725685,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004550555481925743,
      "loss": 5.2469,
      "mean_token_accuracy": 0.1577999398112297,
      "num_tokens": 46927450.0,
      "step": 21420
    },
    {
      "entropy": 5.336559677124024,
      "epoch": 2.1263398173878523,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045503429873065723,
      "loss": 4.9765,
      "mean_token_accuracy": 0.17725519984960555,
      "num_tokens": 46936758.0,
      "step": 21425
    },
    {
      "entropy": 5.261868190765381,
      "epoch": 2.1268360460500197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004550130448042775,
      "loss": 4.9291,
      "mean_token_accuracy": 0.17975317388772966,
      "num_tokens": 46945776.0,
      "step": 21430
    },
    {
      "entropy": 5.322411775588989,
      "epoch": 2.127332274712187,
      "grad_norm": 1.125,
      "learning_rate": 0.0004549917864139622,
      "loss": 5.0558,
      "mean_token_accuracy": 0.17465361207723618,
      "num_tokens": 46955621.0,
      "step": 21435
    },
    {
      "entropy": 5.508475255966187,
      "epoch": 2.1278285033743547,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004549705235602384,
      "loss": 5.1669,
      "mean_token_accuracy": 0.17429178804159165,
      "num_tokens": 46966890.0,
      "step": 21440
    },
    {
      "entropy": 5.434486818313599,
      "epoch": 2.1283247320365226,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004549492562436334,
      "loss": 5.1705,
      "mean_token_accuracy": 0.1665997475385666,
      "num_tokens": 46978380.0,
      "step": 21445
    },
    {
      "entropy": 5.365892028808593,
      "epoch": 2.12882096069869,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045492798446467465,
      "loss": 4.9936,
      "mean_token_accuracy": 0.18029729127883912,
      "num_tokens": 46988286.0,
      "step": 21450
    },
    {
      "entropy": 5.337842750549316,
      "epoch": 2.1293171893608576,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045490670822388957,
      "loss": 5.008,
      "mean_token_accuracy": 0.17681744545698166,
      "num_tokens": 46997795.0,
      "step": 21455
    },
    {
      "entropy": 5.268886756896973,
      "epoch": 2.129813418023025,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004548854275218059,
      "loss": 4.8972,
      "mean_token_accuracy": 0.1867232456803322,
      "num_tokens": 47007970.0,
      "step": 21460
    },
    {
      "entropy": 5.305668115615845,
      "epoch": 2.1303096466851925,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045486414235895115,
      "loss": 5.0239,
      "mean_token_accuracy": 0.17495810687541963,
      "num_tokens": 47019548.0,
      "step": 21465
    },
    {
      "entropy": 5.391514253616333,
      "epoch": 2.13080587534736,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004548428527358532,
      "loss": 5.1284,
      "mean_token_accuracy": 0.17019821405410768,
      "num_tokens": 47029568.0,
      "step": 21470
    },
    {
      "entropy": 5.462993335723877,
      "epoch": 2.1313021040095275,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004548215586530401,
      "loss": 5.1434,
      "mean_token_accuracy": 0.16677082926034928,
      "num_tokens": 47042048.0,
      "step": 21475
    },
    {
      "entropy": 5.36169114112854,
      "epoch": 2.131798332671695,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045480026011103975,
      "loss": 4.9801,
      "mean_token_accuracy": 0.17140425741672516,
      "num_tokens": 47052099.0,
      "step": 21480
    },
    {
      "entropy": 5.360023069381714,
      "epoch": 2.132294561333863,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045477895711038045,
      "loss": 5.0352,
      "mean_token_accuracy": 0.17295927107334136,
      "num_tokens": 47063310.0,
      "step": 21485
    },
    {
      "entropy": 5.378916692733765,
      "epoch": 2.1327907899960303,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004547576496515904,
      "loss": 5.052,
      "mean_token_accuracy": 0.17049139440059663,
      "num_tokens": 47073545.0,
      "step": 21490
    },
    {
      "entropy": 5.359961271286011,
      "epoch": 2.133287018658198,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045473633773519795,
      "loss": 5.1094,
      "mean_token_accuracy": 0.17200727313756942,
      "num_tokens": 47085456.0,
      "step": 21495
    },
    {
      "entropy": 5.39084644317627,
      "epoch": 2.1337832473203653,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045471502136173173,
      "loss": 5.0957,
      "mean_token_accuracy": 0.17270459830760956,
      "num_tokens": 47096590.0,
      "step": 21500
    },
    {
      "entropy": 5.471798229217529,
      "epoch": 2.1342794759825328,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004546937005317201,
      "loss": 5.0311,
      "mean_token_accuracy": 0.1763642430305481,
      "num_tokens": 47107876.0,
      "step": 21505
    },
    {
      "entropy": 5.358714199066162,
      "epoch": 2.1347757046447002,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045467237524569193,
      "loss": 5.0692,
      "mean_token_accuracy": 0.17257637530565262,
      "num_tokens": 47117662.0,
      "step": 21510
    },
    {
      "entropy": 5.330829095840454,
      "epoch": 2.1352719333068677,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045465104550417604,
      "loss": 5.0824,
      "mean_token_accuracy": 0.17052652239799498,
      "num_tokens": 47129283.0,
      "step": 21515
    },
    {
      "entropy": 5.463081693649292,
      "epoch": 2.135768161969035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045462971130770123,
      "loss": 5.1678,
      "mean_token_accuracy": 0.16951697766780854,
      "num_tokens": 47140824.0,
      "step": 21520
    },
    {
      "entropy": 5.358862733840942,
      "epoch": 2.1362643906312027,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004546083726567968,
      "loss": 5.0272,
      "mean_token_accuracy": 0.17507528364658356,
      "num_tokens": 47151733.0,
      "step": 21525
    },
    {
      "entropy": 5.380132341384888,
      "epoch": 2.1367606192933706,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004545870295519917,
      "loss": 5.1448,
      "mean_token_accuracy": 0.16408263444900512,
      "num_tokens": 47162846.0,
      "step": 21530
    },
    {
      "entropy": 5.426844120025635,
      "epoch": 2.137256847955538,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045456568199381524,
      "loss": 5.0373,
      "mean_token_accuracy": 0.17379597574472427,
      "num_tokens": 47174969.0,
      "step": 21535
    },
    {
      "entropy": 5.452709436416626,
      "epoch": 2.1377530766177055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045454432998279676,
      "loss": 5.1434,
      "mean_token_accuracy": 0.1693483203649521,
      "num_tokens": 47186903.0,
      "step": 21540
    },
    {
      "entropy": 5.37514100074768,
      "epoch": 2.138249305279873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004545229735194658,
      "loss": 5.0919,
      "mean_token_accuracy": 0.1717928320169449,
      "num_tokens": 47197716.0,
      "step": 21545
    },
    {
      "entropy": 5.390336370468139,
      "epoch": 2.1387455339420405,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045450161260435196,
      "loss": 5.0297,
      "mean_token_accuracy": 0.17669080942869186,
      "num_tokens": 47208623.0,
      "step": 21550
    },
    {
      "entropy": 5.3517192840576175,
      "epoch": 2.139241762604208,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004544802472379849,
      "loss": 5.0406,
      "mean_token_accuracy": 0.17329699099063872,
      "num_tokens": 47218248.0,
      "step": 21555
    },
    {
      "entropy": 5.396907949447632,
      "epoch": 2.1397379912663754,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004544588774208945,
      "loss": 5.1685,
      "mean_token_accuracy": 0.166157004237175,
      "num_tokens": 47232028.0,
      "step": 21560
    },
    {
      "entropy": 5.441629123687744,
      "epoch": 2.140234219928543,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045443750315361055,
      "loss": 5.0425,
      "mean_token_accuracy": 0.17597909271717072,
      "num_tokens": 47241976.0,
      "step": 21565
    },
    {
      "entropy": 5.382098007202148,
      "epoch": 2.140730448590711,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004544161244366632,
      "loss": 5.034,
      "mean_token_accuracy": 0.17369380891323088,
      "num_tokens": 47253228.0,
      "step": 21570
    },
    {
      "entropy": 5.359360790252685,
      "epoch": 2.1412266772528783,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004543947412705826,
      "loss": 5.0448,
      "mean_token_accuracy": 0.17145697325468062,
      "num_tokens": 47265210.0,
      "step": 21575
    },
    {
      "entropy": 5.326690530776977,
      "epoch": 2.1417229059150458,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045437335365589905,
      "loss": 5.0705,
      "mean_token_accuracy": 0.17061249911785126,
      "num_tokens": 47275723.0,
      "step": 21580
    },
    {
      "entropy": 5.415219068527222,
      "epoch": 2.1422191345772132,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045435196159314276,
      "loss": 5.1271,
      "mean_token_accuracy": 0.16825629472732545,
      "num_tokens": 47286469.0,
      "step": 21585
    },
    {
      "entropy": 5.468785619735717,
      "epoch": 2.1427153632393807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045433056508284425,
      "loss": 5.126,
      "mean_token_accuracy": 0.17347800582647324,
      "num_tokens": 47296461.0,
      "step": 21590
    },
    {
      "entropy": 5.392240810394287,
      "epoch": 2.143211591901548,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004543091641255342,
      "loss": 5.0174,
      "mean_token_accuracy": 0.17696133255958557,
      "num_tokens": 47307217.0,
      "step": 21595
    },
    {
      "entropy": 5.390687942504883,
      "epoch": 2.1437078205637157,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004542877587217432,
      "loss": 5.1533,
      "mean_token_accuracy": 0.16337285190820694,
      "num_tokens": 47317041.0,
      "step": 21600
    },
    {
      "entropy": 5.383497095108032,
      "epoch": 2.144204049225883,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045426634887200224,
      "loss": 5.0789,
      "mean_token_accuracy": 0.17481568604707717,
      "num_tokens": 47328185.0,
      "step": 21605
    },
    {
      "entropy": 5.415882587432861,
      "epoch": 2.1447002778880506,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000454244934576842,
      "loss": 5.1116,
      "mean_token_accuracy": 0.16674584448337554,
      "num_tokens": 47339984.0,
      "step": 21610
    },
    {
      "entropy": 5.383595037460327,
      "epoch": 2.1451965065502185,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045422351583679365,
      "loss": 5.0456,
      "mean_token_accuracy": 0.17692509293556213,
      "num_tokens": 47350911.0,
      "step": 21615
    },
    {
      "entropy": 5.374559354782105,
      "epoch": 2.145692735212386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004542020926523883,
      "loss": 5.0489,
      "mean_token_accuracy": 0.1723474830389023,
      "num_tokens": 47361703.0,
      "step": 21620
    },
    {
      "entropy": 5.3890604972839355,
      "epoch": 2.1461889638745535,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004541806650241572,
      "loss": 5.0328,
      "mean_token_accuracy": 0.17075188159942628,
      "num_tokens": 47372878.0,
      "step": 21625
    },
    {
      "entropy": 5.360749101638794,
      "epoch": 2.146685192536721,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004541592329526317,
      "loss": 5.0744,
      "mean_token_accuracy": 0.16518908590078354,
      "num_tokens": 47383056.0,
      "step": 21630
    },
    {
      "entropy": 5.369113254547119,
      "epoch": 2.1471814211988884,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004541377964383434,
      "loss": 5.0806,
      "mean_token_accuracy": 0.17841193377971648,
      "num_tokens": 47394160.0,
      "step": 21635
    },
    {
      "entropy": 5.410951805114746,
      "epoch": 2.147677649861056,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004541163554818237,
      "loss": 5.0501,
      "mean_token_accuracy": 0.16895425617694854,
      "num_tokens": 47404861.0,
      "step": 21640
    },
    {
      "entropy": 5.40322060585022,
      "epoch": 2.1481738785232234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045409491008360423,
      "loss": 5.1086,
      "mean_token_accuracy": 0.1734311982989311,
      "num_tokens": 47415236.0,
      "step": 21645
    },
    {
      "entropy": 5.391456937789917,
      "epoch": 2.148670107185391,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004540734602442169,
      "loss": 5.0758,
      "mean_token_accuracy": 0.17060115337371826,
      "num_tokens": 47425789.0,
      "step": 21650
    },
    {
      "entropy": 5.335027885437012,
      "epoch": 2.149166335847559,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004540520059641937,
      "loss": 4.9256,
      "mean_token_accuracy": 0.17879532426595687,
      "num_tokens": 47436236.0,
      "step": 21655
    },
    {
      "entropy": 5.302869462966919,
      "epoch": 2.1496625645097263,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004540305472440666,
      "loss": 4.9627,
      "mean_token_accuracy": 0.17825574576854705,
      "num_tokens": 47447041.0,
      "step": 21660
    },
    {
      "entropy": 5.386822748184204,
      "epoch": 2.1501587931718937,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004540090840843677,
      "loss": 5.1098,
      "mean_token_accuracy": 0.16645592600107192,
      "num_tokens": 47458001.0,
      "step": 21665
    },
    {
      "entropy": 5.314797353744507,
      "epoch": 2.150655021834061,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004539876164856292,
      "loss": 4.9841,
      "mean_token_accuracy": 0.17659251838922502,
      "num_tokens": 47468621.0,
      "step": 21670
    },
    {
      "entropy": 5.356137228012085,
      "epoch": 2.1511512504962287,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004539661444483835,
      "loss": 5.11,
      "mean_token_accuracy": 0.17485348433256148,
      "num_tokens": 47478957.0,
      "step": 21675
    },
    {
      "entropy": 5.350113344192505,
      "epoch": 2.151647479158396,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004539446679731631,
      "loss": 5.0374,
      "mean_token_accuracy": 0.17535323947668074,
      "num_tokens": 47489640.0,
      "step": 21680
    },
    {
      "entropy": 5.443061494827271,
      "epoch": 2.1521437078205636,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004539231870605005,
      "loss": 5.1891,
      "mean_token_accuracy": 0.16707523316144943,
      "num_tokens": 47501581.0,
      "step": 21685
    },
    {
      "entropy": 5.440029191970825,
      "epoch": 2.152639936482731,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045390170171092845,
      "loss": 5.1158,
      "mean_token_accuracy": 0.17410192936658858,
      "num_tokens": 47513605.0,
      "step": 21690
    },
    {
      "entropy": 5.441654205322266,
      "epoch": 2.1531361651448986,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045388021192497956,
      "loss": 5.0957,
      "mean_token_accuracy": 0.17239117622375488,
      "num_tokens": 47523784.0,
      "step": 21695
    },
    {
      "entropy": 5.324306583404541,
      "epoch": 2.1536323938070665,
      "grad_norm": 1.046875,
      "learning_rate": 0.000453858717703187,
      "loss": 4.9896,
      "mean_token_accuracy": 0.17158765196800232,
      "num_tokens": 47534643.0,
      "step": 21700
    },
    {
      "entropy": 5.331645965576172,
      "epoch": 2.154128622469234,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004538372190460836,
      "loss": 5.1309,
      "mean_token_accuracy": 0.1696048706769943,
      "num_tokens": 47545920.0,
      "step": 21705
    },
    {
      "entropy": 5.373568058013916,
      "epoch": 2.1546248511314015,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004538157159542025,
      "loss": 5.0371,
      "mean_token_accuracy": 0.17201700657606125,
      "num_tokens": 47557024.0,
      "step": 21710
    },
    {
      "entropy": 5.418411207199097,
      "epoch": 2.155121079793569,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004537942084280771,
      "loss": 5.0091,
      "mean_token_accuracy": 0.17568860799074174,
      "num_tokens": 47567531.0,
      "step": 21715
    },
    {
      "entropy": 5.364583444595337,
      "epoch": 2.1556173084557364,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045377269646824036,
      "loss": 5.1479,
      "mean_token_accuracy": 0.16772240102291108,
      "num_tokens": 47578642.0,
      "step": 21720
    },
    {
      "entropy": 5.359743309020996,
      "epoch": 2.156113537117904,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045375118007522607,
      "loss": 5.189,
      "mean_token_accuracy": 0.16809726655483245,
      "num_tokens": 47589346.0,
      "step": 21725
    },
    {
      "entropy": 5.353248882293701,
      "epoch": 2.1566097657800714,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004537296592495678,
      "loss": 4.999,
      "mean_token_accuracy": 0.17601928412914275,
      "num_tokens": 47599689.0,
      "step": 21730
    },
    {
      "entropy": 5.4632164478302006,
      "epoch": 2.157105994442239,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000453708133991799,
      "loss": 5.0879,
      "mean_token_accuracy": 0.17120047360658647,
      "num_tokens": 47609214.0,
      "step": 21735
    },
    {
      "entropy": 5.400595855712891,
      "epoch": 2.1576022231044063,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045368660430245364,
      "loss": 5.0467,
      "mean_token_accuracy": 0.172533218562603,
      "num_tokens": 47620119.0,
      "step": 21740
    },
    {
      "entropy": 5.356599760055542,
      "epoch": 2.158098451766574,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004536650701820654,
      "loss": 5.0079,
      "mean_token_accuracy": 0.17451198697090148,
      "num_tokens": 47630085.0,
      "step": 21745
    },
    {
      "entropy": 5.3742146492004395,
      "epoch": 2.1585946804287417,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004536435316311685,
      "loss": 4.9985,
      "mean_token_accuracy": 0.17526301443576814,
      "num_tokens": 47641457.0,
      "step": 21750
    },
    {
      "entropy": 5.377593088150024,
      "epoch": 2.159090909090909,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045362198865029696,
      "loss": 5.0258,
      "mean_token_accuracy": 0.17397779375314712,
      "num_tokens": 47652269.0,
      "step": 21755
    },
    {
      "entropy": 5.371991300582886,
      "epoch": 2.1595871377530766,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045360044123998495,
      "loss": 5.072,
      "mean_token_accuracy": 0.17029978334903717,
      "num_tokens": 47664601.0,
      "step": 21760
    },
    {
      "entropy": 5.378527545928955,
      "epoch": 2.160083366415244,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045357888940076685,
      "loss": 5.1847,
      "mean_token_accuracy": 0.16604819744825364,
      "num_tokens": 47676187.0,
      "step": 21765
    },
    {
      "entropy": 5.3374267578125,
      "epoch": 2.1605795950774116,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004535573331331771,
      "loss": 4.9586,
      "mean_token_accuracy": 0.18018792271614076,
      "num_tokens": 47687057.0,
      "step": 21770
    },
    {
      "entropy": 5.3551350116729735,
      "epoch": 2.161075823739579,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045353577243775035,
      "loss": 5.0309,
      "mean_token_accuracy": 0.17421613037586212,
      "num_tokens": 47699074.0,
      "step": 21775
    },
    {
      "entropy": 5.3517577171325685,
      "epoch": 2.1615720524017465,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004535142073150211,
      "loss": 5.0256,
      "mean_token_accuracy": 0.172435861825943,
      "num_tokens": 47710400.0,
      "step": 21780
    },
    {
      "entropy": 5.3282982349395756,
      "epoch": 2.162068281063914,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004534926377655241,
      "loss": 4.9097,
      "mean_token_accuracy": 0.18938955217599868,
      "num_tokens": 47720900.0,
      "step": 21785
    },
    {
      "entropy": 5.293919038772583,
      "epoch": 2.162564509726082,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045347106378979435,
      "loss": 4.9232,
      "mean_token_accuracy": 0.17968342155218125,
      "num_tokens": 47730702.0,
      "step": 21790
    },
    {
      "entropy": 5.320708608627319,
      "epoch": 2.1630607383882494,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045344948538836674,
      "loss": 5.0152,
      "mean_token_accuracy": 0.17401691526174545,
      "num_tokens": 47740451.0,
      "step": 21795
    },
    {
      "entropy": 5.394125318527221,
      "epoch": 2.163556967050417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045342790256177646,
      "loss": 5.0915,
      "mean_token_accuracy": 0.16973927170038222,
      "num_tokens": 47752285.0,
      "step": 21800
    },
    {
      "entropy": 5.427969646453858,
      "epoch": 2.1640531957125844,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004534063153105587,
      "loss": 5.0827,
      "mean_token_accuracy": 0.1731274515390396,
      "num_tokens": 47761895.0,
      "step": 21805
    },
    {
      "entropy": 5.3733477115631105,
      "epoch": 2.164549424374752,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004533847236352487,
      "loss": 4.9841,
      "mean_token_accuracy": 0.17839306443929673,
      "num_tokens": 47771718.0,
      "step": 21810
    },
    {
      "entropy": 5.268028116226196,
      "epoch": 2.1650456530369193,
      "grad_norm": 1.03125,
      "learning_rate": 0.000453363127536382,
      "loss": 5.0126,
      "mean_token_accuracy": 0.17598240226507186,
      "num_tokens": 47782183.0,
      "step": 21815
    },
    {
      "entropy": 5.300457000732422,
      "epoch": 2.165541881699087,
      "grad_norm": 1.125,
      "learning_rate": 0.00045334152701449407,
      "loss": 4.9649,
      "mean_token_accuracy": 0.18025597780942917,
      "num_tokens": 47792190.0,
      "step": 21820
    },
    {
      "entropy": 5.458163070678711,
      "epoch": 2.1660381103612543,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004533199220701205,
      "loss": 5.163,
      "mean_token_accuracy": 0.16786447763442994,
      "num_tokens": 47803245.0,
      "step": 21825
    },
    {
      "entropy": 5.3874945640563965,
      "epoch": 2.166534339023422,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045329831270379714,
      "loss": 5.1156,
      "mean_token_accuracy": 0.16348602175712584,
      "num_tokens": 47814605.0,
      "step": 21830
    },
    {
      "entropy": 5.3859968185424805,
      "epoch": 2.1670305676855897,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045327669891605975,
      "loss": 5.0982,
      "mean_token_accuracy": 0.17172147035598756,
      "num_tokens": 47825945.0,
      "step": 21835
    },
    {
      "entropy": 5.382435750961304,
      "epoch": 2.167526796347757,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004532550807074445,
      "loss": 5.084,
      "mean_token_accuracy": 0.16732012778520583,
      "num_tokens": 47838119.0,
      "step": 21840
    },
    {
      "entropy": 5.411733865737915,
      "epoch": 2.1680230250099246,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004532334580784873,
      "loss": 5.077,
      "mean_token_accuracy": 0.16880767345428466,
      "num_tokens": 47849560.0,
      "step": 21845
    },
    {
      "entropy": 5.428269910812378,
      "epoch": 2.168519253672092,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004532118310297244,
      "loss": 5.0868,
      "mean_token_accuracy": 0.16152296215295792,
      "num_tokens": 47860036.0,
      "step": 21850
    },
    {
      "entropy": 5.295793867111206,
      "epoch": 2.1690154823342596,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004531901995616921,
      "loss": 4.9692,
      "mean_token_accuracy": 0.17535372972488403,
      "num_tokens": 47870826.0,
      "step": 21855
    },
    {
      "entropy": 5.3116247177124025,
      "epoch": 2.169511710996427,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004531685636749268,
      "loss": 5.0866,
      "mean_token_accuracy": 0.1644235610961914,
      "num_tokens": 47881415.0,
      "step": 21860
    },
    {
      "entropy": 5.4708634376525875,
      "epoch": 2.1700079396585945,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004531469233699651,
      "loss": 5.0996,
      "mean_token_accuracy": 0.16952424198389054,
      "num_tokens": 47893343.0,
      "step": 21865
    },
    {
      "entropy": 5.356075382232666,
      "epoch": 2.170504168320762,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004531252786473436,
      "loss": 5.0714,
      "mean_token_accuracy": 0.16835957467556,
      "num_tokens": 47904871.0,
      "step": 21870
    },
    {
      "entropy": 5.329317188262939,
      "epoch": 2.17100039698293,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004531036295075989,
      "loss": 4.9941,
      "mean_token_accuracy": 0.1753849133849144,
      "num_tokens": 47915367.0,
      "step": 21875
    },
    {
      "entropy": 5.375592470169067,
      "epoch": 2.1714966256450974,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045308197595126797,
      "loss": 5.0211,
      "mean_token_accuracy": 0.17160156965255738,
      "num_tokens": 47925987.0,
      "step": 21880
    },
    {
      "entropy": 5.396091938018799,
      "epoch": 2.171992854307265,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004530603179788879,
      "loss": 5.1101,
      "mean_token_accuracy": 0.17315345108509064,
      "num_tokens": 47937688.0,
      "step": 21885
    },
    {
      "entropy": 5.354980993270874,
      "epoch": 2.1724890829694323,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004530386555909955,
      "loss": 5.0557,
      "mean_token_accuracy": 0.17894713431596757,
      "num_tokens": 47948550.0,
      "step": 21890
    },
    {
      "entropy": 5.362069034576416,
      "epoch": 2.1729853116316,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045301698878812813,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1780369460582733,
      "num_tokens": 47957982.0,
      "step": 21895
    },
    {
      "entropy": 5.416823959350586,
      "epoch": 2.1734815402937673,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004529953175708231,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17285542339086532,
      "num_tokens": 47969562.0,
      "step": 21900
    },
    {
      "entropy": 5.3695355415344235,
      "epoch": 2.1739777689559348,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045297364193961763,
      "loss": 5.1046,
      "mean_token_accuracy": 0.16864823997020723,
      "num_tokens": 47981323.0,
      "step": 21905
    },
    {
      "entropy": 5.394772624969482,
      "epoch": 2.1744739976181022,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045295196189504934,
      "loss": 5.05,
      "mean_token_accuracy": 0.17215941995382308,
      "num_tokens": 47991454.0,
      "step": 21910
    },
    {
      "entropy": 5.346798896789551,
      "epoch": 2.17497022628027,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045293027743765595,
      "loss": 4.9382,
      "mean_token_accuracy": 0.1751525029540062,
      "num_tokens": 48001300.0,
      "step": 21915
    },
    {
      "entropy": 5.3895515441894535,
      "epoch": 2.1754664549424376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045290858856797503,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1750716134905815,
      "num_tokens": 48012608.0,
      "step": 21920
    },
    {
      "entropy": 5.374867868423462,
      "epoch": 2.175962683604605,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045288689528654444,
      "loss": 5.0459,
      "mean_token_accuracy": 0.16990364491939544,
      "num_tokens": 48022120.0,
      "step": 21925
    },
    {
      "entropy": 5.364972162246704,
      "epoch": 2.1764589122667726,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045286519759390217,
      "loss": 5.0454,
      "mean_token_accuracy": 0.168553227186203,
      "num_tokens": 48032629.0,
      "step": 21930
    },
    {
      "entropy": 5.33377776145935,
      "epoch": 2.17695514092894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004528434954905863,
      "loss": 4.9639,
      "mean_token_accuracy": 0.17669840455055236,
      "num_tokens": 48042934.0,
      "step": 21935
    },
    {
      "entropy": 5.399502563476562,
      "epoch": 2.1774513695911075,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004528217889771349,
      "loss": 5.0505,
      "mean_token_accuracy": 0.17463553100824356,
      "num_tokens": 48052825.0,
      "step": 21940
    },
    {
      "entropy": 5.4570698738098145,
      "epoch": 2.177947598253275,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045280007805408636,
      "loss": 5.1325,
      "mean_token_accuracy": 0.16696045696735382,
      "num_tokens": 48064565.0,
      "step": 21945
    },
    {
      "entropy": 5.450883054733277,
      "epoch": 2.1784438269154425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045277836272197895,
      "loss": 5.1049,
      "mean_token_accuracy": 0.1726071909070015,
      "num_tokens": 48076006.0,
      "step": 21950
    },
    {
      "entropy": 5.479358148574829,
      "epoch": 2.17894005557761,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045275664298135134,
      "loss": 5.212,
      "mean_token_accuracy": 0.16696474105119705,
      "num_tokens": 48087110.0,
      "step": 21955
    },
    {
      "entropy": 5.412522792816162,
      "epoch": 2.179436284239778,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004527349188327419,
      "loss": 5.0875,
      "mean_token_accuracy": 0.1767511934041977,
      "num_tokens": 48098894.0,
      "step": 21960
    },
    {
      "entropy": 5.408317565917969,
      "epoch": 2.1799325129019453,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004527131902766895,
      "loss": 5.0811,
      "mean_token_accuracy": 0.1710027977824211,
      "num_tokens": 48109729.0,
      "step": 21965
    },
    {
      "entropy": 5.440647029876709,
      "epoch": 2.180428741564113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004526914573137329,
      "loss": 5.0974,
      "mean_token_accuracy": 0.16632296442985534,
      "num_tokens": 48120345.0,
      "step": 21970
    },
    {
      "entropy": 5.367796087265015,
      "epoch": 2.1809249702262803,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045266971994441116,
      "loss": 5.0153,
      "mean_token_accuracy": 0.1788671314716339,
      "num_tokens": 48130185.0,
      "step": 21975
    },
    {
      "entropy": 5.321435976028442,
      "epoch": 2.1814211988884478,
      "grad_norm": 1.046875,
      "learning_rate": 0.000452647978169263,
      "loss": 4.9877,
      "mean_token_accuracy": 0.17499962449073792,
      "num_tokens": 48141020.0,
      "step": 21980
    },
    {
      "entropy": 5.360043478012085,
      "epoch": 2.1819174275506152,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045262623198882797,
      "loss": 5.0703,
      "mean_token_accuracy": 0.18006942123174668,
      "num_tokens": 48153361.0,
      "step": 21985
    },
    {
      "entropy": 5.350394153594971,
      "epoch": 2.1824136562127827,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004526044814036451,
      "loss": 5.0287,
      "mean_token_accuracy": 0.17279748916625975,
      "num_tokens": 48163604.0,
      "step": 21990
    },
    {
      "entropy": 5.371896028518677,
      "epoch": 2.18290988487495,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045258272641425377,
      "loss": 5.0609,
      "mean_token_accuracy": 0.16895462721586227,
      "num_tokens": 48174732.0,
      "step": 21995
    },
    {
      "entropy": 5.385177659988403,
      "epoch": 2.183406113537118,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004525609670211934,
      "loss": 5.0261,
      "mean_token_accuracy": 0.17496683299541474,
      "num_tokens": 48187082.0,
      "step": 22000
    },
    {
      "entropy": 5.363424301147461,
      "epoch": 2.1839023421992856,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045253920322500373,
      "loss": 5.0188,
      "mean_token_accuracy": 0.1770729273557663,
      "num_tokens": 48196547.0,
      "step": 22005
    },
    {
      "entropy": 5.352803564071655,
      "epoch": 2.184398570861453,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045251743502622445,
      "loss": 5.0886,
      "mean_token_accuracy": 0.1699337214231491,
      "num_tokens": 48207404.0,
      "step": 22010
    },
    {
      "entropy": 5.318226146697998,
      "epoch": 2.1848947995236205,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004524956624253952,
      "loss": 5.0598,
      "mean_token_accuracy": 0.17993370890617372,
      "num_tokens": 48217093.0,
      "step": 22015
    },
    {
      "entropy": 5.3861448764801025,
      "epoch": 2.185391028185788,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045247388542305595,
      "loss": 5.0883,
      "mean_token_accuracy": 0.17297727316617967,
      "num_tokens": 48228093.0,
      "step": 22020
    },
    {
      "entropy": 5.3346813201904295,
      "epoch": 2.1858872568479555,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004524521040197469,
      "loss": 5.0419,
      "mean_token_accuracy": 0.1782454952597618,
      "num_tokens": 48238834.0,
      "step": 22025
    },
    {
      "entropy": 5.4141909122467045,
      "epoch": 2.186383485510123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045243031821600794,
      "loss": 5.1047,
      "mean_token_accuracy": 0.18184837102890014,
      "num_tokens": 48250782.0,
      "step": 22030
    },
    {
      "entropy": 5.302735900878906,
      "epoch": 2.1868797141722904,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004524085280123795,
      "loss": 4.9679,
      "mean_token_accuracy": 0.17445237189531326,
      "num_tokens": 48261711.0,
      "step": 22035
    },
    {
      "entropy": 5.351752424240113,
      "epoch": 2.187375942834458,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004523867334094018,
      "loss": 5.0934,
      "mean_token_accuracy": 0.1699676737189293,
      "num_tokens": 48273505.0,
      "step": 22040
    },
    {
      "entropy": 5.331303358078003,
      "epoch": 2.187872171496626,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004523649344076154,
      "loss": 4.9812,
      "mean_token_accuracy": 0.1767857402563095,
      "num_tokens": 48284639.0,
      "step": 22045
    },
    {
      "entropy": 5.397255945205688,
      "epoch": 2.1883684001587933,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045234313100756077,
      "loss": 5.0934,
      "mean_token_accuracy": 0.17123149782419206,
      "num_tokens": 48296275.0,
      "step": 22050
    },
    {
      "entropy": 5.353832864761353,
      "epoch": 2.188864628820961,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045232132320977854,
      "loss": 4.9886,
      "mean_token_accuracy": 0.1776492029428482,
      "num_tokens": 48306859.0,
      "step": 22055
    },
    {
      "entropy": 5.396651887893677,
      "epoch": 2.1893608574831283,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004522995110148097,
      "loss": 5.1157,
      "mean_token_accuracy": 0.1721997156739235,
      "num_tokens": 48317561.0,
      "step": 22060
    },
    {
      "entropy": 5.3708892345428465,
      "epoch": 2.1898570861452957,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000452277694423195,
      "loss": 5.1338,
      "mean_token_accuracy": 0.16297013461589813,
      "num_tokens": 48327814.0,
      "step": 22065
    },
    {
      "entropy": 5.404459714889526,
      "epoch": 2.190353314807463,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004522558734354755,
      "loss": 5.0008,
      "mean_token_accuracy": 0.17576092034578322,
      "num_tokens": 48339658.0,
      "step": 22070
    },
    {
      "entropy": 5.3908905506134035,
      "epoch": 2.1908495434696307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045223404805219225,
      "loss": 5.1091,
      "mean_token_accuracy": 0.17149521559476852,
      "num_tokens": 48352115.0,
      "step": 22075
    },
    {
      "entropy": 5.412671709060669,
      "epoch": 2.191345772131798,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045221221827388655,
      "loss": 5.0547,
      "mean_token_accuracy": 0.17244788855314255,
      "num_tokens": 48362378.0,
      "step": 22080
    },
    {
      "entropy": 5.3583691120147705,
      "epoch": 2.191842000793966,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004521903841010997,
      "loss": 4.9781,
      "mean_token_accuracy": 0.1766326516866684,
      "num_tokens": 48373035.0,
      "step": 22085
    },
    {
      "entropy": 5.335692310333252,
      "epoch": 2.1923382294561335,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004521685455343731,
      "loss": 5.0398,
      "mean_token_accuracy": 0.17788098752498627,
      "num_tokens": 48383737.0,
      "step": 22090
    },
    {
      "entropy": 5.343779706954956,
      "epoch": 2.192834458118301,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045214670257424835,
      "loss": 5.0523,
      "mean_token_accuracy": 0.17299823015928267,
      "num_tokens": 48394547.0,
      "step": 22095
    },
    {
      "entropy": 5.392168045043945,
      "epoch": 2.1933306867804685,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004521248552212671,
      "loss": 5.0981,
      "mean_token_accuracy": 0.17641253769397736,
      "num_tokens": 48404906.0,
      "step": 22100
    },
    {
      "entropy": 5.372693967819214,
      "epoch": 2.193826915442636,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004521030034759711,
      "loss": 5.0506,
      "mean_token_accuracy": 0.17127309143543243,
      "num_tokens": 48416425.0,
      "step": 22105
    },
    {
      "entropy": 5.351895046234131,
      "epoch": 2.1943231441048034,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004520811473389023,
      "loss": 4.9977,
      "mean_token_accuracy": 0.17451567053794861,
      "num_tokens": 48426967.0,
      "step": 22110
    },
    {
      "entropy": 5.391579627990723,
      "epoch": 2.194819372766971,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045205928681060255,
      "loss": 5.124,
      "mean_token_accuracy": 0.17542713582515718,
      "num_tokens": 48439720.0,
      "step": 22115
    },
    {
      "entropy": 5.400501823425293,
      "epoch": 2.1953156014291384,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000452037421891614,
      "loss": 5.1114,
      "mean_token_accuracy": 0.1741048127412796,
      "num_tokens": 48449949.0,
      "step": 22120
    },
    {
      "entropy": 5.310616302490234,
      "epoch": 2.195811830091306,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045201555258247896,
      "loss": 4.9941,
      "mean_token_accuracy": 0.18232494443655015,
      "num_tokens": 48462515.0,
      "step": 22125
    },
    {
      "entropy": 5.298895740509034,
      "epoch": 2.196308058753474,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004519936788837396,
      "loss": 5.0118,
      "mean_token_accuracy": 0.18008548468351365,
      "num_tokens": 48471984.0,
      "step": 22130
    },
    {
      "entropy": 5.465793609619141,
      "epoch": 2.1968042874156413,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045197180079593836,
      "loss": 5.1088,
      "mean_token_accuracy": 0.16972734928131103,
      "num_tokens": 48482814.0,
      "step": 22135
    },
    {
      "entropy": 5.438660335540772,
      "epoch": 2.1973005160778087,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045194991831961777,
      "loss": 5.0603,
      "mean_token_accuracy": 0.17683944702148438,
      "num_tokens": 48493670.0,
      "step": 22140
    },
    {
      "entropy": 5.327695274353028,
      "epoch": 2.197796744739976,
      "grad_norm": 1.0,
      "learning_rate": 0.0004519280314553206,
      "loss": 5.0494,
      "mean_token_accuracy": 0.16586447060108184,
      "num_tokens": 48504758.0,
      "step": 22145
    },
    {
      "entropy": 5.4000168800354,
      "epoch": 2.1982929734021437,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004519061402035894,
      "loss": 5.0718,
      "mean_token_accuracy": 0.1622002065181732,
      "num_tokens": 48515846.0,
      "step": 22150
    },
    {
      "entropy": 5.397673940658569,
      "epoch": 2.198789202064311,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045188424456496723,
      "loss": 5.0643,
      "mean_token_accuracy": 0.17697402089834213,
      "num_tokens": 48526136.0,
      "step": 22155
    },
    {
      "entropy": 5.408515787124633,
      "epoch": 2.1992854307264786,
      "grad_norm": 1.125,
      "learning_rate": 0.00045186234453999693,
      "loss": 5.1651,
      "mean_token_accuracy": 0.17080941349267958,
      "num_tokens": 48536031.0,
      "step": 22160
    },
    {
      "entropy": 5.295034122467041,
      "epoch": 2.199781659388646,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045184044012922154,
      "loss": 4.9765,
      "mean_token_accuracy": 0.1789369761943817,
      "num_tokens": 48546965.0,
      "step": 22165
    },
    {
      "entropy": 5.402659273147583,
      "epoch": 2.200277888050814,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045181853133318425,
      "loss": 5.0695,
      "mean_token_accuracy": 0.16650130301713945,
      "num_tokens": 48557637.0,
      "step": 22170
    },
    {
      "entropy": 5.412844133377075,
      "epoch": 2.2007741167129815,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004517966181524285,
      "loss": 5.0948,
      "mean_token_accuracy": 0.17227240800857543,
      "num_tokens": 48568578.0,
      "step": 22175
    },
    {
      "entropy": 5.441598176956177,
      "epoch": 2.201270345375149,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004517747005874975,
      "loss": 5.1658,
      "mean_token_accuracy": 0.16931914389133454,
      "num_tokens": 48581782.0,
      "step": 22180
    },
    {
      "entropy": 5.285221433639526,
      "epoch": 2.2017665740373165,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004517527786389349,
      "loss": 5.0035,
      "mean_token_accuracy": 0.1811227649450302,
      "num_tokens": 48592432.0,
      "step": 22185
    },
    {
      "entropy": 5.328725051879883,
      "epoch": 2.202262802699484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045173085230728417,
      "loss": 5.0234,
      "mean_token_accuracy": 0.1785391166806221,
      "num_tokens": 48603251.0,
      "step": 22190
    },
    {
      "entropy": 5.405339622497559,
      "epoch": 2.2027590313616514,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004517089215930892,
      "loss": 5.1554,
      "mean_token_accuracy": 0.1700076773762703,
      "num_tokens": 48616118.0,
      "step": 22195
    },
    {
      "entropy": 5.445990705490113,
      "epoch": 2.203255260023819,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045168698649689375,
      "loss": 5.1064,
      "mean_token_accuracy": 0.1697595864534378,
      "num_tokens": 48626903.0,
      "step": 22200
    },
    {
      "entropy": 5.327866506576538,
      "epoch": 2.2037514886859864,
      "grad_norm": 0.953125,
      "learning_rate": 0.00045166504701924176,
      "loss": 4.9769,
      "mean_token_accuracy": 0.17848372012376784,
      "num_tokens": 48638186.0,
      "step": 22205
    },
    {
      "entropy": 5.366583347320557,
      "epoch": 2.204247717348154,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045164310316067727,
      "loss": 5.1042,
      "mean_token_accuracy": 0.1693960666656494,
      "num_tokens": 48649948.0,
      "step": 22210
    },
    {
      "entropy": 5.339461421966552,
      "epoch": 2.2047439460103218,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004516211549217444,
      "loss": 4.9877,
      "mean_token_accuracy": 0.17887333035469055,
      "num_tokens": 48660869.0,
      "step": 22215
    },
    {
      "entropy": 5.338683605194092,
      "epoch": 2.2052401746724892,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004515992023029875,
      "loss": 5.0418,
      "mean_token_accuracy": 0.1714923307299614,
      "num_tokens": 48673383.0,
      "step": 22220
    },
    {
      "entropy": 5.300400829315185,
      "epoch": 2.2057364033346567,
      "grad_norm": 1.03125,
      "learning_rate": 0.000451577245304951,
      "loss": 4.9745,
      "mean_token_accuracy": 0.1832749828696251,
      "num_tokens": 48684071.0,
      "step": 22225
    },
    {
      "entropy": 5.358835983276367,
      "epoch": 2.206232631996824,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004515552839281792,
      "loss": 5.0351,
      "mean_token_accuracy": 0.1722208708524704,
      "num_tokens": 48695432.0,
      "step": 22230
    },
    {
      "entropy": 5.388607311248779,
      "epoch": 2.2067288606589917,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004515333181732169,
      "loss": 5.0355,
      "mean_token_accuracy": 0.17861034274101256,
      "num_tokens": 48706102.0,
      "step": 22235
    },
    {
      "entropy": 5.363927555084229,
      "epoch": 2.207225089321159,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045151134804060863,
      "loss": 5.077,
      "mean_token_accuracy": 0.1692584738135338,
      "num_tokens": 48717748.0,
      "step": 22240
    },
    {
      "entropy": 5.33683295249939,
      "epoch": 2.2077213179833266,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004514893735308993,
      "loss": 4.9905,
      "mean_token_accuracy": 0.17893064618110657,
      "num_tokens": 48728941.0,
      "step": 22245
    },
    {
      "entropy": 5.318995523452759,
      "epoch": 2.208217546645494,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004514673946446338,
      "loss": 5.0175,
      "mean_token_accuracy": 0.17737738490104676,
      "num_tokens": 48738565.0,
      "step": 22250
    },
    {
      "entropy": 5.349276924133301,
      "epoch": 2.208713775307662,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004514454113823572,
      "loss": 5.044,
      "mean_token_accuracy": 0.16487522721290587,
      "num_tokens": 48749537.0,
      "step": 22255
    },
    {
      "entropy": 5.347049951553345,
      "epoch": 2.2092100039698295,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004514234237446146,
      "loss": 5.0309,
      "mean_token_accuracy": 0.16916547119617462,
      "num_tokens": 48760388.0,
      "step": 22260
    },
    {
      "entropy": 5.353564500808716,
      "epoch": 2.209706232631997,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045140143173195117,
      "loss": 5.0448,
      "mean_token_accuracy": 0.17024242579936982,
      "num_tokens": 48770551.0,
      "step": 22265
    },
    {
      "entropy": 5.3545300483703615,
      "epoch": 2.2102024612941644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004513794353449124,
      "loss": 5.0005,
      "mean_token_accuracy": 0.17469172477722167,
      "num_tokens": 48780739.0,
      "step": 22270
    },
    {
      "entropy": 5.369520092010498,
      "epoch": 2.210698689956332,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004513574345840437,
      "loss": 5.1102,
      "mean_token_accuracy": 0.16975859105587005,
      "num_tokens": 48791478.0,
      "step": 22275
    },
    {
      "entropy": 5.410419273376465,
      "epoch": 2.2111949186184994,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004513354294498907,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1698594108223915,
      "num_tokens": 48802202.0,
      "step": 22280
    },
    {
      "entropy": 5.400214147567749,
      "epoch": 2.211691147280667,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045131341994299895,
      "loss": 5.0505,
      "mean_token_accuracy": 0.1775912433862686,
      "num_tokens": 48813619.0,
      "step": 22285
    },
    {
      "entropy": 5.35226583480835,
      "epoch": 2.2121873759428343,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004512914060639143,
      "loss": 5.0166,
      "mean_token_accuracy": 0.18021336942911148,
      "num_tokens": 48824951.0,
      "step": 22290
    },
    {
      "entropy": 5.356893825531006,
      "epoch": 2.212683604605002,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004512693878131827,
      "loss": 5.0268,
      "mean_token_accuracy": 0.17706799656152725,
      "num_tokens": 48834328.0,
      "step": 22295
    },
    {
      "entropy": 5.367729663848877,
      "epoch": 2.2131798332671697,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045124736519135003,
      "loss": 5.0584,
      "mean_token_accuracy": 0.17184543311595918,
      "num_tokens": 48843880.0,
      "step": 22300
    },
    {
      "entropy": 5.3703697681427,
      "epoch": 2.213676061929337,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045122533819896257,
      "loss": 5.063,
      "mean_token_accuracy": 0.17761759757995604,
      "num_tokens": 48854893.0,
      "step": 22305
    },
    {
      "entropy": 5.399434423446655,
      "epoch": 2.2141722905915047,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004512033068365664,
      "loss": 5.1399,
      "mean_token_accuracy": 0.16559647917747497,
      "num_tokens": 48867308.0,
      "step": 22310
    },
    {
      "entropy": 5.449048089981079,
      "epoch": 2.214668519253672,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004511812711047079,
      "loss": 5.1159,
      "mean_token_accuracy": 0.17427487671375275,
      "num_tokens": 48880222.0,
      "step": 22315
    },
    {
      "entropy": 5.4072741031646725,
      "epoch": 2.2151647479158396,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045115923100393356,
      "loss": 5.0487,
      "mean_token_accuracy": 0.17111050933599473,
      "num_tokens": 48890345.0,
      "step": 22320
    },
    {
      "entropy": 5.3341383934021,
      "epoch": 2.215660976578007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045113718653478984,
      "loss": 4.9921,
      "mean_token_accuracy": 0.17607253640890122,
      "num_tokens": 48901267.0,
      "step": 22325
    },
    {
      "entropy": 5.401247501373291,
      "epoch": 2.2161572052401746,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004511151376978234,
      "loss": 5.0615,
      "mean_token_accuracy": 0.17451687902212143,
      "num_tokens": 48911739.0,
      "step": 22330
    },
    {
      "entropy": 5.371271848678589,
      "epoch": 2.216653433902342,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045109308449358114,
      "loss": 5.0256,
      "mean_token_accuracy": 0.17645665109157563,
      "num_tokens": 48922376.0,
      "step": 22335
    },
    {
      "entropy": 5.312776803970337,
      "epoch": 2.21714966256451,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004510710269226098,
      "loss": 4.9351,
      "mean_token_accuracy": 0.17910029143095016,
      "num_tokens": 48934101.0,
      "step": 22340
    },
    {
      "entropy": 5.262359762191773,
      "epoch": 2.2176458912266774,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004510489649854564,
      "loss": 5.0265,
      "mean_token_accuracy": 0.17769453972578048,
      "num_tokens": 48944684.0,
      "step": 22345
    },
    {
      "entropy": 5.369534015655518,
      "epoch": 2.218142119888845,
      "grad_norm": 1.0625,
      "learning_rate": 0.000451026898682668,
      "loss": 5.0284,
      "mean_token_accuracy": 0.1686280682682991,
      "num_tokens": 48955255.0,
      "step": 22350
    },
    {
      "entropy": 5.378615236282348,
      "epoch": 2.2186383485510124,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004510048280147918,
      "loss": 5.046,
      "mean_token_accuracy": 0.17433767020702362,
      "num_tokens": 48967034.0,
      "step": 22355
    },
    {
      "entropy": 5.36211519241333,
      "epoch": 2.21913457721318,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004509827529823751,
      "loss": 5.151,
      "mean_token_accuracy": 0.16978170573711396,
      "num_tokens": 48977903.0,
      "step": 22360
    },
    {
      "entropy": 5.363869047164917,
      "epoch": 2.2196308058753473,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045096067358596536,
      "loss": 5.0251,
      "mean_token_accuracy": 0.17855021506547927,
      "num_tokens": 48988889.0,
      "step": 22365
    },
    {
      "entropy": 5.448483991622925,
      "epoch": 2.220127034537515,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045093858982611006,
      "loss": 5.1326,
      "mean_token_accuracy": 0.15985297709703444,
      "num_tokens": 48998946.0,
      "step": 22370
    },
    {
      "entropy": 5.384618949890137,
      "epoch": 2.2206232631996823,
      "grad_norm": 1.0,
      "learning_rate": 0.0004509165017033569,
      "loss": 5.0349,
      "mean_token_accuracy": 0.1818299487233162,
      "num_tokens": 49008141.0,
      "step": 22375
    },
    {
      "entropy": 5.289153909683227,
      "epoch": 2.2211194918618498,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045089440921825354,
      "loss": 4.9447,
      "mean_token_accuracy": 0.18240997940301895,
      "num_tokens": 49019387.0,
      "step": 22380
    },
    {
      "entropy": 5.312760353088379,
      "epoch": 2.2216157205240172,
      "grad_norm": 1.0,
      "learning_rate": 0.00045087231237134787,
      "loss": 5.0033,
      "mean_token_accuracy": 0.17667342275381087,
      "num_tokens": 49030280.0,
      "step": 22385
    },
    {
      "entropy": 5.3585694313049315,
      "epoch": 2.222111949186185,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045085021116318783,
      "loss": 5.0019,
      "mean_token_accuracy": 0.17614959478378295,
      "num_tokens": 49041766.0,
      "step": 22390
    },
    {
      "entropy": 5.469907903671265,
      "epoch": 2.2226081778483526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004508281055943214,
      "loss": 5.2358,
      "mean_token_accuracy": 0.16442484259605408,
      "num_tokens": 49052931.0,
      "step": 22395
    },
    {
      "entropy": 5.368815279006958,
      "epoch": 2.22310440651052,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004508059956652969,
      "loss": 5.0145,
      "mean_token_accuracy": 0.1739718958735466,
      "num_tokens": 49064153.0,
      "step": 22400
    },
    {
      "entropy": 5.377976512908935,
      "epoch": 2.2236006351726876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004507838813766625,
      "loss": 5.0453,
      "mean_token_accuracy": 0.17426479011774063,
      "num_tokens": 49075857.0,
      "step": 22405
    },
    {
      "entropy": 5.44958086013794,
      "epoch": 2.224096863834855,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004507617627289666,
      "loss": 5.0875,
      "mean_token_accuracy": 0.17384012192487716,
      "num_tokens": 49086632.0,
      "step": 22410
    },
    {
      "entropy": 5.394014883041382,
      "epoch": 2.2245930924970225,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004507396397227577,
      "loss": 5.0377,
      "mean_token_accuracy": 0.17432418018579482,
      "num_tokens": 49098086.0,
      "step": 22415
    },
    {
      "entropy": 5.313254547119141,
      "epoch": 2.22508932115919,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045071751235858454,
      "loss": 4.955,
      "mean_token_accuracy": 0.1753916099667549,
      "num_tokens": 49108857.0,
      "step": 22420
    },
    {
      "entropy": 5.358402919769287,
      "epoch": 2.2255855498213575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004506953806369956,
      "loss": 4.9774,
      "mean_token_accuracy": 0.17778799086809158,
      "num_tokens": 49119609.0,
      "step": 22425
    },
    {
      "entropy": 5.367636108398438,
      "epoch": 2.2260817784835254,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004506732445585399,
      "loss": 5.0818,
      "mean_token_accuracy": 0.17572299987077714,
      "num_tokens": 49130047.0,
      "step": 22430
    },
    {
      "entropy": 5.317305183410644,
      "epoch": 2.226578007145693,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004506511041237662,
      "loss": 5.0135,
      "mean_token_accuracy": 0.17435785979032517,
      "num_tokens": 49140517.0,
      "step": 22435
    },
    {
      "entropy": 5.404394197463989,
      "epoch": 2.2270742358078603,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045062895933322366,
      "loss": 5.0409,
      "mean_token_accuracy": 0.17142440378665924,
      "num_tokens": 49151497.0,
      "step": 22440
    },
    {
      "entropy": 5.329073429107666,
      "epoch": 2.227570464470028,
      "grad_norm": 1.0,
      "learning_rate": 0.0004506068101874614,
      "loss": 4.9542,
      "mean_token_accuracy": 0.1778205469250679,
      "num_tokens": 49161568.0,
      "step": 22445
    },
    {
      "entropy": 5.379180669784546,
      "epoch": 2.2280666931321953,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045058465668702856,
      "loss": 5.0751,
      "mean_token_accuracy": 0.1739594653248787,
      "num_tokens": 49173554.0,
      "step": 22450
    },
    {
      "entropy": 5.390225553512574,
      "epoch": 2.2285629217943628,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045056249883247457,
      "loss": 5.0726,
      "mean_token_accuracy": 0.1738201007246971,
      "num_tokens": 49185399.0,
      "step": 22455
    },
    {
      "entropy": 5.379238605499268,
      "epoch": 2.2290591504565302,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045054033662434903,
      "loss": 5.0009,
      "mean_token_accuracy": 0.17741573452949524,
      "num_tokens": 49195489.0,
      "step": 22460
    },
    {
      "entropy": 5.295685291290283,
      "epoch": 2.2295553791186977,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045051817006320127,
      "loss": 4.9797,
      "mean_token_accuracy": 0.17503834664821624,
      "num_tokens": 49204652.0,
      "step": 22465
    },
    {
      "entropy": 5.316080617904663,
      "epoch": 2.230051607780865,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045049599914958124,
      "loss": 5.0709,
      "mean_token_accuracy": 0.1679056853055954,
      "num_tokens": 49215420.0,
      "step": 22470
    },
    {
      "entropy": 5.402918004989624,
      "epoch": 2.230547836443033,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004504738238840385,
      "loss": 5.0198,
      "mean_token_accuracy": 0.17857805341482164,
      "num_tokens": 49225375.0,
      "step": 22475
    },
    {
      "entropy": 5.408377933502197,
      "epoch": 2.2310440651052006,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004504516442671231,
      "loss": 5.0253,
      "mean_token_accuracy": 0.1716996416449547,
      "num_tokens": 49238399.0,
      "step": 22480
    },
    {
      "entropy": 5.347697114944458,
      "epoch": 2.231540293767368,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045042946029938495,
      "loss": 5.0111,
      "mean_token_accuracy": 0.1722542643547058,
      "num_tokens": 49251069.0,
      "step": 22485
    },
    {
      "entropy": 5.358260297775269,
      "epoch": 2.2320365224295355,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045040727198137416,
      "loss": 5.0551,
      "mean_token_accuracy": 0.1722538873553276,
      "num_tokens": 49263859.0,
      "step": 22490
    },
    {
      "entropy": 5.405295515060425,
      "epoch": 2.232532751091703,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045038507931364103,
      "loss": 5.0488,
      "mean_token_accuracy": 0.1725186288356781,
      "num_tokens": 49274846.0,
      "step": 22495
    },
    {
      "entropy": 5.311789608001709,
      "epoch": 2.2330289797538705,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045036288229673583,
      "loss": 4.9843,
      "mean_token_accuracy": 0.1809956058859825,
      "num_tokens": 49285197.0,
      "step": 22500
    },
    {
      "entropy": 5.292363500595092,
      "epoch": 2.233525208416038,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045034068093120907,
      "loss": 4.975,
      "mean_token_accuracy": 0.17241333276033402,
      "num_tokens": 49296238.0,
      "step": 22505
    },
    {
      "entropy": 5.469737243652344,
      "epoch": 2.2340214370782054,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004503184752176112,
      "loss": 5.2202,
      "mean_token_accuracy": 0.16537212580442429,
      "num_tokens": 49308262.0,
      "step": 22510
    },
    {
      "entropy": 5.396468734741211,
      "epoch": 2.2345176657403734,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000450296265156493,
      "loss": 5.0003,
      "mean_token_accuracy": 0.1746342658996582,
      "num_tokens": 49318881.0,
      "step": 22515
    },
    {
      "entropy": 5.433895397186279,
      "epoch": 2.235013894402541,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045027405074840506,
      "loss": 5.1374,
      "mean_token_accuracy": 0.17037462890148164,
      "num_tokens": 49329470.0,
      "step": 22520
    },
    {
      "entropy": 5.362544059753418,
      "epoch": 2.2355101230647083,
      "grad_norm": 1.0,
      "learning_rate": 0.0004502518319938984,
      "loss": 5.0362,
      "mean_token_accuracy": 0.17630160003900527,
      "num_tokens": 49339806.0,
      "step": 22525
    },
    {
      "entropy": 5.438360786437988,
      "epoch": 2.236006351726876,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045022960889352394,
      "loss": 5.1383,
      "mean_token_accuracy": 0.17111808508634568,
      "num_tokens": 49350295.0,
      "step": 22530
    },
    {
      "entropy": 5.3534255027771,
      "epoch": 2.2365025803890433,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045020738144783273,
      "loss": 5.045,
      "mean_token_accuracy": 0.17602352946996688,
      "num_tokens": 49360735.0,
      "step": 22535
    },
    {
      "entropy": 5.341494989395142,
      "epoch": 2.2369988090512107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004501851496573761,
      "loss": 5.0148,
      "mean_token_accuracy": 0.17164534479379653,
      "num_tokens": 49372278.0,
      "step": 22540
    },
    {
      "entropy": 5.362668085098266,
      "epoch": 2.237495037713378,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004501629135227051,
      "loss": 5.0943,
      "mean_token_accuracy": 0.16710684150457383,
      "num_tokens": 49382842.0,
      "step": 22545
    },
    {
      "entropy": 5.41583480834961,
      "epoch": 2.2379912663755457,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004501406730443714,
      "loss": 5.1083,
      "mean_token_accuracy": 0.17093432396650315,
      "num_tokens": 49394633.0,
      "step": 22550
    },
    {
      "entropy": 5.440032005310059,
      "epoch": 2.238487495037713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045011842822292636,
      "loss": 5.0721,
      "mean_token_accuracy": 0.17457079738378525,
      "num_tokens": 49404731.0,
      "step": 22555
    },
    {
      "entropy": 5.421986722946167,
      "epoch": 2.238983723699881,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004500961790589216,
      "loss": 5.0632,
      "mean_token_accuracy": 0.17392938137054442,
      "num_tokens": 49415918.0,
      "step": 22560
    },
    {
      "entropy": 5.319574165344238,
      "epoch": 2.2394799523620486,
      "grad_norm": 1.09375,
      "learning_rate": 0.000450073925552909,
      "loss": 5.0156,
      "mean_token_accuracy": 0.17571015208959578,
      "num_tokens": 49427458.0,
      "step": 22565
    },
    {
      "entropy": 5.36943416595459,
      "epoch": 2.239976181024216,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004500516677054403,
      "loss": 5.0297,
      "mean_token_accuracy": 0.1783345401287079,
      "num_tokens": 49437696.0,
      "step": 22570
    },
    {
      "entropy": 5.4639380931854244,
      "epoch": 2.2404724096863835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045002940551706726,
      "loss": 5.2431,
      "mean_token_accuracy": 0.16499289274215698,
      "num_tokens": 49449220.0,
      "step": 22575
    },
    {
      "entropy": 5.420271635055542,
      "epoch": 2.240968638348551,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045000713898834225,
      "loss": 5.052,
      "mean_token_accuracy": 0.17399043440818787,
      "num_tokens": 49459624.0,
      "step": 22580
    },
    {
      "entropy": 5.375232553482055,
      "epoch": 2.2414648670107185,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004499848681198173,
      "loss": 4.9852,
      "mean_token_accuracy": 0.18007365763187408,
      "num_tokens": 49471813.0,
      "step": 22585
    },
    {
      "entropy": 5.388800525665284,
      "epoch": 2.241961095672886,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004499625929120447,
      "loss": 5.0569,
      "mean_token_accuracy": 0.17420250326395034,
      "num_tokens": 49482624.0,
      "step": 22590
    },
    {
      "entropy": 5.402924394607544,
      "epoch": 2.2424573243350534,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004499403133655768,
      "loss": 5.0799,
      "mean_token_accuracy": 0.16981296092271805,
      "num_tokens": 49493791.0,
      "step": 22595
    },
    {
      "entropy": 5.389824628829956,
      "epoch": 2.2429535529972213,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004499180294809661,
      "loss": 5.0995,
      "mean_token_accuracy": 0.17327599823474885,
      "num_tokens": 49504060.0,
      "step": 22600
    },
    {
      "entropy": 5.293384599685669,
      "epoch": 2.243449781659389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044989574125876516,
      "loss": 4.9321,
      "mean_token_accuracy": 0.18103266805410384,
      "num_tokens": 49515962.0,
      "step": 22605
    },
    {
      "entropy": 5.404565477371216,
      "epoch": 2.2439460103215563,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004498734486995267,
      "loss": 5.065,
      "mean_token_accuracy": 0.16984232217073442,
      "num_tokens": 49526676.0,
      "step": 22610
    },
    {
      "entropy": 5.365847969055176,
      "epoch": 2.2444422389837237,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044985115180380355,
      "loss": 5.0699,
      "mean_token_accuracy": 0.16803995966911317,
      "num_tokens": 49537347.0,
      "step": 22615
    },
    {
      "entropy": 5.388544940948487,
      "epoch": 2.2449384676458912,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044982885057214857,
      "loss": 5.0732,
      "mean_token_accuracy": 0.16782470494508744,
      "num_tokens": 49547422.0,
      "step": 22620
    },
    {
      "entropy": 5.336597061157226,
      "epoch": 2.2454346963080587,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044980654500511485,
      "loss": 4.9146,
      "mean_token_accuracy": 0.18375383764505387,
      "num_tokens": 49557250.0,
      "step": 22625
    },
    {
      "entropy": 5.433314228057862,
      "epoch": 2.245930924970226,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044978423510325544,
      "loss": 5.126,
      "mean_token_accuracy": 0.1669748067855835,
      "num_tokens": 49567106.0,
      "step": 22630
    },
    {
      "entropy": 5.383455801010132,
      "epoch": 2.2464271536323936,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004497619208671237,
      "loss": 5.0691,
      "mean_token_accuracy": 0.1706724852323532,
      "num_tokens": 49578978.0,
      "step": 22635
    },
    {
      "entropy": 5.44146237373352,
      "epoch": 2.246923382294561,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004497396022972728,
      "loss": 5.1091,
      "mean_token_accuracy": 0.16828266978263856,
      "num_tokens": 49589730.0,
      "step": 22640
    },
    {
      "entropy": 5.361346864700318,
      "epoch": 2.247419610956729,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004497172793942564,
      "loss": 5.0216,
      "mean_token_accuracy": 0.1767912596464157,
      "num_tokens": 49600881.0,
      "step": 22645
    },
    {
      "entropy": 5.370608758926392,
      "epoch": 2.2479158396188965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004496949521586279,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17675362080335616,
      "num_tokens": 49611120.0,
      "step": 22650
    },
    {
      "entropy": 5.387104845046997,
      "epoch": 2.248412068281064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044967262059094097,
      "loss": 5.0931,
      "mean_token_accuracy": 0.17105765342712403,
      "num_tokens": 49622978.0,
      "step": 22655
    },
    {
      "entropy": 5.35486741065979,
      "epoch": 2.2489082969432315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044965028469174956,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17557128816843032,
      "num_tokens": 49633678.0,
      "step": 22660
    },
    {
      "entropy": 5.377434635162354,
      "epoch": 2.249404525605399,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044962794446160736,
      "loss": 5.0556,
      "mean_token_accuracy": 0.1718231499195099,
      "num_tokens": 49644886.0,
      "step": 22665
    },
    {
      "entropy": 5.381330871582032,
      "epoch": 2.2499007542675664,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044960559990106844,
      "loss": 5.027,
      "mean_token_accuracy": 0.18187296241521836,
      "num_tokens": 49656790.0,
      "step": 22670
    },
    {
      "entropy": 5.384118127822876,
      "epoch": 2.250396982929734,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004495832510106869,
      "loss": 5.0618,
      "mean_token_accuracy": 0.16886563301086427,
      "num_tokens": 49667943.0,
      "step": 22675
    },
    {
      "entropy": 5.37179708480835,
      "epoch": 2.2508932115919014,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004495608977910169,
      "loss": 5.0399,
      "mean_token_accuracy": 0.16788456887006759,
      "num_tokens": 49678652.0,
      "step": 22680
    },
    {
      "entropy": 5.329069328308106,
      "epoch": 2.2513894402540693,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004495385402426128,
      "loss": 4.9968,
      "mean_token_accuracy": 0.17906121164560318,
      "num_tokens": 49688951.0,
      "step": 22685
    },
    {
      "entropy": 5.349430894851684,
      "epoch": 2.2518856689162368,
      "grad_norm": 1.03125,
      "learning_rate": 0.000449516178366029,
      "loss": 5.0541,
      "mean_token_accuracy": 0.17023944109678268,
      "num_tokens": 49698750.0,
      "step": 22690
    },
    {
      "entropy": 5.371826314926148,
      "epoch": 2.2523818975784042,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044949381216182,
      "loss": 5.1508,
      "mean_token_accuracy": 0.1660812094807625,
      "num_tokens": 49709442.0,
      "step": 22695
    },
    {
      "entropy": 5.374296474456787,
      "epoch": 2.2528781262405717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044947144163054036,
      "loss": 5.0481,
      "mean_token_accuracy": 0.1780099630355835,
      "num_tokens": 49720517.0,
      "step": 22700
    },
    {
      "entropy": 5.4168306350708,
      "epoch": 2.253374354902739,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004494490667727451,
      "loss": 5.0296,
      "mean_token_accuracy": 0.1737184703350067,
      "num_tokens": 49731413.0,
      "step": 22705
    },
    {
      "entropy": 5.470418691635132,
      "epoch": 2.2538705835649067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044942668758898873,
      "loss": 5.2095,
      "mean_token_accuracy": 0.16114892810583115,
      "num_tokens": 49742492.0,
      "step": 22710
    },
    {
      "entropy": 5.373818969726562,
      "epoch": 2.254366812227074,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004494043040798265,
      "loss": 5.0513,
      "mean_token_accuracy": 0.16837262213230134,
      "num_tokens": 49753981.0,
      "step": 22715
    },
    {
      "entropy": 5.373852872848511,
      "epoch": 2.2548630408892416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004493819162458133,
      "loss": 5.0809,
      "mean_token_accuracy": 0.173592671751976,
      "num_tokens": 49764718.0,
      "step": 22720
    },
    {
      "entropy": 5.415003967285156,
      "epoch": 2.255359269551409,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004493595240875043,
      "loss": 5.0324,
      "mean_token_accuracy": 0.17139716744422911,
      "num_tokens": 49775361.0,
      "step": 22725
    },
    {
      "entropy": 5.344558525085449,
      "epoch": 2.255855498213577,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044933712760545486,
      "loss": 5.0229,
      "mean_token_accuracy": 0.17108300626277922,
      "num_tokens": 49786670.0,
      "step": 22730
    },
    {
      "entropy": 5.40809211730957,
      "epoch": 2.2563517268757445,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004493147268002203,
      "loss": 5.1574,
      "mean_token_accuracy": 0.17028712928295137,
      "num_tokens": 49796514.0,
      "step": 22735
    },
    {
      "entropy": 5.335297060012818,
      "epoch": 2.256847955537912,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044929232167235613,
      "loss": 4.991,
      "mean_token_accuracy": 0.18380380123853685,
      "num_tokens": 49807042.0,
      "step": 22740
    },
    {
      "entropy": 5.413664436340332,
      "epoch": 2.2573441842000794,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004492699122224179,
      "loss": 5.0684,
      "mean_token_accuracy": 0.1713469758629799,
      "num_tokens": 49818779.0,
      "step": 22745
    },
    {
      "entropy": 5.408750581741333,
      "epoch": 2.257840412862247,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044924749845096144,
      "loss": 5.0897,
      "mean_token_accuracy": 0.17181849479675293,
      "num_tokens": 49829901.0,
      "step": 22750
    },
    {
      "entropy": 5.3596494674682615,
      "epoch": 2.2583366415244144,
      "grad_norm": 1.125,
      "learning_rate": 0.0004492250803585425,
      "loss": 5.0629,
      "mean_token_accuracy": 0.17281948328018187,
      "num_tokens": 49840837.0,
      "step": 22755
    },
    {
      "entropy": 5.391885232925415,
      "epoch": 2.258832870186582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004492026579457169,
      "loss": 5.0453,
      "mean_token_accuracy": 0.16628299355506898,
      "num_tokens": 49853052.0,
      "step": 22760
    },
    {
      "entropy": 5.362270164489746,
      "epoch": 2.2593290988487493,
      "grad_norm": 1.0,
      "learning_rate": 0.0004491802312130408,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1710537776350975,
      "num_tokens": 49864210.0,
      "step": 22765
    },
    {
      "entropy": 5.339938020706176,
      "epoch": 2.2598253275109172,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004491578001610703,
      "loss": 5.0152,
      "mean_token_accuracy": 0.1803627148270607,
      "num_tokens": 49874995.0,
      "step": 22770
    },
    {
      "entropy": 5.320356130599976,
      "epoch": 2.2603215561730847,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044913536479036163,
      "loss": 5.0146,
      "mean_token_accuracy": 0.17244176268577577,
      "num_tokens": 49886240.0,
      "step": 22775
    },
    {
      "entropy": 5.313217210769653,
      "epoch": 2.260817784835252,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044911292510147117,
      "loss": 5.0231,
      "mean_token_accuracy": 0.17355303466320038,
      "num_tokens": 49897580.0,
      "step": 22780
    },
    {
      "entropy": 5.345343208312988,
      "epoch": 2.2613140134974197,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004490904810949553,
      "loss": 5.0626,
      "mean_token_accuracy": 0.1741154670715332,
      "num_tokens": 49907280.0,
      "step": 22785
    },
    {
      "entropy": 5.395234632492065,
      "epoch": 2.261810242159587,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044906803277137064,
      "loss": 5.0676,
      "mean_token_accuracy": 0.176661017537117,
      "num_tokens": 49917552.0,
      "step": 22790
    },
    {
      "entropy": 5.34765658378601,
      "epoch": 2.2623064708217546,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004490455801312738,
      "loss": 5.0508,
      "mean_token_accuracy": 0.17010846138000488,
      "num_tokens": 49930610.0,
      "step": 22795
    },
    {
      "entropy": 5.387277030944825,
      "epoch": 2.262802699483922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004490231231752217,
      "loss": 5.081,
      "mean_token_accuracy": 0.16591572761535645,
      "num_tokens": 49941054.0,
      "step": 22800
    },
    {
      "entropy": 5.301331186294556,
      "epoch": 2.2632989281460896,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004490006619037711,
      "loss": 4.9285,
      "mean_token_accuracy": 0.1831444561481476,
      "num_tokens": 49950885.0,
      "step": 22805
    },
    {
      "entropy": 5.33036241531372,
      "epoch": 2.263795156808257,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044897819631747903,
      "loss": 5.0467,
      "mean_token_accuracy": 0.1733948916196823,
      "num_tokens": 49961723.0,
      "step": 22810
    },
    {
      "entropy": 5.37024998664856,
      "epoch": 2.2642913854704245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044895572641690256,
      "loss": 5.0109,
      "mean_token_accuracy": 0.17636981457471848,
      "num_tokens": 49974124.0,
      "step": 22815
    },
    {
      "entropy": 5.365680885314942,
      "epoch": 2.2647876141325924,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000448933252202599,
      "loss": 5.0462,
      "mean_token_accuracy": 0.1696328654885292,
      "num_tokens": 49985792.0,
      "step": 22820
    },
    {
      "entropy": 5.353872966766358,
      "epoch": 2.26528384279476,
      "grad_norm": 1.0,
      "learning_rate": 0.00044891077367512544,
      "loss": 4.9761,
      "mean_token_accuracy": 0.17573423981666564,
      "num_tokens": 49996935.0,
      "step": 22825
    },
    {
      "entropy": 5.445383977890015,
      "epoch": 2.2657800714569274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004488882908350396,
      "loss": 5.1455,
      "mean_token_accuracy": 0.16712917685508727,
      "num_tokens": 50007545.0,
      "step": 22830
    },
    {
      "entropy": 5.393108320236206,
      "epoch": 2.266276300119095,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004488658036828987,
      "loss": 5.0607,
      "mean_token_accuracy": 0.17001772224903106,
      "num_tokens": 50018797.0,
      "step": 22835
    },
    {
      "entropy": 5.3629903316497805,
      "epoch": 2.2667725287812623,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004488433122192606,
      "loss": 5.029,
      "mean_token_accuracy": 0.17761739045381547,
      "num_tokens": 50030884.0,
      "step": 22840
    },
    {
      "entropy": 5.428975391387939,
      "epoch": 2.26726875744343,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044882081644468286,
      "loss": 5.1602,
      "mean_token_accuracy": 0.16632796674966813,
      "num_tokens": 50042327.0,
      "step": 22845
    },
    {
      "entropy": 5.418456411361694,
      "epoch": 2.2677649861055973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044879831635972345,
      "loss": 5.1143,
      "mean_token_accuracy": 0.1745055079460144,
      "num_tokens": 50053733.0,
      "step": 22850
    },
    {
      "entropy": 5.366400480270386,
      "epoch": 2.268261214767765,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004487758119649403,
      "loss": 5.0097,
      "mean_token_accuracy": 0.1797109991312027,
      "num_tokens": 50064153.0,
      "step": 22855
    },
    {
      "entropy": 5.3112798690795895,
      "epoch": 2.2687574434299327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044875330326089153,
      "loss": 4.9725,
      "mean_token_accuracy": 0.18851456344127654,
      "num_tokens": 50074397.0,
      "step": 22860
    },
    {
      "entropy": 5.304088830947876,
      "epoch": 2.2692536720921,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004487307902481352,
      "loss": 4.9742,
      "mean_token_accuracy": 0.17648680210113527,
      "num_tokens": 50084074.0,
      "step": 22865
    },
    {
      "entropy": 5.36228928565979,
      "epoch": 2.2697499007542676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044870827292722967,
      "loss": 4.9948,
      "mean_token_accuracy": 0.16644156873226165,
      "num_tokens": 50094535.0,
      "step": 22870
    },
    {
      "entropy": 5.419567012786866,
      "epoch": 2.270246129416435,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004486857512987333,
      "loss": 5.0879,
      "mean_token_accuracy": 0.16938829571008682,
      "num_tokens": 50105907.0,
      "step": 22875
    },
    {
      "entropy": 5.34932312965393,
      "epoch": 2.2707423580786026,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044866322536320447,
      "loss": 5.0501,
      "mean_token_accuracy": 0.17870703786611558,
      "num_tokens": 50117361.0,
      "step": 22880
    },
    {
      "entropy": 5.388257074356079,
      "epoch": 2.27123858674077,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044864069512120185,
      "loss": 5.0468,
      "mean_token_accuracy": 0.17965768724679948,
      "num_tokens": 50128598.0,
      "step": 22885
    },
    {
      "entropy": 5.375472354888916,
      "epoch": 2.2717348154029375,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044861816057328424,
      "loss": 5.0528,
      "mean_token_accuracy": 0.16962587237358093,
      "num_tokens": 50140470.0,
      "step": 22890
    },
    {
      "entropy": 5.391121959686279,
      "epoch": 2.272231044065105,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044859562172001036,
      "loss": 5.1054,
      "mean_token_accuracy": 0.17060508877038955,
      "num_tokens": 50150950.0,
      "step": 22895
    },
    {
      "entropy": 5.319776487350464,
      "epoch": 2.2727272727272725,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004485730785619391,
      "loss": 4.9752,
      "mean_token_accuracy": 0.17912409603595733,
      "num_tokens": 50161173.0,
      "step": 22900
    },
    {
      "entropy": 5.409841394424438,
      "epoch": 2.2732235013894404,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004485505310996295,
      "loss": 5.1632,
      "mean_token_accuracy": 0.16716630309820174,
      "num_tokens": 50173874.0,
      "step": 22905
    },
    {
      "entropy": 5.494783592224121,
      "epoch": 2.273719730051608,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004485279793336406,
      "loss": 5.1191,
      "mean_token_accuracy": 0.16688870936632155,
      "num_tokens": 50185123.0,
      "step": 22910
    },
    {
      "entropy": 5.397196054458618,
      "epoch": 2.2742159587137754,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004485054232645319,
      "loss": 5.0862,
      "mean_token_accuracy": 0.17337239235639573,
      "num_tokens": 50195699.0,
      "step": 22915
    },
    {
      "entropy": 5.421840238571167,
      "epoch": 2.274712187375943,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004484828628928625,
      "loss": 5.0925,
      "mean_token_accuracy": 0.1764795958995819,
      "num_tokens": 50206623.0,
      "step": 22920
    },
    {
      "entropy": 5.440144681930542,
      "epoch": 2.2752084160381103,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044846029821919193,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1693646490573883,
      "num_tokens": 50216086.0,
      "step": 22925
    },
    {
      "entropy": 5.344475555419922,
      "epoch": 2.275704644700278,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004484377292440797,
      "loss": 5.0599,
      "mean_token_accuracy": 0.16884406954050063,
      "num_tokens": 50228472.0,
      "step": 22930
    },
    {
      "entropy": 5.3987305641174315,
      "epoch": 2.2762008733624453,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044841515596808553,
      "loss": 5.0741,
      "mean_token_accuracy": 0.16760554611682893,
      "num_tokens": 50238607.0,
      "step": 22935
    },
    {
      "entropy": 5.381375169754028,
      "epoch": 2.276697102024613,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044839257839176914,
      "loss": 5.0312,
      "mean_token_accuracy": 0.17327709197998048,
      "num_tokens": 50249890.0,
      "step": 22940
    },
    {
      "entropy": 5.3408080577850345,
      "epoch": 2.2771933306867806,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044836999651569055,
      "loss": 5.0026,
      "mean_token_accuracy": 0.18089056611061097,
      "num_tokens": 50260234.0,
      "step": 22945
    },
    {
      "entropy": 5.360384845733643,
      "epoch": 2.277689559348948,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004483474103404095,
      "loss": 5.0839,
      "mean_token_accuracy": 0.17564980387687684,
      "num_tokens": 50271605.0,
      "step": 22950
    },
    {
      "entropy": 5.380985164642334,
      "epoch": 2.2781857880111156,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044832481986648633,
      "loss": 4.9362,
      "mean_token_accuracy": 0.17911087572574616,
      "num_tokens": 50281837.0,
      "step": 22955
    },
    {
      "entropy": 5.311028051376343,
      "epoch": 2.278682016673283,
      "grad_norm": 1.0625,
      "learning_rate": 0.000448302225094481,
      "loss": 5.016,
      "mean_token_accuracy": 0.17085451185703276,
      "num_tokens": 50292524.0,
      "step": 22960
    },
    {
      "entropy": 5.386628293991089,
      "epoch": 2.2791782453354505,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044827962602495393,
      "loss": 5.0095,
      "mean_token_accuracy": 0.17574279457330705,
      "num_tokens": 50304009.0,
      "step": 22965
    },
    {
      "entropy": 5.38130259513855,
      "epoch": 2.279674473997618,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044825702265846557,
      "loss": 5.0963,
      "mean_token_accuracy": 0.17231747806072234,
      "num_tokens": 50316103.0,
      "step": 22970
    },
    {
      "entropy": 5.376050043106079,
      "epoch": 2.2801707026597855,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004482344149955763,
      "loss": 5.0937,
      "mean_token_accuracy": 0.17449478507041932,
      "num_tokens": 50326127.0,
      "step": 22975
    },
    {
      "entropy": 5.329447841644287,
      "epoch": 2.280666931321953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004482118030368469,
      "loss": 4.9804,
      "mean_token_accuracy": 0.17767466604709625,
      "num_tokens": 50337044.0,
      "step": 22980
    },
    {
      "entropy": 5.372458600997925,
      "epoch": 2.2811631599841204,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000448189186782838,
      "loss": 5.1802,
      "mean_token_accuracy": 0.1658843457698822,
      "num_tokens": 50347796.0,
      "step": 22985
    },
    {
      "entropy": 5.349428272247314,
      "epoch": 2.2816593886462884,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044816656623411044,
      "loss": 5.0573,
      "mean_token_accuracy": 0.17399250119924545,
      "num_tokens": 50358454.0,
      "step": 22990
    },
    {
      "entropy": 5.3466148853302,
      "epoch": 2.282155617308456,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004481439413912252,
      "loss": 4.977,
      "mean_token_accuracy": 0.1735505536198616,
      "num_tokens": 50369415.0,
      "step": 22995
    },
    {
      "entropy": 5.359826850891113,
      "epoch": 2.2826518459706233,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044812131225474333,
      "loss": 5.0565,
      "mean_token_accuracy": 0.1690649628639221,
      "num_tokens": 50379977.0,
      "step": 23000
    },
    {
      "entropy": 5.410555982589722,
      "epoch": 2.283148074632791,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004480986788252259,
      "loss": 5.075,
      "mean_token_accuracy": 0.1718619480729103,
      "num_tokens": 50390465.0,
      "step": 23005
    },
    {
      "entropy": 5.419961261749267,
      "epoch": 2.2836443032949583,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004480760411032343,
      "loss": 5.0931,
      "mean_token_accuracy": 0.17376925200223922,
      "num_tokens": 50402230.0,
      "step": 23010
    },
    {
      "entropy": 5.293108367919922,
      "epoch": 2.2841405319571257,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004480533990893298,
      "loss": 4.9568,
      "mean_token_accuracy": 0.18222106248140335,
      "num_tokens": 50413819.0,
      "step": 23015
    },
    {
      "entropy": 5.521878290176391,
      "epoch": 2.284636760619293,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004480307527840739,
      "loss": 5.2618,
      "mean_token_accuracy": 0.15550881177186965,
      "num_tokens": 50424977.0,
      "step": 23020
    },
    {
      "entropy": 5.448246622085572,
      "epoch": 2.285132989281461,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004480081021880282,
      "loss": 5.0635,
      "mean_token_accuracy": 0.17379000633955002,
      "num_tokens": 50438357.0,
      "step": 23025
    },
    {
      "entropy": 5.359100866317749,
      "epoch": 2.2856292179436286,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004479854473017543,
      "loss": 5.075,
      "mean_token_accuracy": 0.1715774729847908,
      "num_tokens": 50449081.0,
      "step": 23030
    },
    {
      "entropy": 5.28122296333313,
      "epoch": 2.286125446605796,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044796278812581417,
      "loss": 4.9765,
      "mean_token_accuracy": 0.17857685685157776,
      "num_tokens": 50460358.0,
      "step": 23035
    },
    {
      "entropy": 5.382869386672974,
      "epoch": 2.2866216752679636,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004479401246607695,
      "loss": 5.0375,
      "mean_token_accuracy": 0.17566006034612655,
      "num_tokens": 50470955.0,
      "step": 23040
    },
    {
      "entropy": 5.372187376022339,
      "epoch": 2.287117903930131,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004479174569071823,
      "loss": 5.0094,
      "mean_token_accuracy": 0.1773396074771881,
      "num_tokens": 50482250.0,
      "step": 23045
    },
    {
      "entropy": 5.3434113502502445,
      "epoch": 2.2876141325922985,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004478947848656149,
      "loss": 5.0616,
      "mean_token_accuracy": 0.17366839498281478,
      "num_tokens": 50493868.0,
      "step": 23050
    },
    {
      "entropy": 5.391402292251587,
      "epoch": 2.288110361254466,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044787210853662943,
      "loss": 5.1064,
      "mean_token_accuracy": 0.17906535863876344,
      "num_tokens": 50504012.0,
      "step": 23055
    },
    {
      "entropy": 5.330997467041016,
      "epoch": 2.2886065899166335,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004478494279207881,
      "loss": 4.9909,
      "mean_token_accuracy": 0.18114234507083893,
      "num_tokens": 50515503.0,
      "step": 23060
    },
    {
      "entropy": 5.39494047164917,
      "epoch": 2.289102818578801,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044782674301865346,
      "loss": 5.0443,
      "mean_token_accuracy": 0.17424046695232392,
      "num_tokens": 50526180.0,
      "step": 23065
    },
    {
      "entropy": 5.37973370552063,
      "epoch": 2.2895990472409684,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004478040538307879,
      "loss": 5.0633,
      "mean_token_accuracy": 0.16908363699913026,
      "num_tokens": 50536711.0,
      "step": 23070
    },
    {
      "entropy": 5.395523881912231,
      "epoch": 2.2900952759031363,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004477813603577543,
      "loss": 5.1189,
      "mean_token_accuracy": 0.1664086565375328,
      "num_tokens": 50547736.0,
      "step": 23075
    },
    {
      "entropy": 5.349597024917602,
      "epoch": 2.290591504565304,
      "grad_norm": 1.0,
      "learning_rate": 0.00044775866260011516,
      "loss": 5.0093,
      "mean_token_accuracy": 0.17820306718349457,
      "num_tokens": 50558173.0,
      "step": 23080
    },
    {
      "entropy": 5.363086605072022,
      "epoch": 2.2910877332274713,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004477359605584335,
      "loss": 5.0511,
      "mean_token_accuracy": 0.17337346971035003,
      "num_tokens": 50568421.0,
      "step": 23085
    },
    {
      "entropy": 5.321041488647461,
      "epoch": 2.2915839618896388,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044771325423327227,
      "loss": 5.0065,
      "mean_token_accuracy": 0.17447673231363298,
      "num_tokens": 50579182.0,
      "step": 23090
    },
    {
      "entropy": 5.3597839832305905,
      "epoch": 2.2920801905518062,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004476905436251945,
      "loss": 5.0421,
      "mean_token_accuracy": 0.176125967502594,
      "num_tokens": 50590916.0,
      "step": 23095
    },
    {
      "entropy": 5.256513833999634,
      "epoch": 2.2925764192139737,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044766782873476333,
      "loss": 4.9239,
      "mean_token_accuracy": 0.1834631621837616,
      "num_tokens": 50600682.0,
      "step": 23100
    },
    {
      "entropy": 5.443897676467896,
      "epoch": 2.293072647876141,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000447645109562542,
      "loss": 5.2033,
      "mean_token_accuracy": 0.161502443253994,
      "num_tokens": 50612103.0,
      "step": 23105
    },
    {
      "entropy": 5.462101793289184,
      "epoch": 2.293568876538309,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044762238610909414,
      "loss": 5.1347,
      "mean_token_accuracy": 0.1677398905158043,
      "num_tokens": 50623894.0,
      "step": 23110
    },
    {
      "entropy": 5.439551734924317,
      "epoch": 2.2940651052004766,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004475996583749829,
      "loss": 5.0748,
      "mean_token_accuracy": 0.1719006761908531,
      "num_tokens": 50634495.0,
      "step": 23115
    },
    {
      "entropy": 5.440600824356079,
      "epoch": 2.294561333862644,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044757692636077215,
      "loss": 5.2226,
      "mean_token_accuracy": 0.16398790404200553,
      "num_tokens": 50645921.0,
      "step": 23120
    },
    {
      "entropy": 5.457008171081543,
      "epoch": 2.2950575625248115,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044755419006702545,
      "loss": 5.0754,
      "mean_token_accuracy": 0.16863217800855637,
      "num_tokens": 50657047.0,
      "step": 23125
    },
    {
      "entropy": 5.405387115478516,
      "epoch": 2.295553791186979,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044753144949430667,
      "loss": 5.0169,
      "mean_token_accuracy": 0.1751677304506302,
      "num_tokens": 50668271.0,
      "step": 23130
    },
    {
      "entropy": 5.308016538619995,
      "epoch": 2.2960500198491465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004475087046431798,
      "loss": 4.9926,
      "mean_token_accuracy": 0.17927819192409516,
      "num_tokens": 50678870.0,
      "step": 23135
    },
    {
      "entropy": 5.312892580032349,
      "epoch": 2.296546248511314,
      "grad_norm": 1.0,
      "learning_rate": 0.0004474859555142087,
      "loss": 5.0552,
      "mean_token_accuracy": 0.17789046466350555,
      "num_tokens": 50690322.0,
      "step": 23140
    },
    {
      "entropy": 5.354588365554809,
      "epoch": 2.2970424771734814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044746320210795753,
      "loss": 4.9879,
      "mean_token_accuracy": 0.17957944720983504,
      "num_tokens": 50701958.0,
      "step": 23145
    },
    {
      "entropy": 5.32082724571228,
      "epoch": 2.297538705835649,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004474404444249906,
      "loss": 5.0627,
      "mean_token_accuracy": 0.17661184817552567,
      "num_tokens": 50712962.0,
      "step": 23150
    },
    {
      "entropy": 5.379194498062134,
      "epoch": 2.2980349344978164,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044741768246587226,
      "loss": 5.0947,
      "mean_token_accuracy": 0.17538570910692214,
      "num_tokens": 50723442.0,
      "step": 23155
    },
    {
      "entropy": 5.383048820495605,
      "epoch": 2.2985311631599843,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004473949162311669,
      "loss": 4.9582,
      "mean_token_accuracy": 0.17495406419038773,
      "num_tokens": 50733188.0,
      "step": 23160
    },
    {
      "entropy": 5.393789148330688,
      "epoch": 2.2990273918221518,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004473721457214391,
      "loss": 5.113,
      "mean_token_accuracy": 0.16474495977163314,
      "num_tokens": 50744023.0,
      "step": 23165
    },
    {
      "entropy": 5.47245831489563,
      "epoch": 2.2995236204843192,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004473493709372535,
      "loss": 5.2666,
      "mean_token_accuracy": 0.1692916989326477,
      "num_tokens": 50756807.0,
      "step": 23170
    },
    {
      "entropy": 5.424104642868042,
      "epoch": 2.3000198491464867,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004473265918791749,
      "loss": 5.1498,
      "mean_token_accuracy": 0.16616376638412475,
      "num_tokens": 50768981.0,
      "step": 23175
    },
    {
      "entropy": 5.4353374481201175,
      "epoch": 2.300516077808654,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004473038085477682,
      "loss": 5.0602,
      "mean_token_accuracy": 0.17790271043777467,
      "num_tokens": 50780111.0,
      "step": 23180
    },
    {
      "entropy": 5.415660333633423,
      "epoch": 2.3010123064708217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044728102094359826,
      "loss": 5.0871,
      "mean_token_accuracy": 0.16918237805366515,
      "num_tokens": 50790318.0,
      "step": 23185
    },
    {
      "entropy": 5.402129793167115,
      "epoch": 2.301508535132989,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004472582290672303,
      "loss": 5.1213,
      "mean_token_accuracy": 0.17149370312690734,
      "num_tokens": 50800734.0,
      "step": 23190
    },
    {
      "entropy": 5.280769205093383,
      "epoch": 2.3020047637951566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044723543291922936,
      "loss": 5.017,
      "mean_token_accuracy": 0.17996638268232346,
      "num_tokens": 50811236.0,
      "step": 23195
    },
    {
      "entropy": 5.299712228775024,
      "epoch": 2.3025009924573245,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044721263250016086,
      "loss": 5.0345,
      "mean_token_accuracy": 0.17447767704725264,
      "num_tokens": 50821776.0,
      "step": 23200
    },
    {
      "entropy": 5.454625797271729,
      "epoch": 2.302997221119492,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044718982781059017,
      "loss": 5.1595,
      "mean_token_accuracy": 0.16933314949274064,
      "num_tokens": 50834156.0,
      "step": 23205
    },
    {
      "entropy": 5.418417692184448,
      "epoch": 2.3034934497816595,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044716701885108275,
      "loss": 5.0249,
      "mean_token_accuracy": 0.171786330640316,
      "num_tokens": 50845086.0,
      "step": 23210
    },
    {
      "entropy": 5.400807237625122,
      "epoch": 2.303989678443827,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044714420562220423,
      "loss": 5.0491,
      "mean_token_accuracy": 0.17280738204717636,
      "num_tokens": 50858637.0,
      "step": 23215
    },
    {
      "entropy": 5.40351037979126,
      "epoch": 2.3044859071059944,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004471213881245204,
      "loss": 5.1646,
      "mean_token_accuracy": 0.17022796273231505,
      "num_tokens": 50871159.0,
      "step": 23220
    },
    {
      "entropy": 5.3547135353088375,
      "epoch": 2.304982135768162,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000447098566358597,
      "loss": 5.0364,
      "mean_token_accuracy": 0.1739700421690941,
      "num_tokens": 50884072.0,
      "step": 23225
    },
    {
      "entropy": 5.323694658279419,
      "epoch": 2.3054783644303294,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044707574032500005,
      "loss": 5.0419,
      "mean_token_accuracy": 0.17744735777378082,
      "num_tokens": 50896627.0,
      "step": 23230
    },
    {
      "entropy": 5.31010069847107,
      "epoch": 2.305974593092497,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004470529100242955,
      "loss": 4.9592,
      "mean_token_accuracy": 0.1820032238960266,
      "num_tokens": 50907476.0,
      "step": 23235
    },
    {
      "entropy": 5.364790248870849,
      "epoch": 2.3064708217546643,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004470300754570495,
      "loss": 5.0491,
      "mean_token_accuracy": 0.17645647674798964,
      "num_tokens": 50918104.0,
      "step": 23240
    },
    {
      "entropy": 5.390502071380615,
      "epoch": 2.3069670504168323,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004470072366238284,
      "loss": 5.0182,
      "mean_token_accuracy": 0.17591462880373002,
      "num_tokens": 50930575.0,
      "step": 23245
    },
    {
      "entropy": 5.321107387542725,
      "epoch": 2.3074632790789997,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044698439352519836,
      "loss": 4.991,
      "mean_token_accuracy": 0.17214764207601546,
      "num_tokens": 50940592.0,
      "step": 23250
    },
    {
      "entropy": 5.344041728973389,
      "epoch": 2.307959507741167,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044696154616172604,
      "loss": 5.0551,
      "mean_token_accuracy": 0.16830667108297348,
      "num_tokens": 50952134.0,
      "step": 23255
    },
    {
      "entropy": 5.372017860412598,
      "epoch": 2.3084557364033347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004469386945339779,
      "loss": 5.0865,
      "mean_token_accuracy": 0.17594979256391524,
      "num_tokens": 50962721.0,
      "step": 23260
    },
    {
      "entropy": 5.331207895278931,
      "epoch": 2.308951965065502,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004469158386425206,
      "loss": 5.0276,
      "mean_token_accuracy": 0.17250469624996184,
      "num_tokens": 50974893.0,
      "step": 23265
    },
    {
      "entropy": 5.3399497985839846,
      "epoch": 2.3094481937276696,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000446892978487921,
      "loss": 5.0018,
      "mean_token_accuracy": 0.1740655153989792,
      "num_tokens": 50986225.0,
      "step": 23270
    },
    {
      "entropy": 5.35940203666687,
      "epoch": 2.309944422389837,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044687011407074583,
      "loss": 5.0868,
      "mean_token_accuracy": 0.16566449105739595,
      "num_tokens": 50997924.0,
      "step": 23275
    },
    {
      "entropy": 5.390340328216553,
      "epoch": 2.3104406510520046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004468472453915623,
      "loss": 4.9764,
      "mean_token_accuracy": 0.17201390564441682,
      "num_tokens": 51008179.0,
      "step": 23280
    },
    {
      "entropy": 5.359320068359375,
      "epoch": 2.3109368797141725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004468243724509373,
      "loss": 4.9886,
      "mean_token_accuracy": 0.18142406046390533,
      "num_tokens": 51018462.0,
      "step": 23285
    },
    {
      "entropy": 5.372231674194336,
      "epoch": 2.31143310837634,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004468014952494382,
      "loss": 5.0887,
      "mean_token_accuracy": 0.17352184802293777,
      "num_tokens": 51029769.0,
      "step": 23290
    },
    {
      "entropy": 5.334834432601928,
      "epoch": 2.3119293370385074,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044677861378763205,
      "loss": 5.0767,
      "mean_token_accuracy": 0.17259311378002168,
      "num_tokens": 51041058.0,
      "step": 23295
    },
    {
      "entropy": 5.342276668548584,
      "epoch": 2.312425565700675,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044675572806608655,
      "loss": 5.0011,
      "mean_token_accuracy": 0.17807672917842865,
      "num_tokens": 51050958.0,
      "step": 23300
    },
    {
      "entropy": 5.390648984909058,
      "epoch": 2.3129217943628424,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044673283808536915,
      "loss": 5.0291,
      "mean_token_accuracy": 0.18216815888881682,
      "num_tokens": 51062459.0,
      "step": 23305
    },
    {
      "entropy": 5.4385683059692385,
      "epoch": 2.31341802302501,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004467099438460474,
      "loss": 5.2191,
      "mean_token_accuracy": 0.16830465495586394,
      "num_tokens": 51074596.0,
      "step": 23310
    },
    {
      "entropy": 5.437622737884522,
      "epoch": 2.3139142516871773,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044668704534868886,
      "loss": 5.0789,
      "mean_token_accuracy": 0.16864141076803207,
      "num_tokens": 51085265.0,
      "step": 23315
    },
    {
      "entropy": 5.412437677383423,
      "epoch": 2.314410480349345,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044666414259386173,
      "loss": 5.0118,
      "mean_token_accuracy": 0.17106833904981614,
      "num_tokens": 51095986.0,
      "step": 23320
    },
    {
      "entropy": 5.372608757019043,
      "epoch": 2.3149067090115123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044664123558213367,
      "loss": 5.0291,
      "mean_token_accuracy": 0.17252686470746995,
      "num_tokens": 51107518.0,
      "step": 23325
    },
    {
      "entropy": 5.341517448425293,
      "epoch": 2.31540293767368,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004466183243140729,
      "loss": 4.9905,
      "mean_token_accuracy": 0.1793963760137558,
      "num_tokens": 51119715.0,
      "step": 23330
    },
    {
      "entropy": 5.27083911895752,
      "epoch": 2.3158991663358477,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044659540879024746,
      "loss": 5.0058,
      "mean_token_accuracy": 0.17380531132221222,
      "num_tokens": 51130801.0,
      "step": 23335
    },
    {
      "entropy": 5.327224254608154,
      "epoch": 2.316395394998015,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004465724890112256,
      "loss": 5.0177,
      "mean_token_accuracy": 0.1759017065167427,
      "num_tokens": 51142509.0,
      "step": 23340
    },
    {
      "entropy": 5.437225389480591,
      "epoch": 2.3168916236601826,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044654956497757576,
      "loss": 5.0726,
      "mean_token_accuracy": 0.1747567355632782,
      "num_tokens": 51154839.0,
      "step": 23345
    },
    {
      "entropy": 5.355621767044068,
      "epoch": 2.31738785232235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044652663668986634,
      "loss": 5.0121,
      "mean_token_accuracy": 0.1788795605301857,
      "num_tokens": 51165293.0,
      "step": 23350
    },
    {
      "entropy": 5.3402941703796385,
      "epoch": 2.3178840809845176,
      "grad_norm": 1.0,
      "learning_rate": 0.000446503704148666,
      "loss": 5.0962,
      "mean_token_accuracy": 0.16996345669031143,
      "num_tokens": 51176371.0,
      "step": 23355
    },
    {
      "entropy": 5.376697635650634,
      "epoch": 2.318380309646685,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004464807673545433,
      "loss": 5.0187,
      "mean_token_accuracy": 0.1823025643825531,
      "num_tokens": 51187240.0,
      "step": 23360
    },
    {
      "entropy": 5.427601861953735,
      "epoch": 2.3188765383088525,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004464578263080671,
      "loss": 5.1066,
      "mean_token_accuracy": 0.17334885597229005,
      "num_tokens": 51199084.0,
      "step": 23365
    },
    {
      "entropy": 5.325928354263306,
      "epoch": 2.3193727669710205,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044643488100980623,
      "loss": 4.9706,
      "mean_token_accuracy": 0.176742585003376,
      "num_tokens": 51209885.0,
      "step": 23370
    },
    {
      "entropy": 5.3756695747375485,
      "epoch": 2.319868995633188,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004464119314603298,
      "loss": 5.0348,
      "mean_token_accuracy": 0.1790466919541359,
      "num_tokens": 51220644.0,
      "step": 23375
    },
    {
      "entropy": 5.351066446304321,
      "epoch": 2.3203652242953554,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044638897766020673,
      "loss": 5.0025,
      "mean_token_accuracy": 0.1762324631214142,
      "num_tokens": 51231342.0,
      "step": 23380
    },
    {
      "entropy": 5.2829790115356445,
      "epoch": 2.320861452957523,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004463660196100064,
      "loss": 4.9175,
      "mean_token_accuracy": 0.18526383340358735,
      "num_tokens": 51241368.0,
      "step": 23385
    },
    {
      "entropy": 5.347844648361206,
      "epoch": 2.3213576816196904,
      "grad_norm": 1.0,
      "learning_rate": 0.00044634305731029803,
      "loss": 5.1182,
      "mean_token_accuracy": 0.17234142273664474,
      "num_tokens": 51252784.0,
      "step": 23390
    },
    {
      "entropy": 5.396998500823974,
      "epoch": 2.321853910281858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044632009076165096,
      "loss": 4.9939,
      "mean_token_accuracy": 0.1772867888212204,
      "num_tokens": 51264246.0,
      "step": 23395
    },
    {
      "entropy": 5.288774156570435,
      "epoch": 2.3223501389440253,
      "grad_norm": 1.046875,
      "learning_rate": 0.000446297119964635,
      "loss": 4.9622,
      "mean_token_accuracy": 0.1813057005405426,
      "num_tokens": 51274575.0,
      "step": 23400
    },
    {
      "entropy": 5.32437539100647,
      "epoch": 2.322846367606193,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004462741449198194,
      "loss": 4.9902,
      "mean_token_accuracy": 0.1716400906443596,
      "num_tokens": 51285384.0,
      "step": 23405
    },
    {
      "entropy": 5.403724622726441,
      "epoch": 2.3233425962683603,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044625116562777413,
      "loss": 5.0859,
      "mean_token_accuracy": 0.1793188899755478,
      "num_tokens": 51297983.0,
      "step": 23410
    },
    {
      "entropy": 5.289322328567505,
      "epoch": 2.323838824930528,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004462281820890689,
      "loss": 4.9796,
      "mean_token_accuracy": 0.18460406363010406,
      "num_tokens": 51308795.0,
      "step": 23415
    },
    {
      "entropy": 5.3109523296356205,
      "epoch": 2.3243350535926957,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044620519430427376,
      "loss": 4.9965,
      "mean_token_accuracy": 0.1815990537405014,
      "num_tokens": 51318597.0,
      "step": 23420
    },
    {
      "entropy": 5.317487812042236,
      "epoch": 2.324831282254863,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004461822022739588,
      "loss": 4.9467,
      "mean_token_accuracy": 0.17848799973726273,
      "num_tokens": 51329238.0,
      "step": 23425
    },
    {
      "entropy": 5.382061100006103,
      "epoch": 2.3253275109170306,
      "grad_norm": 1.015625,
      "learning_rate": 0.000446159205998694,
      "loss": 5.0562,
      "mean_token_accuracy": 0.17763799279928208,
      "num_tokens": 51340295.0,
      "step": 23430
    },
    {
      "entropy": 5.319087123870849,
      "epoch": 2.325823739579198,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004461362054790497,
      "loss": 5.0237,
      "mean_token_accuracy": 0.17462182343006133,
      "num_tokens": 51350702.0,
      "step": 23435
    },
    {
      "entropy": 5.355668306350708,
      "epoch": 2.3263199682413656,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044611320071559637,
      "loss": 5.0354,
      "mean_token_accuracy": 0.17548919469118118,
      "num_tokens": 51359794.0,
      "step": 23440
    },
    {
      "entropy": 5.358128356933594,
      "epoch": 2.326816196903533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044609019170890424,
      "loss": 4.9987,
      "mean_token_accuracy": 0.17658048421144484,
      "num_tokens": 51368932.0,
      "step": 23445
    },
    {
      "entropy": 5.372833347320556,
      "epoch": 2.3273124255657005,
      "grad_norm": 1.0,
      "learning_rate": 0.00044606717845954415,
      "loss": 5.0693,
      "mean_token_accuracy": 0.176838718354702,
      "num_tokens": 51381111.0,
      "step": 23450
    },
    {
      "entropy": 5.308470821380615,
      "epoch": 2.3278086542278684,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044604416096808656,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18671648055315018,
      "num_tokens": 51392976.0,
      "step": 23455
    },
    {
      "entropy": 5.344123697280883,
      "epoch": 2.328304882890036,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044602113923510235,
      "loss": 4.9836,
      "mean_token_accuracy": 0.17604075819253923,
      "num_tokens": 51403586.0,
      "step": 23460
    },
    {
      "entropy": 5.3812243938446045,
      "epoch": 2.3288011115522034,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044599811326116246,
      "loss": 5.0658,
      "mean_token_accuracy": 0.17234191298484802,
      "num_tokens": 51414035.0,
      "step": 23465
    },
    {
      "entropy": 5.399755859375,
      "epoch": 2.329297340214371,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044597508304683784,
      "loss": 5.0565,
      "mean_token_accuracy": 0.1728152737021446,
      "num_tokens": 51424111.0,
      "step": 23470
    },
    {
      "entropy": 5.33221697807312,
      "epoch": 2.3297935688765383,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004459520485926995,
      "loss": 4.9916,
      "mean_token_accuracy": 0.17953828126192092,
      "num_tokens": 51434566.0,
      "step": 23475
    },
    {
      "entropy": 5.361140871047974,
      "epoch": 2.330289797538706,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004459290098993188,
      "loss": 5.0397,
      "mean_token_accuracy": 0.1727583572268486,
      "num_tokens": 51445568.0,
      "step": 23480
    },
    {
      "entropy": 5.386128282546997,
      "epoch": 2.3307860262008733,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044590596696726686,
      "loss": 5.0245,
      "mean_token_accuracy": 0.1807289808988571,
      "num_tokens": 51456482.0,
      "step": 23485
    },
    {
      "entropy": 5.39080777168274,
      "epoch": 2.3312822548630407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004458829197971153,
      "loss": 5.1132,
      "mean_token_accuracy": 0.16859484612941741,
      "num_tokens": 51467809.0,
      "step": 23490
    },
    {
      "entropy": 5.40198278427124,
      "epoch": 2.3317784835252082,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044585986838943556,
      "loss": 5.0495,
      "mean_token_accuracy": 0.1704128324985504,
      "num_tokens": 51477927.0,
      "step": 23495
    },
    {
      "entropy": 5.332223415374756,
      "epoch": 2.3322747121873757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044583681274479927,
      "loss": 5.0695,
      "mean_token_accuracy": 0.17532191723585128,
      "num_tokens": 51488432.0,
      "step": 23500
    },
    {
      "entropy": 5.397152090072632,
      "epoch": 2.3327709408495436,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004458137528637781,
      "loss": 5.0332,
      "mean_token_accuracy": 0.17143319994211198,
      "num_tokens": 51499149.0,
      "step": 23505
    },
    {
      "entropy": 5.449923181533814,
      "epoch": 2.333267169511711,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044579068874694386,
      "loss": 5.0721,
      "mean_token_accuracy": 0.16955820322036744,
      "num_tokens": 51510672.0,
      "step": 23510
    },
    {
      "entropy": 5.3625044345855715,
      "epoch": 2.3337633981738786,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004457676203948687,
      "loss": 5.069,
      "mean_token_accuracy": 0.17191506922245026,
      "num_tokens": 51521855.0,
      "step": 23515
    },
    {
      "entropy": 5.400622177124023,
      "epoch": 2.334259626836046,
      "grad_norm": 1.0,
      "learning_rate": 0.0004457445478081244,
      "loss": 5.0334,
      "mean_token_accuracy": 0.17523660957813264,
      "num_tokens": 51531970.0,
      "step": 23520
    },
    {
      "entropy": 5.376819038391114,
      "epoch": 2.3347558554982135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044572147098728335,
      "loss": 5.1077,
      "mean_token_accuracy": 0.17091049998998642,
      "num_tokens": 51543110.0,
      "step": 23525
    },
    {
      "entropy": 5.327816009521484,
      "epoch": 2.335252084160381,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044569838993291757,
      "loss": 5.0477,
      "mean_token_accuracy": 0.16993411183357238,
      "num_tokens": 51552991.0,
      "step": 23530
    },
    {
      "entropy": 5.424234533309937,
      "epoch": 2.3357483128225485,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044567530464559963,
      "loss": 5.0788,
      "mean_token_accuracy": 0.17535715103149413,
      "num_tokens": 51564472.0,
      "step": 23535
    },
    {
      "entropy": 5.3467755794525145,
      "epoch": 2.3362445414847164,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004456522151259019,
      "loss": 5.0952,
      "mean_token_accuracy": 0.17473195791244506,
      "num_tokens": 51576051.0,
      "step": 23540
    },
    {
      "entropy": 5.433764982223511,
      "epoch": 2.336740770146884,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000445629121374397,
      "loss": 5.0757,
      "mean_token_accuracy": 0.16967634707689286,
      "num_tokens": 51586658.0,
      "step": 23545
    },
    {
      "entropy": 5.330868101119995,
      "epoch": 2.3372369988090513,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044560602339165746,
      "loss": 5.0415,
      "mean_token_accuracy": 0.16911462396383287,
      "num_tokens": 51596614.0,
      "step": 23550
    },
    {
      "entropy": 5.327672052383423,
      "epoch": 2.337733227471219,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004455829211782562,
      "loss": 5.0567,
      "mean_token_accuracy": 0.17289797961711884,
      "num_tokens": 51607102.0,
      "step": 23555
    },
    {
      "entropy": 5.448483324050903,
      "epoch": 2.3382294561333863,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004455598147347661,
      "loss": 5.1007,
      "mean_token_accuracy": 0.17174533605575562,
      "num_tokens": 51619079.0,
      "step": 23560
    },
    {
      "entropy": 5.352362251281738,
      "epoch": 2.3387256847955538,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004455367040617601,
      "loss": 4.9745,
      "mean_token_accuracy": 0.1816625103354454,
      "num_tokens": 51630273.0,
      "step": 23565
    },
    {
      "entropy": 5.297429370880127,
      "epoch": 2.3392219134577212,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044551358915981135,
      "loss": 5.0136,
      "mean_token_accuracy": 0.1846915602684021,
      "num_tokens": 51640480.0,
      "step": 23570
    },
    {
      "entropy": 5.347823143005371,
      "epoch": 2.3397181421198887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004454904700294929,
      "loss": 5.078,
      "mean_token_accuracy": 0.1742543488740921,
      "num_tokens": 51651831.0,
      "step": 23575
    },
    {
      "entropy": 5.384792804718018,
      "epoch": 2.340214370782056,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004454673466713783,
      "loss": 5.0761,
      "mean_token_accuracy": 0.17434240728616715,
      "num_tokens": 51664061.0,
      "step": 23580
    },
    {
      "entropy": 5.363372182846069,
      "epoch": 2.3407105994442237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044544421908604073,
      "loss": 5.0355,
      "mean_token_accuracy": 0.1753165230154991,
      "num_tokens": 51675081.0,
      "step": 23585
    },
    {
      "entropy": 5.329639434814453,
      "epoch": 2.3412068281063916,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044542108727405386,
      "loss": 4.939,
      "mean_token_accuracy": 0.1851007267832756,
      "num_tokens": 51684477.0,
      "step": 23590
    },
    {
      "entropy": 5.444061517715454,
      "epoch": 2.341703056768559,
      "grad_norm": 1.125,
      "learning_rate": 0.00044539795123599117,
      "loss": 5.0897,
      "mean_token_accuracy": 0.1807327851653099,
      "num_tokens": 51694420.0,
      "step": 23595
    },
    {
      "entropy": 5.2159041404724125,
      "epoch": 2.3421992854307265,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044537481097242654,
      "loss": 4.9346,
      "mean_token_accuracy": 0.18810392916202545,
      "num_tokens": 51704635.0,
      "step": 23600
    },
    {
      "entropy": 5.422498750686645,
      "epoch": 2.342695514092894,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044535166648393377,
      "loss": 5.1466,
      "mean_token_accuracy": 0.1702172115445137,
      "num_tokens": 51716149.0,
      "step": 23605
    },
    {
      "entropy": 5.334575748443603,
      "epoch": 2.3431917427550615,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044532851777108663,
      "loss": 4.9884,
      "mean_token_accuracy": 0.17818379402160645,
      "num_tokens": 51726861.0,
      "step": 23610
    },
    {
      "entropy": 5.422310400009155,
      "epoch": 2.343687971417229,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004453053648344593,
      "loss": 5.2368,
      "mean_token_accuracy": 0.16424040347337723,
      "num_tokens": 51738111.0,
      "step": 23615
    },
    {
      "entropy": 5.359122848510742,
      "epoch": 2.3441842000793964,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044528220767462595,
      "loss": 5.0465,
      "mean_token_accuracy": 0.1734995275735855,
      "num_tokens": 51749004.0,
      "step": 23620
    },
    {
      "entropy": 5.340857648849488,
      "epoch": 2.3446804287415643,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004452590462921607,
      "loss": 5.0286,
      "mean_token_accuracy": 0.17002685368061066,
      "num_tokens": 51759557.0,
      "step": 23625
    },
    {
      "entropy": 5.4176593780517575,
      "epoch": 2.345176657403732,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044523588068763804,
      "loss": 5.0903,
      "mean_token_accuracy": 0.1738931804895401,
      "num_tokens": 51769921.0,
      "step": 23630
    },
    {
      "entropy": 5.321840524673462,
      "epoch": 2.3456728860658993,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004452127108616324,
      "loss": 4.9461,
      "mean_token_accuracy": 0.18164153546094894,
      "num_tokens": 51780158.0,
      "step": 23635
    },
    {
      "entropy": 5.410857105255127,
      "epoch": 2.3461691147280668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044518953681471824,
      "loss": 5.0614,
      "mean_token_accuracy": 0.17468585669994355,
      "num_tokens": 51790502.0,
      "step": 23640
    },
    {
      "entropy": 5.359015321731567,
      "epoch": 2.3466653433902342,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004451663585474703,
      "loss": 5.0727,
      "mean_token_accuracy": 0.1686547055840492,
      "num_tokens": 51801420.0,
      "step": 23645
    },
    {
      "entropy": 5.388235139846802,
      "epoch": 2.3471615720524017,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044514317606046336,
      "loss": 5.0723,
      "mean_token_accuracy": 0.18045054823160173,
      "num_tokens": 51812399.0,
      "step": 23650
    },
    {
      "entropy": 5.42985463142395,
      "epoch": 2.347657800714569,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004451199893542723,
      "loss": 5.1254,
      "mean_token_accuracy": 0.1686950922012329,
      "num_tokens": 51824268.0,
      "step": 23655
    },
    {
      "entropy": 5.324703407287598,
      "epoch": 2.3481540293767367,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004450967984294721,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17275310158729554,
      "num_tokens": 51835842.0,
      "step": 23660
    },
    {
      "entropy": 5.32946367263794,
      "epoch": 2.348650258038904,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004450736032866377,
      "loss": 4.9815,
      "mean_token_accuracy": 0.180229115486145,
      "num_tokens": 51847501.0,
      "step": 23665
    },
    {
      "entropy": 5.3565572738647464,
      "epoch": 2.3491464867010716,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044505040392634447,
      "loss": 5.0525,
      "mean_token_accuracy": 0.17931318730115892,
      "num_tokens": 51858056.0,
      "step": 23670
    },
    {
      "entropy": 5.404448175430298,
      "epoch": 2.3496427153632395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004450272003491677,
      "loss": 5.0078,
      "mean_token_accuracy": 0.18393402695655822,
      "num_tokens": 51869248.0,
      "step": 23675
    },
    {
      "entropy": 5.371134090423584,
      "epoch": 2.350138944025407,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044500399255568274,
      "loss": 5.0906,
      "mean_token_accuracy": 0.17074304521083833,
      "num_tokens": 51881496.0,
      "step": 23680
    },
    {
      "entropy": 5.314603567123413,
      "epoch": 2.3506351726875745,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044498078054646507,
      "loss": 5.0473,
      "mean_token_accuracy": 0.17386519461870192,
      "num_tokens": 51891851.0,
      "step": 23685
    },
    {
      "entropy": 5.410032320022583,
      "epoch": 2.351131401349742,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044495756432209025,
      "loss": 5.1278,
      "mean_token_accuracy": 0.169502155482769,
      "num_tokens": 51903581.0,
      "step": 23690
    },
    {
      "entropy": 5.336469841003418,
      "epoch": 2.3516276300119094,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044493434388313423,
      "loss": 4.9618,
      "mean_token_accuracy": 0.1791615977883339,
      "num_tokens": 51913650.0,
      "step": 23695
    },
    {
      "entropy": 5.298092460632324,
      "epoch": 2.352123858674077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044491111923017254,
      "loss": 4.972,
      "mean_token_accuracy": 0.1813747838139534,
      "num_tokens": 51924704.0,
      "step": 23700
    },
    {
      "entropy": 5.40219521522522,
      "epoch": 2.3526200873362444,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044488789036378117,
      "loss": 5.1199,
      "mean_token_accuracy": 0.1732394739985466,
      "num_tokens": 51937576.0,
      "step": 23705
    },
    {
      "entropy": 5.33323106765747,
      "epoch": 2.3531163159984123,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004448646572845363,
      "loss": 4.9473,
      "mean_token_accuracy": 0.17488353848457336,
      "num_tokens": 51947359.0,
      "step": 23710
    },
    {
      "entropy": 5.368526840209961,
      "epoch": 2.35361254466058,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004448414199930139,
      "loss": 5.0946,
      "mean_token_accuracy": 0.17046739757061005,
      "num_tokens": 51958016.0,
      "step": 23715
    },
    {
      "entropy": 5.446837282180786,
      "epoch": 2.3541087733227473,
      "grad_norm": 1.0,
      "learning_rate": 0.00044481817848979026,
      "loss": 5.0731,
      "mean_token_accuracy": 0.17131375074386596,
      "num_tokens": 51968770.0,
      "step": 23720
    },
    {
      "entropy": 5.357419443130493,
      "epoch": 2.3546050019849147,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004447949327754418,
      "loss": 4.9659,
      "mean_token_accuracy": 0.18314096033573152,
      "num_tokens": 51980052.0,
      "step": 23725
    },
    {
      "entropy": 5.391411304473877,
      "epoch": 2.355101230647082,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004447716828505447,
      "loss": 5.1259,
      "mean_token_accuracy": 0.17216980308294297,
      "num_tokens": 51992080.0,
      "step": 23730
    },
    {
      "entropy": 5.360816907882691,
      "epoch": 2.3555974593092497,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004447484287156758,
      "loss": 5.0694,
      "mean_token_accuracy": 0.1798478826880455,
      "num_tokens": 52004313.0,
      "step": 23735
    },
    {
      "entropy": 5.415277194976807,
      "epoch": 2.356093687971417,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004447251703714117,
      "loss": 5.0392,
      "mean_token_accuracy": 0.17299067229032516,
      "num_tokens": 52015207.0,
      "step": 23740
    },
    {
      "entropy": 5.365356540679931,
      "epoch": 2.3565899166335846,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000444701907818329,
      "loss": 5.0379,
      "mean_token_accuracy": 0.1746157482266426,
      "num_tokens": 52026565.0,
      "step": 23745
    },
    {
      "entropy": 5.38454418182373,
      "epoch": 2.357086145295752,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004446786410570048,
      "loss": 5.073,
      "mean_token_accuracy": 0.17074529081583023,
      "num_tokens": 52037654.0,
      "step": 23750
    },
    {
      "entropy": 5.346745491027832,
      "epoch": 2.3575823739579196,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044465537008801575,
      "loss": 4.9689,
      "mean_token_accuracy": 0.17774230986833572,
      "num_tokens": 52049772.0,
      "step": 23755
    },
    {
      "entropy": 5.3428267478942875,
      "epoch": 2.3580786026200875,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004446320949119392,
      "loss": 5.1125,
      "mean_token_accuracy": 0.1711305409669876,
      "num_tokens": 52061363.0,
      "step": 23760
    },
    {
      "entropy": 5.3390154361724855,
      "epoch": 2.358574831282255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044460881552935215,
      "loss": 4.977,
      "mean_token_accuracy": 0.1740237295627594,
      "num_tokens": 52073153.0,
      "step": 23765
    },
    {
      "entropy": 5.339682483673096,
      "epoch": 2.3590710599444225,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044458553194083204,
      "loss": 5.0013,
      "mean_token_accuracy": 0.17778213322162628,
      "num_tokens": 52083369.0,
      "step": 23770
    },
    {
      "entropy": 5.299059247970581,
      "epoch": 2.35956728860659,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004445622441469562,
      "loss": 4.9568,
      "mean_token_accuracy": 0.18288030922412873,
      "num_tokens": 52093701.0,
      "step": 23775
    },
    {
      "entropy": 5.331163120269776,
      "epoch": 2.3600635172687574,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044453895214830195,
      "loss": 5.0153,
      "mean_token_accuracy": 0.17605394273996353,
      "num_tokens": 52105863.0,
      "step": 23780
    },
    {
      "entropy": 5.291473865509033,
      "epoch": 2.360559745930925,
      "grad_norm": 1.0,
      "learning_rate": 0.0004445156559454471,
      "loss": 5.0104,
      "mean_token_accuracy": 0.1761175125837326,
      "num_tokens": 52116476.0,
      "step": 23785
    },
    {
      "entropy": 5.444654321670532,
      "epoch": 2.3610559745930924,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044449235553896914,
      "loss": 5.1241,
      "mean_token_accuracy": 0.1682300314307213,
      "num_tokens": 52127262.0,
      "step": 23790
    },
    {
      "entropy": 5.45071120262146,
      "epoch": 2.36155220325526,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004444690509294461,
      "loss": 5.0677,
      "mean_token_accuracy": 0.17392871975898744,
      "num_tokens": 52138575.0,
      "step": 23795
    },
    {
      "entropy": 5.338071489334107,
      "epoch": 2.3620484319174277,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004444457421174558,
      "loss": 5.0334,
      "mean_token_accuracy": 0.17530251741409303,
      "num_tokens": 52149620.0,
      "step": 23800
    },
    {
      "entropy": 5.421306800842285,
      "epoch": 2.3625446605795952,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004444224291035761,
      "loss": 5.0747,
      "mean_token_accuracy": 0.1716116636991501,
      "num_tokens": 52161929.0,
      "step": 23805
    },
    {
      "entropy": 5.285071802139282,
      "epoch": 2.3630408892417627,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044439911188838533,
      "loss": 4.9801,
      "mean_token_accuracy": 0.1847517266869545,
      "num_tokens": 52172283.0,
      "step": 23810
    },
    {
      "entropy": 5.3181108951568605,
      "epoch": 2.36353711790393,
      "grad_norm": 1.0,
      "learning_rate": 0.0004443757904724615,
      "loss": 5.0908,
      "mean_token_accuracy": 0.17562458366155625,
      "num_tokens": 52184130.0,
      "step": 23815
    },
    {
      "entropy": 5.35390796661377,
      "epoch": 2.3640333465660976,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00044435246485638317,
      "loss": 5.0647,
      "mean_token_accuracy": 0.17732380479574203,
      "num_tokens": 52195358.0,
      "step": 23820
    },
    {
      "entropy": 5.335308790206909,
      "epoch": 2.364529575228265,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004443291350407285,
      "loss": 4.9686,
      "mean_token_accuracy": 0.17837252765893935,
      "num_tokens": 52207070.0,
      "step": 23825
    },
    {
      "entropy": 5.376240158081055,
      "epoch": 2.3650258038904326,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004443058010260762,
      "loss": 5.1116,
      "mean_token_accuracy": 0.16996528804302216,
      "num_tokens": 52217658.0,
      "step": 23830
    },
    {
      "entropy": 5.339537286758423,
      "epoch": 2.3655220325526,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044428246281300493,
      "loss": 4.994,
      "mean_token_accuracy": 0.1763981446623802,
      "num_tokens": 52229350.0,
      "step": 23835
    },
    {
      "entropy": 5.45172348022461,
      "epoch": 2.3660182612147675,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044425912040209326,
      "loss": 5.1631,
      "mean_token_accuracy": 0.17092523574829102,
      "num_tokens": 52241260.0,
      "step": 23840
    },
    {
      "entropy": 5.449008798599243,
      "epoch": 2.3665144898769355,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044423577379392023,
      "loss": 5.057,
      "mean_token_accuracy": 0.17484749257564544,
      "num_tokens": 52253049.0,
      "step": 23845
    },
    {
      "entropy": 5.340025615692139,
      "epoch": 2.367010718539103,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004442124229890646,
      "loss": 5.0399,
      "mean_token_accuracy": 0.17323856055736542,
      "num_tokens": 52263405.0,
      "step": 23850
    },
    {
      "entropy": 5.34362382888794,
      "epoch": 2.3675069472012704,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044418906798810554,
      "loss": 5.0501,
      "mean_token_accuracy": 0.1738727271556854,
      "num_tokens": 52273075.0,
      "step": 23855
    },
    {
      "entropy": 5.508619260787964,
      "epoch": 2.368003175863438,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044416570879162213,
      "loss": 5.2709,
      "mean_token_accuracy": 0.16505997776985168,
      "num_tokens": 52284275.0,
      "step": 23860
    },
    {
      "entropy": 5.393843650817871,
      "epoch": 2.3684994045256054,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004441423454001936,
      "loss": 5.0557,
      "mean_token_accuracy": 0.16884201020002365,
      "num_tokens": 52295084.0,
      "step": 23865
    },
    {
      "entropy": 5.343979835510254,
      "epoch": 2.368995633187773,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044411897781439946,
      "loss": 5.02,
      "mean_token_accuracy": 0.17700401693582535,
      "num_tokens": 52304172.0,
      "step": 23870
    },
    {
      "entropy": 5.267445421218872,
      "epoch": 2.3694918618499403,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004440956060348191,
      "loss": 4.9656,
      "mean_token_accuracy": 0.18484351187944412,
      "num_tokens": 52315370.0,
      "step": 23875
    },
    {
      "entropy": 5.368021488189697,
      "epoch": 2.369988090512108,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004440722300620321,
      "loss": 5.0349,
      "mean_token_accuracy": 0.17535640597343444,
      "num_tokens": 52326517.0,
      "step": 23880
    },
    {
      "entropy": 5.410694789886475,
      "epoch": 2.3704843191742757,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044404884989661804,
      "loss": 5.0912,
      "mean_token_accuracy": 0.17615650296211244,
      "num_tokens": 52337110.0,
      "step": 23885
    },
    {
      "entropy": 5.355504035949707,
      "epoch": 2.370980547836443,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004440254655391568,
      "loss": 4.9957,
      "mean_token_accuracy": 0.1815697133541107,
      "num_tokens": 52347509.0,
      "step": 23890
    },
    {
      "entropy": 5.486967325210571,
      "epoch": 2.3714767764986107,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004440020769902282,
      "loss": 5.1324,
      "mean_token_accuracy": 0.17153545022010802,
      "num_tokens": 52357920.0,
      "step": 23895
    },
    {
      "entropy": 5.392173528671265,
      "epoch": 2.371973005160778,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044397868425041234,
      "loss": 5.0989,
      "mean_token_accuracy": 0.17058905959129333,
      "num_tokens": 52368138.0,
      "step": 23900
    },
    {
      "entropy": 5.42748327255249,
      "epoch": 2.3724692338229456,
      "grad_norm": 1.0,
      "learning_rate": 0.0004439552873202891,
      "loss": 5.0361,
      "mean_token_accuracy": 0.17769884765148164,
      "num_tokens": 52380503.0,
      "step": 23905
    },
    {
      "entropy": 5.344425249099731,
      "epoch": 2.372965462485113,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044393188620043893,
      "loss": 5.0479,
      "mean_token_accuracy": 0.17245973348617555,
      "num_tokens": 52390850.0,
      "step": 23910
    },
    {
      "entropy": 5.274942111968994,
      "epoch": 2.3734616911472806,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004439084808914419,
      "loss": 4.9837,
      "mean_token_accuracy": 0.18164736777544022,
      "num_tokens": 52400829.0,
      "step": 23915
    },
    {
      "entropy": 5.447281408309936,
      "epoch": 2.373957919809448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044388507139387864,
      "loss": 5.0862,
      "mean_token_accuracy": 0.17327000945806503,
      "num_tokens": 52412892.0,
      "step": 23920
    },
    {
      "entropy": 5.358316373825073,
      "epoch": 2.3744541484716155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004438616577083293,
      "loss": 5.0106,
      "mean_token_accuracy": 0.1787156045436859,
      "num_tokens": 52423662.0,
      "step": 23925
    },
    {
      "entropy": 5.362601709365845,
      "epoch": 2.3749503771337834,
      "grad_norm": 0.984375,
      "learning_rate": 0.000443838239835375,
      "loss": 5.0498,
      "mean_token_accuracy": 0.17649957090616225,
      "num_tokens": 52433541.0,
      "step": 23930
    },
    {
      "entropy": 5.387660360336303,
      "epoch": 2.375446605795951,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004438148177755959,
      "loss": 5.1014,
      "mean_token_accuracy": 0.1704227238893509,
      "num_tokens": 52445504.0,
      "step": 23935
    },
    {
      "entropy": 5.411738920211792,
      "epoch": 2.3759428344581184,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044379139152957327,
      "loss": 5.0691,
      "mean_token_accuracy": 0.16209717839956284,
      "num_tokens": 52456116.0,
      "step": 23940
    },
    {
      "entropy": 5.342976093292236,
      "epoch": 2.376439063120286,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044376796109788777,
      "loss": 4.9401,
      "mean_token_accuracy": 0.1823066994547844,
      "num_tokens": 52465462.0,
      "step": 23945
    },
    {
      "entropy": 5.339077758789062,
      "epoch": 2.3769352917824533,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044374452648112046,
      "loss": 5.1059,
      "mean_token_accuracy": 0.17331092953681945,
      "num_tokens": 52477213.0,
      "step": 23950
    },
    {
      "entropy": 5.392141103744507,
      "epoch": 2.377431520444621,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044372108767985253,
      "loss": 5.0887,
      "mean_token_accuracy": 0.1689410015940666,
      "num_tokens": 52488095.0,
      "step": 23955
    },
    {
      "entropy": 5.421023559570313,
      "epoch": 2.3779277491067883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004436976446946652,
      "loss": 5.1466,
      "mean_token_accuracy": 0.16544752568006516,
      "num_tokens": 52499078.0,
      "step": 23960
    },
    {
      "entropy": 5.4020500659942625,
      "epoch": 2.3784239777689558,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044367419752613983,
      "loss": 5.0257,
      "mean_token_accuracy": 0.17453015148639678,
      "num_tokens": 52510829.0,
      "step": 23965
    },
    {
      "entropy": 5.427231693267823,
      "epoch": 2.3789202064311237,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004436507461748577,
      "loss": 5.0292,
      "mean_token_accuracy": 0.17477159202098846,
      "num_tokens": 52521466.0,
      "step": 23970
    },
    {
      "entropy": 5.385166311264038,
      "epoch": 2.379416435093291,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004436272906414007,
      "loss": 4.9441,
      "mean_token_accuracy": 0.18292285203933717,
      "num_tokens": 52532617.0,
      "step": 23975
    },
    {
      "entropy": 5.260566568374633,
      "epoch": 2.3799126637554586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004436038309263501,
      "loss": 5.0261,
      "mean_token_accuracy": 0.173289854824543,
      "num_tokens": 52543826.0,
      "step": 23980
    },
    {
      "entropy": 5.367899560928345,
      "epoch": 2.380408892417626,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044358036703028773,
      "loss": 5.0654,
      "mean_token_accuracy": 0.17118189632892608,
      "num_tokens": 52556752.0,
      "step": 23985
    },
    {
      "entropy": 5.44351053237915,
      "epoch": 2.3809051210797936,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044355689895379563,
      "loss": 5.1257,
      "mean_token_accuracy": 0.16676561981439592,
      "num_tokens": 52567268.0,
      "step": 23990
    },
    {
      "entropy": 5.363738489151001,
      "epoch": 2.381401349741961,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044353342669745566,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1780358910560608,
      "num_tokens": 52576894.0,
      "step": 23995
    },
    {
      "entropy": 5.338898181915283,
      "epoch": 2.3818975784041285,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044350995026184976,
      "loss": 5.0475,
      "mean_token_accuracy": 0.1743764325976372,
      "num_tokens": 52587434.0,
      "step": 24000
    },
    {
      "epoch": 2.3818975784041285,
      "eval_entropy": 5.224403728807757,
      "eval_loss": 5.205906391143799,
      "eval_mean_token_accuracy": 0.17298515881111912,
      "eval_num_tokens": 52587434.0,
      "eval_runtime": 26.7578,
      "eval_samples_per_second": 1258.286,
      "eval_steps_per_second": 157.3,
      "step": 24000
    },
    {
      "entropy": 5.413251829147339,
      "epoch": 2.382393807066296,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044348646964756036,
      "loss": 5.1178,
      "mean_token_accuracy": 0.17424178570508958,
      "num_tokens": 52599455.0,
      "step": 24005
    },
    {
      "entropy": 5.426706171035766,
      "epoch": 2.3828900357284635,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004434629848551694,
      "loss": 5.0938,
      "mean_token_accuracy": 0.1717309311032295,
      "num_tokens": 52611020.0,
      "step": 24010
    },
    {
      "entropy": 5.354325199127198,
      "epoch": 2.3833862643906314,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004434394958852595,
      "loss": 5.032,
      "mean_token_accuracy": 0.17588514238595962,
      "num_tokens": 52621743.0,
      "step": 24015
    },
    {
      "entropy": 5.383949232101441,
      "epoch": 2.383882493052799,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044341600273841304,
      "loss": 5.0821,
      "mean_token_accuracy": 0.17783569693565368,
      "num_tokens": 52633437.0,
      "step": 24020
    },
    {
      "entropy": 5.312158727645874,
      "epoch": 2.3843787217149663,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044339250541521264,
      "loss": 5.001,
      "mean_token_accuracy": 0.17992278784513474,
      "num_tokens": 52643849.0,
      "step": 24025
    },
    {
      "entropy": 5.476568365097046,
      "epoch": 2.384874950377134,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044336900391624096,
      "loss": 5.1645,
      "mean_token_accuracy": 0.1657537817955017,
      "num_tokens": 52656417.0,
      "step": 24030
    },
    {
      "entropy": 5.3411332130432125,
      "epoch": 2.3853711790393013,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004433454982420808,
      "loss": 4.9676,
      "mean_token_accuracy": 0.17686924040317537,
      "num_tokens": 52668491.0,
      "step": 24035
    },
    {
      "entropy": 5.381711339950561,
      "epoch": 2.3858674077014688,
      "grad_norm": 1.0625,
      "learning_rate": 0.000443321988393315,
      "loss": 5.0946,
      "mean_token_accuracy": 0.16923352628946303,
      "num_tokens": 52679532.0,
      "step": 24040
    },
    {
      "entropy": 5.331835556030273,
      "epoch": 2.3863636363636362,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004432984743705265,
      "loss": 4.9704,
      "mean_token_accuracy": 0.17620761096477508,
      "num_tokens": 52689850.0,
      "step": 24045
    },
    {
      "entropy": 5.27823052406311,
      "epoch": 2.3868598650258037,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044327495617429866,
      "loss": 4.9544,
      "mean_token_accuracy": 0.18177102953195573,
      "num_tokens": 52699529.0,
      "step": 24050
    },
    {
      "entropy": 5.2624626636505125,
      "epoch": 2.3873560936879716,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004432514338052144,
      "loss": 4.9716,
      "mean_token_accuracy": 0.18478393852710723,
      "num_tokens": 52711449.0,
      "step": 24055
    },
    {
      "entropy": 5.442509698867798,
      "epoch": 2.387852322350139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004432279072638571,
      "loss": 5.0847,
      "mean_token_accuracy": 0.16333047300577164,
      "num_tokens": 52721528.0,
      "step": 24060
    },
    {
      "entropy": 5.376565647125244,
      "epoch": 2.3883485510123066,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00044320437655081033,
      "loss": 5.0396,
      "mean_token_accuracy": 0.17544911056756973,
      "num_tokens": 52731367.0,
      "step": 24065
    },
    {
      "entropy": 5.304096937179565,
      "epoch": 2.388844779674474,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044318084166665737,
      "loss": 4.932,
      "mean_token_accuracy": 0.18081146329641343,
      "num_tokens": 52740822.0,
      "step": 24070
    },
    {
      "entropy": 5.3104969501495365,
      "epoch": 2.3893410083366415,
      "grad_norm": 0.96875,
      "learning_rate": 0.000443157302611982,
      "loss": 5.0318,
      "mean_token_accuracy": 0.16794940829277039,
      "num_tokens": 52751609.0,
      "step": 24075
    },
    {
      "entropy": 5.393575191497803,
      "epoch": 2.389837236998809,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004431337593873678,
      "loss": 5.076,
      "mean_token_accuracy": 0.16909282505512238,
      "num_tokens": 52762914.0,
      "step": 24080
    },
    {
      "entropy": 5.485241651535034,
      "epoch": 2.3903334656609765,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004431102119933987,
      "loss": 5.12,
      "mean_token_accuracy": 0.17188985794782638,
      "num_tokens": 52773894.0,
      "step": 24085
    },
    {
      "entropy": 5.315875720977783,
      "epoch": 2.390829694323144,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004430866604306586,
      "loss": 5.0331,
      "mean_token_accuracy": 0.1780285507440567,
      "num_tokens": 52784606.0,
      "step": 24090
    },
    {
      "entropy": 5.343503093719482,
      "epoch": 2.3913259229853114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044306310469973147,
      "loss": 5.0105,
      "mean_token_accuracy": 0.17342295497655869,
      "num_tokens": 52795275.0,
      "step": 24095
    },
    {
      "entropy": 5.43363881111145,
      "epoch": 2.391822151647479,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044303954480120155,
      "loss": 5.0937,
      "mean_token_accuracy": 0.17636725306510925,
      "num_tokens": 52805058.0,
      "step": 24100
    },
    {
      "entropy": 5.381444835662842,
      "epoch": 2.392318380309647,
      "grad_norm": 1.125,
      "learning_rate": 0.00044301598073565295,
      "loss": 5.0323,
      "mean_token_accuracy": 0.1716475710272789,
      "num_tokens": 52815486.0,
      "step": 24105
    },
    {
      "entropy": 5.335614442825317,
      "epoch": 2.3928146089718143,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004429924125036702,
      "loss": 5.0384,
      "mean_token_accuracy": 0.17456250488758088,
      "num_tokens": 52827039.0,
      "step": 24110
    },
    {
      "entropy": 5.363163948059082,
      "epoch": 2.393310837633982,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004429688401058375,
      "loss": 5.0386,
      "mean_token_accuracy": 0.17634951919317246,
      "num_tokens": 52837712.0,
      "step": 24115
    },
    {
      "entropy": 5.333619737625122,
      "epoch": 2.3938070662961493,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004429452635427394,
      "loss": 4.9311,
      "mean_token_accuracy": 0.18443846553564072,
      "num_tokens": 52846321.0,
      "step": 24120
    },
    {
      "entropy": 5.347241449356079,
      "epoch": 2.3943032949583167,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044292168281496086,
      "loss": 4.9773,
      "mean_token_accuracy": 0.18089332431554794,
      "num_tokens": 52856342.0,
      "step": 24125
    },
    {
      "entropy": 5.381012296676635,
      "epoch": 2.394799523620484,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004428980979230863,
      "loss": 5.1282,
      "mean_token_accuracy": 0.17163293212652206,
      "num_tokens": 52866602.0,
      "step": 24130
    },
    {
      "entropy": 5.33086667060852,
      "epoch": 2.3952957522826517,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044287450886770075,
      "loss": 4.9497,
      "mean_token_accuracy": 0.18012239336967467,
      "num_tokens": 52877066.0,
      "step": 24135
    },
    {
      "entropy": 5.315430736541748,
      "epoch": 2.3957919809448196,
      "grad_norm": 1.125,
      "learning_rate": 0.0004428509156493891,
      "loss": 4.985,
      "mean_token_accuracy": 0.17908789813518525,
      "num_tokens": 52887907.0,
      "step": 24140
    },
    {
      "entropy": 5.422709512710571,
      "epoch": 2.396288209606987,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004428273182687364,
      "loss": 5.1291,
      "mean_token_accuracy": 0.1699674144387245,
      "num_tokens": 52899308.0,
      "step": 24145
    },
    {
      "entropy": 5.334198808670044,
      "epoch": 2.3967844382691545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004428037167263279,
      "loss": 4.946,
      "mean_token_accuracy": 0.17962228655815124,
      "num_tokens": 52909127.0,
      "step": 24150
    },
    {
      "entropy": 5.317044973373413,
      "epoch": 2.397280666931322,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044278011102274887,
      "loss": 4.9777,
      "mean_token_accuracy": 0.1823316916823387,
      "num_tokens": 52920503.0,
      "step": 24155
    },
    {
      "entropy": 5.332784795761109,
      "epoch": 2.3977768955934895,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044275650115858445,
      "loss": 5.0113,
      "mean_token_accuracy": 0.17563499957323075,
      "num_tokens": 52931904.0,
      "step": 24160
    },
    {
      "entropy": 5.388986682891845,
      "epoch": 2.398273124255657,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004427328871344204,
      "loss": 5.1591,
      "mean_token_accuracy": 0.16403813511133195,
      "num_tokens": 52945101.0,
      "step": 24165
    },
    {
      "entropy": 5.375605630874634,
      "epoch": 2.3987693529178244,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004427092689508422,
      "loss": 5.0449,
      "mean_token_accuracy": 0.17661650478839874,
      "num_tokens": 52955999.0,
      "step": 24170
    },
    {
      "entropy": 5.364274406433106,
      "epoch": 2.399265581579992,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004426856466084354,
      "loss": 5.0595,
      "mean_token_accuracy": 0.1734430357813835,
      "num_tokens": 52967273.0,
      "step": 24175
    },
    {
      "entropy": 5.427544116973877,
      "epoch": 2.3997618102421594,
      "grad_norm": 1.125,
      "learning_rate": 0.000442662020107786,
      "loss": 5.1242,
      "mean_token_accuracy": 0.1760244369506836,
      "num_tokens": 52977415.0,
      "step": 24180
    },
    {
      "entropy": 5.315626859664917,
      "epoch": 2.400258038904327,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004426383894494797,
      "loss": 5.0082,
      "mean_token_accuracy": 0.1768188863992691,
      "num_tokens": 52990071.0,
      "step": 24185
    },
    {
      "entropy": 5.2807512283325195,
      "epoch": 2.400754267566495,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004426147546341027,
      "loss": 4.9842,
      "mean_token_accuracy": 0.17263754904270173,
      "num_tokens": 52999874.0,
      "step": 24190
    },
    {
      "entropy": 5.432112121582032,
      "epoch": 2.4012504962286623,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044259111566224084,
      "loss": 5.0243,
      "mean_token_accuracy": 0.17143243551254272,
      "num_tokens": 53011349.0,
      "step": 24195
    },
    {
      "entropy": 5.2940356731414795,
      "epoch": 2.4017467248908297,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044256747253448045,
      "loss": 4.9823,
      "mean_token_accuracy": 0.18331675976514816,
      "num_tokens": 53024256.0,
      "step": 24200
    },
    {
      "entropy": 5.3229106903076175,
      "epoch": 2.402242953552997,
      "grad_norm": 1.125,
      "learning_rate": 0.00044254382525140784,
      "loss": 4.9708,
      "mean_token_accuracy": 0.1868906646966934,
      "num_tokens": 53035617.0,
      "step": 24205
    },
    {
      "entropy": 5.347861957550049,
      "epoch": 2.4027391822151647,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044252017381360935,
      "loss": 4.9753,
      "mean_token_accuracy": 0.18326762020587922,
      "num_tokens": 53046219.0,
      "step": 24210
    },
    {
      "entropy": 5.4443124294281,
      "epoch": 2.403235410877332,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004424965182216716,
      "loss": 5.1104,
      "mean_token_accuracy": 0.16891320496797563,
      "num_tokens": 53057114.0,
      "step": 24215
    },
    {
      "entropy": 5.420514249801636,
      "epoch": 2.4037316395394996,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044247285847618105,
      "loss": 5.034,
      "mean_token_accuracy": 0.18013737797737123,
      "num_tokens": 53067481.0,
      "step": 24220
    },
    {
      "entropy": 5.441090488433838,
      "epoch": 2.4042278682016676,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004424491945777245,
      "loss": 5.1507,
      "mean_token_accuracy": 0.17238020151853561,
      "num_tokens": 53078129.0,
      "step": 24225
    },
    {
      "entropy": 5.360264778137207,
      "epoch": 2.404724096863835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044242552652688885,
      "loss": 4.9772,
      "mean_token_accuracy": 0.1774058997631073,
      "num_tokens": 53088919.0,
      "step": 24230
    },
    {
      "entropy": 5.350648832321167,
      "epoch": 2.4052203255260025,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004424018543242608,
      "loss": 5.0545,
      "mean_token_accuracy": 0.18418318182229995,
      "num_tokens": 53099716.0,
      "step": 24235
    },
    {
      "entropy": 5.301181840896606,
      "epoch": 2.40571655418817,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004423781779704275,
      "loss": 4.943,
      "mean_token_accuracy": 0.18180761486291885,
      "num_tokens": 53108798.0,
      "step": 24240
    },
    {
      "entropy": 5.364476871490479,
      "epoch": 2.4062127828503375,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004423544974659761,
      "loss": 5.072,
      "mean_token_accuracy": 0.17334414422512054,
      "num_tokens": 53120655.0,
      "step": 24245
    },
    {
      "entropy": 5.2794920921325685,
      "epoch": 2.406709011512505,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044233081281149373,
      "loss": 4.9118,
      "mean_token_accuracy": 0.1797805517911911,
      "num_tokens": 53131760.0,
      "step": 24250
    },
    {
      "entropy": 5.381933689117432,
      "epoch": 2.4072052401746724,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044230712400756774,
      "loss": 5.0028,
      "mean_token_accuracy": 0.17554878294467927,
      "num_tokens": 53142347.0,
      "step": 24255
    },
    {
      "entropy": 5.353096628189087,
      "epoch": 2.40770146883684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044228343105478564,
      "loss": 5.0302,
      "mean_token_accuracy": 0.173873408138752,
      "num_tokens": 53153678.0,
      "step": 24260
    },
    {
      "entropy": 5.322562503814697,
      "epoch": 2.4081976974990074,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004422597339537349,
      "loss": 5.0859,
      "mean_token_accuracy": 0.1770435690879822,
      "num_tokens": 53165089.0,
      "step": 24265
    },
    {
      "entropy": 5.384848642349243,
      "epoch": 2.408693926161175,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004422360327050032,
      "loss": 5.002,
      "mean_token_accuracy": 0.17055773735046387,
      "num_tokens": 53175614.0,
      "step": 24270
    },
    {
      "entropy": 5.375102138519287,
      "epoch": 2.4091901548233428,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044221232730917823,
      "loss": 4.9706,
      "mean_token_accuracy": 0.1775892198085785,
      "num_tokens": 53186306.0,
      "step": 24275
    },
    {
      "entropy": 5.351121044158935,
      "epoch": 2.4096863834855102,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044218861776684776,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17716164886951447,
      "num_tokens": 53197730.0,
      "step": 24280
    },
    {
      "entropy": 5.3087419986724855,
      "epoch": 2.4101826121476777,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044216490407859995,
      "loss": 4.9599,
      "mean_token_accuracy": 0.18102093935012817,
      "num_tokens": 53208486.0,
      "step": 24285
    },
    {
      "entropy": 5.37816743850708,
      "epoch": 2.410678840809845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044214118624502267,
      "loss": 5.0508,
      "mean_token_accuracy": 0.17507839053869248,
      "num_tokens": 53220045.0,
      "step": 24290
    },
    {
      "entropy": 5.394984006881714,
      "epoch": 2.4111750694720127,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004421174642667042,
      "loss": 5.1335,
      "mean_token_accuracy": 0.17088280767202377,
      "num_tokens": 53231385.0,
      "step": 24295
    },
    {
      "entropy": 5.427643251419068,
      "epoch": 2.41167129813418,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044209373814423265,
      "loss": 5.0817,
      "mean_token_accuracy": 0.17368377298116683,
      "num_tokens": 53243038.0,
      "step": 24300
    },
    {
      "entropy": 5.405091428756714,
      "epoch": 2.4121675267963476,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004420700078781965,
      "loss": 5.059,
      "mean_token_accuracy": 0.16958539485931395,
      "num_tokens": 53252959.0,
      "step": 24305
    },
    {
      "entropy": 5.287730312347412,
      "epoch": 2.4126637554585155,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044204627346918407,
      "loss": 4.9852,
      "mean_token_accuracy": 0.17773352414369584,
      "num_tokens": 53263542.0,
      "step": 24310
    },
    {
      "entropy": 5.357538318634033,
      "epoch": 2.413159984120683,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044202253491778404,
      "loss": 4.993,
      "mean_token_accuracy": 0.18466148674488067,
      "num_tokens": 53273511.0,
      "step": 24315
    },
    {
      "entropy": 5.38882966041565,
      "epoch": 2.4136562127828505,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004419987922245852,
      "loss": 5.1258,
      "mean_token_accuracy": 0.1732481002807617,
      "num_tokens": 53286173.0,
      "step": 24320
    },
    {
      "entropy": 5.362208461761474,
      "epoch": 2.414152441445018,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000441975045390176,
      "loss": 4.9929,
      "mean_token_accuracy": 0.18367340713739394,
      "num_tokens": 53296784.0,
      "step": 24325
    },
    {
      "entropy": 5.406254768371582,
      "epoch": 2.4146486701071854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044195129441514547,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1726016491651535,
      "num_tokens": 53308044.0,
      "step": 24330
    },
    {
      "entropy": 5.326471853256225,
      "epoch": 2.415144898769353,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044192753930008267,
      "loss": 4.9882,
      "mean_token_accuracy": 0.182477405667305,
      "num_tokens": 53319161.0,
      "step": 24335
    },
    {
      "entropy": 5.285283374786377,
      "epoch": 2.4156411274315204,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004419037800455765,
      "loss": 4.9582,
      "mean_token_accuracy": 0.1822102412581444,
      "num_tokens": 53329766.0,
      "step": 24340
    },
    {
      "entropy": 5.385812616348266,
      "epoch": 2.416137356093688,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004418800166522163,
      "loss": 5.0512,
      "mean_token_accuracy": 0.17790469527244568,
      "num_tokens": 53340329.0,
      "step": 24345
    },
    {
      "entropy": 5.458166885375976,
      "epoch": 2.4166335847558553,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044185624912059125,
      "loss": 5.0612,
      "mean_token_accuracy": 0.17449921518564224,
      "num_tokens": 53352592.0,
      "step": 24350
    },
    {
      "entropy": 5.415979480743408,
      "epoch": 2.417129813418023,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004418324774512908,
      "loss": 5.0397,
      "mean_token_accuracy": 0.17331262826919555,
      "num_tokens": 53362877.0,
      "step": 24355
    },
    {
      "entropy": 5.364007043838501,
      "epoch": 2.4176260420801907,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004418087016449043,
      "loss": 5.0428,
      "mean_token_accuracy": 0.17261012345552446,
      "num_tokens": 53374612.0,
      "step": 24360
    },
    {
      "entropy": 5.396334409713745,
      "epoch": 2.418122270742358,
      "grad_norm": 1.0,
      "learning_rate": 0.00044178492170202153,
      "loss": 5.0519,
      "mean_token_accuracy": 0.16974952071905136,
      "num_tokens": 53385129.0,
      "step": 24365
    },
    {
      "entropy": 5.388797092437744,
      "epoch": 2.4186184994045257,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044176113762323213,
      "loss": 5.1213,
      "mean_token_accuracy": 0.16984976083040237,
      "num_tokens": 53395403.0,
      "step": 24370
    },
    {
      "entropy": 5.379655551910401,
      "epoch": 2.419114728066693,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044173734940912586,
      "loss": 5.0989,
      "mean_token_accuracy": 0.16881851702928544,
      "num_tokens": 53406684.0,
      "step": 24375
    },
    {
      "entropy": 5.369517421722412,
      "epoch": 2.4196109567288606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044171355706029247,
      "loss": 4.9693,
      "mean_token_accuracy": 0.18242324739694596,
      "num_tokens": 53417256.0,
      "step": 24380
    },
    {
      "entropy": 5.347319507598877,
      "epoch": 2.420107185391028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044168976057732225,
      "loss": 4.9677,
      "mean_token_accuracy": 0.17411447763442994,
      "num_tokens": 53428305.0,
      "step": 24385
    },
    {
      "entropy": 5.308351755142212,
      "epoch": 2.4206034140531956,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004416659599608051,
      "loss": 4.9986,
      "mean_token_accuracy": 0.17386749833822251,
      "num_tokens": 53439301.0,
      "step": 24390
    },
    {
      "entropy": 5.39564208984375,
      "epoch": 2.4210996427153635,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004416421552113312,
      "loss": 5.1165,
      "mean_token_accuracy": 0.16437510997056962,
      "num_tokens": 53450478.0,
      "step": 24395
    },
    {
      "entropy": 5.387485027313232,
      "epoch": 2.421595871377531,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000441618346329491,
      "loss": 5.0426,
      "mean_token_accuracy": 0.17455680072307586,
      "num_tokens": 53461293.0,
      "step": 24400
    },
    {
      "entropy": 5.315534973144532,
      "epoch": 2.4220921000396984,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044159453331587493,
      "loss": 5.0139,
      "mean_token_accuracy": 0.17919946312904358,
      "num_tokens": 53472014.0,
      "step": 24405
    },
    {
      "entropy": 5.387729454040527,
      "epoch": 2.422588328701866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044157071617107325,
      "loss": 5.0857,
      "mean_token_accuracy": 0.17237299531698227,
      "num_tokens": 53483010.0,
      "step": 24410
    },
    {
      "entropy": 5.407374143600464,
      "epoch": 2.4230845573640334,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044154689489567684,
      "loss": 5.0624,
      "mean_token_accuracy": 0.17324400395154954,
      "num_tokens": 53494151.0,
      "step": 24415
    },
    {
      "entropy": 5.391531467437744,
      "epoch": 2.423580786026201,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044152306949027625,
      "loss": 5.0499,
      "mean_token_accuracy": 0.17409007251262665,
      "num_tokens": 53504462.0,
      "step": 24420
    },
    {
      "entropy": 5.375015306472778,
      "epoch": 2.4240770146883683,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044149923995546244,
      "loss": 5.0302,
      "mean_token_accuracy": 0.17463871091604233,
      "num_tokens": 53514896.0,
      "step": 24425
    },
    {
      "entropy": 5.343123483657837,
      "epoch": 2.424573243350536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004414754062918262,
      "loss": 4.9867,
      "mean_token_accuracy": 0.1825680360198021,
      "num_tokens": 53524656.0,
      "step": 24430
    },
    {
      "entropy": 5.452423906326294,
      "epoch": 2.4250694720127033,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004414515684999587,
      "loss": 5.0316,
      "mean_token_accuracy": 0.17734029591083528,
      "num_tokens": 53534896.0,
      "step": 24435
    },
    {
      "entropy": 5.284127330780029,
      "epoch": 2.4255657006748708,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004414277265804508,
      "loss": 4.9811,
      "mean_token_accuracy": 0.17960916608572006,
      "num_tokens": 53546108.0,
      "step": 24440
    },
    {
      "entropy": 5.263179159164428,
      "epoch": 2.4260619293370387,
      "grad_norm": 1.046875,
      "learning_rate": 0.000441403880533894,
      "loss": 5.0042,
      "mean_token_accuracy": 0.1735494390130043,
      "num_tokens": 53557325.0,
      "step": 24445
    },
    {
      "entropy": 5.281814384460449,
      "epoch": 2.426558157999206,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044138003036087956,
      "loss": 4.9305,
      "mean_token_accuracy": 0.180743969976902,
      "num_tokens": 53568737.0,
      "step": 24450
    },
    {
      "entropy": 5.398468112945556,
      "epoch": 2.4270543866613736,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004413561760619988,
      "loss": 5.0362,
      "mean_token_accuracy": 0.17444141507148742,
      "num_tokens": 53579312.0,
      "step": 24455
    },
    {
      "entropy": 5.326541566848755,
      "epoch": 2.427550615323541,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004413323176378434,
      "loss": 4.9365,
      "mean_token_accuracy": 0.18442853540182114,
      "num_tokens": 53590476.0,
      "step": 24460
    },
    {
      "entropy": 5.370265817642212,
      "epoch": 2.4280468439857086,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004413084550890049,
      "loss": 5.0887,
      "mean_token_accuracy": 0.17428426295518876,
      "num_tokens": 53601284.0,
      "step": 24465
    },
    {
      "entropy": 5.333635663986206,
      "epoch": 2.428543072647876,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004412845884160751,
      "loss": 5.0539,
      "mean_token_accuracy": 0.17063104808330537,
      "num_tokens": 53612590.0,
      "step": 24470
    },
    {
      "entropy": 5.376225185394287,
      "epoch": 2.4290393013100435,
      "grad_norm": 1.0,
      "learning_rate": 0.0004412607176196458,
      "loss": 5.0713,
      "mean_token_accuracy": 0.17084375768899918,
      "num_tokens": 53623049.0,
      "step": 24475
    },
    {
      "entropy": 5.351805877685547,
      "epoch": 2.429535529972211,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044123684270030907,
      "loss": 5.0147,
      "mean_token_accuracy": 0.17817028611898422,
      "num_tokens": 53633961.0,
      "step": 24480
    },
    {
      "entropy": 5.337976455688477,
      "epoch": 2.430031758634379,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044121296365865674,
      "loss": 5.0504,
      "mean_token_accuracy": 0.1753819078207016,
      "num_tokens": 53644634.0,
      "step": 24485
    },
    {
      "entropy": 5.327024459838867,
      "epoch": 2.4305279872965464,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044118908049528105,
      "loss": 4.9838,
      "mean_token_accuracy": 0.1824416473507881,
      "num_tokens": 53655982.0,
      "step": 24490
    },
    {
      "entropy": 5.442230129241944,
      "epoch": 2.431024215958714,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044116519321077435,
      "loss": 5.0792,
      "mean_token_accuracy": 0.17298808246850966,
      "num_tokens": 53667367.0,
      "step": 24495
    },
    {
      "entropy": 5.331802272796631,
      "epoch": 2.4315204446208813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044114130180572883,
      "loss": 4.9819,
      "mean_token_accuracy": 0.18083319216966628,
      "num_tokens": 53677191.0,
      "step": 24500
    },
    {
      "entropy": 5.33215708732605,
      "epoch": 2.432016673283049,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004411174062807371,
      "loss": 5.0486,
      "mean_token_accuracy": 0.17455772906541825,
      "num_tokens": 53688636.0,
      "step": 24505
    },
    {
      "entropy": 5.375634860992432,
      "epoch": 2.4325129019452163,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004410935066363916,
      "loss": 4.9835,
      "mean_token_accuracy": 0.17764726728200914,
      "num_tokens": 53699588.0,
      "step": 24510
    },
    {
      "entropy": 5.3843213558197025,
      "epoch": 2.4330091306073838,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044106960287328503,
      "loss": 5.1114,
      "mean_token_accuracy": 0.16809025555849075,
      "num_tokens": 53710427.0,
      "step": 24515
    },
    {
      "entropy": 5.412915945053101,
      "epoch": 2.4335053592695512,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004410456949920102,
      "loss": 5.0905,
      "mean_token_accuracy": 0.1675503209233284,
      "num_tokens": 53722319.0,
      "step": 24520
    },
    {
      "entropy": 5.356510019302368,
      "epoch": 2.4340015879317187,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044102178299315994,
      "loss": 5.0962,
      "mean_token_accuracy": 0.17219000458717346,
      "num_tokens": 53734059.0,
      "step": 24525
    },
    {
      "entropy": 5.3595599174499515,
      "epoch": 2.4344978165938866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044099786687732713,
      "loss": 5.0564,
      "mean_token_accuracy": 0.173480723798275,
      "num_tokens": 53745795.0,
      "step": 24530
    },
    {
      "entropy": 5.404663038253784,
      "epoch": 2.434994045256054,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004409739466451049,
      "loss": 5.0523,
      "mean_token_accuracy": 0.17371581345796586,
      "num_tokens": 53756781.0,
      "step": 24535
    },
    {
      "entropy": 5.4731851577758786,
      "epoch": 2.4354902739182216,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044095002229708655,
      "loss": 5.1158,
      "mean_token_accuracy": 0.16899864077568055,
      "num_tokens": 53769725.0,
      "step": 24540
    },
    {
      "entropy": 5.305251598358154,
      "epoch": 2.435986502580389,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044092609383386517,
      "loss": 5.014,
      "mean_token_accuracy": 0.18094824105501175,
      "num_tokens": 53781217.0,
      "step": 24545
    },
    {
      "entropy": 5.380830240249634,
      "epoch": 2.4364827312425565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044090216125603426,
      "loss": 5.0966,
      "mean_token_accuracy": 0.17547460943460463,
      "num_tokens": 53791478.0,
      "step": 24550
    },
    {
      "entropy": 5.341966724395752,
      "epoch": 2.436978959904724,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004408782245641872,
      "loss": 4.9846,
      "mean_token_accuracy": 0.18050077706575393,
      "num_tokens": 53802844.0,
      "step": 24555
    },
    {
      "entropy": 5.377667331695557,
      "epoch": 2.4374751885668915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004408542837589176,
      "loss": 5.042,
      "mean_token_accuracy": 0.17428137511014938,
      "num_tokens": 53812938.0,
      "step": 24560
    },
    {
      "entropy": 5.380779170989991,
      "epoch": 2.437971417229059,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044083033884081914,
      "loss": 5.002,
      "mean_token_accuracy": 0.1758854240179062,
      "num_tokens": 53824105.0,
      "step": 24565
    },
    {
      "entropy": 5.412218189239502,
      "epoch": 2.438467645891227,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004408063898104856,
      "loss": 5.1551,
      "mean_token_accuracy": 0.16970351338386536,
      "num_tokens": 53834956.0,
      "step": 24570
    },
    {
      "entropy": 5.387321615219117,
      "epoch": 2.4389638745533944,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044078243666851094,
      "loss": 5.0462,
      "mean_token_accuracy": 0.17428627461194993,
      "num_tokens": 53846257.0,
      "step": 24575
    },
    {
      "entropy": 5.399728727340698,
      "epoch": 2.439460103215562,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044075847941548905,
      "loss": 4.9716,
      "mean_token_accuracy": 0.1805548533797264,
      "num_tokens": 53857107.0,
      "step": 24580
    },
    {
      "entropy": 5.374495506286621,
      "epoch": 2.4399563318777293,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00044073451805201396,
      "loss": 5.0227,
      "mean_token_accuracy": 0.17515914887189865,
      "num_tokens": 53867779.0,
      "step": 24585
    },
    {
      "entropy": 5.3505490779876705,
      "epoch": 2.440452560539897,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044071055257868,
      "loss": 5.0154,
      "mean_token_accuracy": 0.17378681600093843,
      "num_tokens": 53878661.0,
      "step": 24590
    },
    {
      "entropy": 5.393152475357056,
      "epoch": 2.4409487892020643,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044068658299608145,
      "loss": 5.0885,
      "mean_token_accuracy": 0.17146968096494675,
      "num_tokens": 53891300.0,
      "step": 24595
    },
    {
      "entropy": 5.391712284088134,
      "epoch": 2.4414450178642317,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044066260930481264,
      "loss": 4.9761,
      "mean_token_accuracy": 0.17916012853384017,
      "num_tokens": 53901086.0,
      "step": 24600
    },
    {
      "entropy": 5.386125898361206,
      "epoch": 2.441941246526399,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004406386315054681,
      "loss": 5.0735,
      "mean_token_accuracy": 0.16419337540864945,
      "num_tokens": 53912599.0,
      "step": 24605
    },
    {
      "entropy": 5.296385002136231,
      "epoch": 2.4424374751885667,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044061464959864243,
      "loss": 5.055,
      "mean_token_accuracy": 0.16969068050384523,
      "num_tokens": 53924293.0,
      "step": 24610
    },
    {
      "entropy": 5.362279844284058,
      "epoch": 2.4429337038507346,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004405906635849302,
      "loss": 5.1033,
      "mean_token_accuracy": 0.17046958804130555,
      "num_tokens": 53935074.0,
      "step": 24615
    },
    {
      "entropy": 5.409727144241333,
      "epoch": 2.443429932512902,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004405666734649265,
      "loss": 5.044,
      "mean_token_accuracy": 0.1744973123073578,
      "num_tokens": 53946770.0,
      "step": 24620
    },
    {
      "entropy": 5.3921490669250485,
      "epoch": 2.4439261611750696,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044054267923922594,
      "loss": 5.0227,
      "mean_token_accuracy": 0.1733122080564499,
      "num_tokens": 53957077.0,
      "step": 24625
    },
    {
      "entropy": 5.322025489807129,
      "epoch": 2.444422389837237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044051868090842376,
      "loss": 5.0283,
      "mean_token_accuracy": 0.18075695186853408,
      "num_tokens": 53968455.0,
      "step": 24630
    },
    {
      "entropy": 5.29988923072815,
      "epoch": 2.4449186184994045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044049467847311486,
      "loss": 5.0009,
      "mean_token_accuracy": 0.17448631078004836,
      "num_tokens": 53980429.0,
      "step": 24635
    },
    {
      "entropy": 5.314863681793213,
      "epoch": 2.445414847161572,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004404706719338946,
      "loss": 5.0,
      "mean_token_accuracy": 0.18227920532226563,
      "num_tokens": 53990937.0,
      "step": 24640
    },
    {
      "entropy": 5.361550569534302,
      "epoch": 2.4459110758237395,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044044666129135825,
      "loss": 5.0178,
      "mean_token_accuracy": 0.176066030561924,
      "num_tokens": 54003526.0,
      "step": 24645
    },
    {
      "entropy": 5.31281418800354,
      "epoch": 2.446407304485907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044042264654610115,
      "loss": 4.9803,
      "mean_token_accuracy": 0.18835097700357437,
      "num_tokens": 54013875.0,
      "step": 24650
    },
    {
      "entropy": 5.353661251068115,
      "epoch": 2.446903533148075,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044039862769871884,
      "loss": 5.1427,
      "mean_token_accuracy": 0.16687737852334977,
      "num_tokens": 54024103.0,
      "step": 24655
    },
    {
      "entropy": 5.385623979568481,
      "epoch": 2.4473997618102423,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044037460474980703,
      "loss": 5.1103,
      "mean_token_accuracy": 0.16703845858573912,
      "num_tokens": 54035417.0,
      "step": 24660
    },
    {
      "entropy": 5.445803689956665,
      "epoch": 2.44789599047241,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004403505776999614,
      "loss": 5.0843,
      "mean_token_accuracy": 0.1673841580748558,
      "num_tokens": 54045180.0,
      "step": 24665
    },
    {
      "entropy": 5.397284841537475,
      "epoch": 2.4483922191345773,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044032654654977766,
      "loss": 5.01,
      "mean_token_accuracy": 0.18112823814153672,
      "num_tokens": 54056493.0,
      "step": 24670
    },
    {
      "entropy": 5.2781883716583256,
      "epoch": 2.4488884477967447,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004403025112998519,
      "loss": 5.0092,
      "mean_token_accuracy": 0.17476174235343933,
      "num_tokens": 54067445.0,
      "step": 24675
    },
    {
      "entropy": 5.34616985321045,
      "epoch": 2.4493846764589122,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004402784719507799,
      "loss": 4.9933,
      "mean_token_accuracy": 0.18302221298217775,
      "num_tokens": 54076937.0,
      "step": 24680
    },
    {
      "entropy": 5.415417861938477,
      "epoch": 2.4498809051210797,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004402544285031581,
      "loss": 5.0686,
      "mean_token_accuracy": 0.1808757498860359,
      "num_tokens": 54088014.0,
      "step": 24685
    },
    {
      "entropy": 5.406236982345581,
      "epoch": 2.450377133783247,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004402303809575825,
      "loss": 4.9997,
      "mean_token_accuracy": 0.17546266466379165,
      "num_tokens": 54098984.0,
      "step": 24690
    },
    {
      "entropy": 5.376453828811646,
      "epoch": 2.4508733624454146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004402063293146495,
      "loss": 5.1511,
      "mean_token_accuracy": 0.1708860993385315,
      "num_tokens": 54110490.0,
      "step": 24695
    },
    {
      "entropy": 5.376290559768677,
      "epoch": 2.451369591107582,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004401822735749555,
      "loss": 5.0524,
      "mean_token_accuracy": 0.17200125753879547,
      "num_tokens": 54120785.0,
      "step": 24700
    },
    {
      "entropy": 5.347737836837768,
      "epoch": 2.45186581976975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044015821373909707,
      "loss": 4.9631,
      "mean_token_accuracy": 0.17573967576026917,
      "num_tokens": 54131339.0,
      "step": 24705
    },
    {
      "entropy": 5.351630830764771,
      "epoch": 2.4523620484319175,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044013414980767084,
      "loss": 5.0801,
      "mean_token_accuracy": 0.1677129700779915,
      "num_tokens": 54142270.0,
      "step": 24710
    },
    {
      "entropy": 5.324189281463623,
      "epoch": 2.452858277094085,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004401100817812735,
      "loss": 5.0247,
      "mean_token_accuracy": 0.17328174263238907,
      "num_tokens": 54153045.0,
      "step": 24715
    },
    {
      "entropy": 5.3117430210113525,
      "epoch": 2.4533545057562525,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044008600966050196,
      "loss": 4.9647,
      "mean_token_accuracy": 0.18025840520858766,
      "num_tokens": 54164069.0,
      "step": 24720
    },
    {
      "entropy": 5.384675979614258,
      "epoch": 2.45385073441842,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004400619334459531,
      "loss": 5.0575,
      "mean_token_accuracy": 0.17294350117444993,
      "num_tokens": 54176902.0,
      "step": 24725
    },
    {
      "entropy": 5.31982421875,
      "epoch": 2.4543469630805874,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044003785313822386,
      "loss": 5.0128,
      "mean_token_accuracy": 0.17983949035406113,
      "num_tokens": 54188515.0,
      "step": 24730
    },
    {
      "entropy": 5.399835395812988,
      "epoch": 2.454843191742755,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044001376873791166,
      "loss": 5.0685,
      "mean_token_accuracy": 0.17419227510690688,
      "num_tokens": 54199077.0,
      "step": 24735
    },
    {
      "entropy": 5.4539719581604,
      "epoch": 2.455339420404923,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043998968024561337,
      "loss": 5.1032,
      "mean_token_accuracy": 0.1727812707424164,
      "num_tokens": 54210216.0,
      "step": 24740
    },
    {
      "entropy": 5.332045602798462,
      "epoch": 2.4558356490670903,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004399655876619267,
      "loss": 5.0223,
      "mean_token_accuracy": 0.17980478703975677,
      "num_tokens": 54221143.0,
      "step": 24745
    },
    {
      "entropy": 5.384216785430908,
      "epoch": 2.4563318777292578,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043994149098744884,
      "loss": 5.0904,
      "mean_token_accuracy": 0.17526769191026687,
      "num_tokens": 54233070.0,
      "step": 24750
    },
    {
      "entropy": 5.335721778869629,
      "epoch": 2.4568281063914252,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004399173902227775,
      "loss": 5.0158,
      "mean_token_accuracy": 0.17775343209505082,
      "num_tokens": 54243793.0,
      "step": 24755
    },
    {
      "entropy": 5.305764150619507,
      "epoch": 2.4573243350535927,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043989328536851014,
      "loss": 5.0159,
      "mean_token_accuracy": 0.1774770736694336,
      "num_tokens": 54256983.0,
      "step": 24760
    },
    {
      "entropy": 5.307118225097656,
      "epoch": 2.45782056371576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043986917642524465,
      "loss": 4.9891,
      "mean_token_accuracy": 0.17726021260023117,
      "num_tokens": 54268789.0,
      "step": 24765
    },
    {
      "entropy": 5.447189569473267,
      "epoch": 2.4583167923779277,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043984506339357887,
      "loss": 5.0388,
      "mean_token_accuracy": 0.1704381063580513,
      "num_tokens": 54279517.0,
      "step": 24770
    },
    {
      "entropy": 5.365664863586426,
      "epoch": 2.458813021040095,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004398209462741107,
      "loss": 5.0206,
      "mean_token_accuracy": 0.1754514455795288,
      "num_tokens": 54290865.0,
      "step": 24775
    },
    {
      "entropy": 5.466447353363037,
      "epoch": 2.4593092497022626,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004397968250674383,
      "loss": 5.1146,
      "mean_token_accuracy": 0.17673116624355317,
      "num_tokens": 54302658.0,
      "step": 24780
    },
    {
      "entropy": 5.355204439163208,
      "epoch": 2.45980547836443,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004397726997741597,
      "loss": 5.0538,
      "mean_token_accuracy": 0.1766948163509369,
      "num_tokens": 54314000.0,
      "step": 24785
    },
    {
      "entropy": 5.271871709823609,
      "epoch": 2.460301707026598,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004397485703948732,
      "loss": 5.0282,
      "mean_token_accuracy": 0.17754343748092652,
      "num_tokens": 54325952.0,
      "step": 24790
    },
    {
      "entropy": 5.403670883178711,
      "epoch": 2.4607979356887655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004397244369301771,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17045315653085708,
      "num_tokens": 54338582.0,
      "step": 24795
    },
    {
      "entropy": 5.378180646896363,
      "epoch": 2.461294164350933,
      "grad_norm": 1.125,
      "learning_rate": 0.00043970029938066995,
      "loss": 5.0893,
      "mean_token_accuracy": 0.16786551475524902,
      "num_tokens": 54349513.0,
      "step": 24800
    },
    {
      "entropy": 5.400853013992309,
      "epoch": 2.4617903930131004,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004396761577469503,
      "loss": 5.0104,
      "mean_token_accuracy": 0.1749692440032959,
      "num_tokens": 54360264.0,
      "step": 24805
    },
    {
      "entropy": 5.379650449752807,
      "epoch": 2.462286621675268,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043965201202961673,
      "loss": 5.0688,
      "mean_token_accuracy": 0.1676955446600914,
      "num_tokens": 54371379.0,
      "step": 24810
    },
    {
      "entropy": 5.343711709976196,
      "epoch": 2.4627828503374354,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043962786222926807,
      "loss": 5.0625,
      "mean_token_accuracy": 0.1800290435552597,
      "num_tokens": 54382844.0,
      "step": 24815
    },
    {
      "entropy": 5.333056974411011,
      "epoch": 2.463279078999603,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043960370834650325,
      "loss": 5.008,
      "mean_token_accuracy": 0.17692189216613768,
      "num_tokens": 54392807.0,
      "step": 24820
    },
    {
      "entropy": 5.324676656723023,
      "epoch": 2.4637753076617708,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004395795503819211,
      "loss": 5.0298,
      "mean_token_accuracy": 0.17927694022655488,
      "num_tokens": 54404060.0,
      "step": 24825
    },
    {
      "entropy": 5.310246276855469,
      "epoch": 2.4642715363239382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004395553883361207,
      "loss": 4.9379,
      "mean_token_accuracy": 0.1792901113629341,
      "num_tokens": 54414636.0,
      "step": 24830
    },
    {
      "entropy": 5.322856855392456,
      "epoch": 2.4647677649861057,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043953122220970127,
      "loss": 4.992,
      "mean_token_accuracy": 0.1798480659723282,
      "num_tokens": 54426185.0,
      "step": 24835
    },
    {
      "entropy": 5.3330625057220455,
      "epoch": 2.465263993648273,
      "grad_norm": 1.0,
      "learning_rate": 0.00043950705200326205,
      "loss": 5.0034,
      "mean_token_accuracy": 0.1761929303407669,
      "num_tokens": 54436468.0,
      "step": 24840
    },
    {
      "entropy": 5.339693069458008,
      "epoch": 2.4657602223104407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004394828777174025,
      "loss": 5.0076,
      "mean_token_accuracy": 0.18448252975940704,
      "num_tokens": 54446228.0,
      "step": 24845
    },
    {
      "entropy": 5.331423377990722,
      "epoch": 2.466256450972608,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043945869935272194,
      "loss": 5.0699,
      "mean_token_accuracy": 0.16965340524911882,
      "num_tokens": 54456704.0,
      "step": 24850
    },
    {
      "entropy": 5.360025453567505,
      "epoch": 2.4667526796347756,
      "grad_norm": 1.515625,
      "learning_rate": 0.00043943451690982006,
      "loss": 5.0391,
      "mean_token_accuracy": 0.1807718187570572,
      "num_tokens": 54467780.0,
      "step": 24855
    },
    {
      "entropy": 5.462622404098511,
      "epoch": 2.467248908296943,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004394103303892964,
      "loss": 5.1474,
      "mean_token_accuracy": 0.17418680489063262,
      "num_tokens": 54478428.0,
      "step": 24860
    },
    {
      "entropy": 5.310592269897461,
      "epoch": 2.4677451369591106,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043938613979175087,
      "loss": 4.9321,
      "mean_token_accuracy": 0.18200430124998093,
      "num_tokens": 54489544.0,
      "step": 24865
    },
    {
      "entropy": 5.283468723297119,
      "epoch": 2.468241365621278,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004393619451177834,
      "loss": 5.0267,
      "mean_token_accuracy": 0.18086652159690858,
      "num_tokens": 54501438.0,
      "step": 24870
    },
    {
      "entropy": 5.2674518585205075,
      "epoch": 2.468737594283446,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043933774636799365,
      "loss": 4.9339,
      "mean_token_accuracy": 0.18060911744832991,
      "num_tokens": 54511746.0,
      "step": 24875
    },
    {
      "entropy": 5.366827297210693,
      "epoch": 2.4692338229456134,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043931354354298204,
      "loss": 5.0609,
      "mean_token_accuracy": 0.1732812985777855,
      "num_tokens": 54523141.0,
      "step": 24880
    },
    {
      "entropy": 5.494414281845093,
      "epoch": 2.469730051607781,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004392893366433486,
      "loss": 5.1718,
      "mean_token_accuracy": 0.1719530075788498,
      "num_tokens": 54533275.0,
      "step": 24885
    },
    {
      "entropy": 5.4132408618927,
      "epoch": 2.4702262802699484,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043926512566969366,
      "loss": 5.0076,
      "mean_token_accuracy": 0.17810919880867004,
      "num_tokens": 54545309.0,
      "step": 24890
    },
    {
      "entropy": 5.294353199005127,
      "epoch": 2.470722508932116,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043924091062261743,
      "loss": 4.9891,
      "mean_token_accuracy": 0.18003311455249787,
      "num_tokens": 54554905.0,
      "step": 24895
    },
    {
      "entropy": 5.346034002304077,
      "epoch": 2.4712187375942833,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004392166915027206,
      "loss": 5.0333,
      "mean_token_accuracy": 0.17456697672605515,
      "num_tokens": 54566050.0,
      "step": 24900
    },
    {
      "entropy": 5.383196830749512,
      "epoch": 2.471714966256451,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004391924683106037,
      "loss": 5.0972,
      "mean_token_accuracy": 0.17132419645786284,
      "num_tokens": 54577791.0,
      "step": 24905
    },
    {
      "entropy": 5.388339710235596,
      "epoch": 2.4722111949186187,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004391682410468674,
      "loss": 5.0282,
      "mean_token_accuracy": 0.18076155334711075,
      "num_tokens": 54589183.0,
      "step": 24910
    },
    {
      "entropy": 5.318300867080689,
      "epoch": 2.472707423580786,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043914400971211255,
      "loss": 5.0357,
      "mean_token_accuracy": 0.17247350066900252,
      "num_tokens": 54600776.0,
      "step": 24915
    },
    {
      "entropy": 5.32877926826477,
      "epoch": 2.4732036522429537,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004391197743069398,
      "loss": 4.9787,
      "mean_token_accuracy": 0.1835220605134964,
      "num_tokens": 54611167.0,
      "step": 24920
    },
    {
      "entropy": 5.360300016403198,
      "epoch": 2.473699880905121,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004390955348319503,
      "loss": 5.025,
      "mean_token_accuracy": 0.17651212960481644,
      "num_tokens": 54622811.0,
      "step": 24925
    },
    {
      "entropy": 5.376177787780762,
      "epoch": 2.4741961095672886,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043907129128774525,
      "loss": 5.0402,
      "mean_token_accuracy": 0.1782752349972725,
      "num_tokens": 54635109.0,
      "step": 24930
    },
    {
      "entropy": 5.434084177017212,
      "epoch": 2.474692338229456,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043904704367492563,
      "loss": 5.0769,
      "mean_token_accuracy": 0.17701766043901443,
      "num_tokens": 54644260.0,
      "step": 24935
    },
    {
      "entropy": 5.348824787139892,
      "epoch": 2.4751885668916236,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004390227919940928,
      "loss": 4.9597,
      "mean_token_accuracy": 0.1792246162891388,
      "num_tokens": 54654433.0,
      "step": 24940
    },
    {
      "entropy": 5.342943859100342,
      "epoch": 2.475684795553791,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043899853624584825,
      "loss": 5.0945,
      "mean_token_accuracy": 0.1702755481004715,
      "num_tokens": 54666582.0,
      "step": 24945
    },
    {
      "entropy": 5.3504993438720705,
      "epoch": 2.4761810242159585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043897427643079336,
      "loss": 5.0049,
      "mean_token_accuracy": 0.17747773379087448,
      "num_tokens": 54677247.0,
      "step": 24950
    },
    {
      "entropy": 5.379823493957519,
      "epoch": 2.476677252878126,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043895001254952963,
      "loss": 5.0546,
      "mean_token_accuracy": 0.17260948717594146,
      "num_tokens": 54688423.0,
      "step": 24955
    },
    {
      "entropy": 5.357532215118408,
      "epoch": 2.477173481540294,
      "grad_norm": 1.015625,
      "learning_rate": 0.000438925744602659,
      "loss": 5.056,
      "mean_token_accuracy": 0.17868893146514891,
      "num_tokens": 54698517.0,
      "step": 24960
    },
    {
      "entropy": 5.413807487487793,
      "epoch": 2.4776697102024614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043890147259078314,
      "loss": 5.093,
      "mean_token_accuracy": 0.1704331234097481,
      "num_tokens": 54711635.0,
      "step": 24965
    },
    {
      "entropy": 5.376549291610718,
      "epoch": 2.478165938864629,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043887719651450384,
      "loss": 5.0379,
      "mean_token_accuracy": 0.17927781343460084,
      "num_tokens": 54723051.0,
      "step": 24970
    },
    {
      "entropy": 5.449807834625244,
      "epoch": 2.4786621675267964,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004388529163744233,
      "loss": 5.028,
      "mean_token_accuracy": 0.17871348410844803,
      "num_tokens": 54734191.0,
      "step": 24975
    },
    {
      "entropy": 5.315324449539185,
      "epoch": 2.479158396188964,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004388286321711435,
      "loss": 5.0037,
      "mean_token_accuracy": 0.17765514105558394,
      "num_tokens": 54745632.0,
      "step": 24980
    },
    {
      "entropy": 5.359863948822022,
      "epoch": 2.4796546248511313,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004388043439052665,
      "loss": 5.0588,
      "mean_token_accuracy": 0.17062133103609084,
      "num_tokens": 54755871.0,
      "step": 24985
    },
    {
      "entropy": 5.330174732208252,
      "epoch": 2.480150853513299,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043878005157739486,
      "loss": 4.9649,
      "mean_token_accuracy": 0.18304739892482758,
      "num_tokens": 54768036.0,
      "step": 24990
    },
    {
      "entropy": 5.325761461257935,
      "epoch": 2.4806470821754667,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004387557551881308,
      "loss": 5.0031,
      "mean_token_accuracy": 0.1715209186077118,
      "num_tokens": 54778518.0,
      "step": 24995
    },
    {
      "entropy": 5.2969183921813965,
      "epoch": 2.481143310837634,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000438731454738077,
      "loss": 4.986,
      "mean_token_accuracy": 0.17756396532058716,
      "num_tokens": 54790786.0,
      "step": 25000
    },
    {
      "entropy": 5.318441867828369,
      "epoch": 2.4816395394998016,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043870715022783575,
      "loss": 4.9682,
      "mean_token_accuracy": 0.18099630028009414,
      "num_tokens": 54800063.0,
      "step": 25005
    },
    {
      "entropy": 5.372000789642334,
      "epoch": 2.482135768161969,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004386828416580101,
      "loss": 5.0501,
      "mean_token_accuracy": 0.1699716940522194,
      "num_tokens": 54810605.0,
      "step": 25010
    },
    {
      "entropy": 5.3572587966918945,
      "epoch": 2.4826319968241366,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004386585290292026,
      "loss": 4.9791,
      "mean_token_accuracy": 0.17937591373920442,
      "num_tokens": 54820626.0,
      "step": 25015
    },
    {
      "entropy": 5.338228940963745,
      "epoch": 2.483128225486304,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004386342123420161,
      "loss": 4.9759,
      "mean_token_accuracy": 0.1762418642640114,
      "num_tokens": 54831395.0,
      "step": 25020
    },
    {
      "entropy": 5.303357887268066,
      "epoch": 2.4836244541484715,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043860989159705396,
      "loss": 4.9862,
      "mean_token_accuracy": 0.17528152465820312,
      "num_tokens": 54841238.0,
      "step": 25025
    },
    {
      "entropy": 5.40837721824646,
      "epoch": 2.484120682810639,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043858556679491896,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17422469556331635,
      "num_tokens": 54851301.0,
      "step": 25030
    },
    {
      "entropy": 5.384293794631958,
      "epoch": 2.4846169114728065,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043856123793621437,
      "loss": 4.9913,
      "mean_token_accuracy": 0.18155156821012497,
      "num_tokens": 54863135.0,
      "step": 25035
    },
    {
      "entropy": 5.355708074569702,
      "epoch": 2.485113140134974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043853690502154355,
      "loss": 5.0677,
      "mean_token_accuracy": 0.17085759490728378,
      "num_tokens": 54874780.0,
      "step": 25040
    },
    {
      "entropy": 5.401432228088379,
      "epoch": 2.485609368797142,
      "grad_norm": 1.0,
      "learning_rate": 0.0004385125680515099,
      "loss": 5.0466,
      "mean_token_accuracy": 0.17986894249916077,
      "num_tokens": 54885237.0,
      "step": 25045
    },
    {
      "entropy": 5.382382011413574,
      "epoch": 2.4861055974593094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004384882270267168,
      "loss": 5.0598,
      "mean_token_accuracy": 0.18197105824947357,
      "num_tokens": 54896801.0,
      "step": 25050
    },
    {
      "entropy": 5.366536378860474,
      "epoch": 2.486601826121477,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004384638819477681,
      "loss": 5.0146,
      "mean_token_accuracy": 0.1714157998561859,
      "num_tokens": 54908427.0,
      "step": 25055
    },
    {
      "entropy": 5.277341508865357,
      "epoch": 2.4870980547836443,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043843953281526727,
      "loss": 4.9703,
      "mean_token_accuracy": 0.18457310050725936,
      "num_tokens": 54918394.0,
      "step": 25060
    },
    {
      "entropy": 5.340137577056884,
      "epoch": 2.487594283445812,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043841517962981823,
      "loss": 4.9923,
      "mean_token_accuracy": 0.1826310634613037,
      "num_tokens": 54928613.0,
      "step": 25065
    },
    {
      "entropy": 5.39379768371582,
      "epoch": 2.4880905121079793,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043839082239202496,
      "loss": 4.9551,
      "mean_token_accuracy": 0.17859004437923431,
      "num_tokens": 54938938.0,
      "step": 25070
    },
    {
      "entropy": 5.417671632766724,
      "epoch": 2.4885867407701467,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004383664611024912,
      "loss": 5.0904,
      "mean_token_accuracy": 0.17588224411010742,
      "num_tokens": 54950152.0,
      "step": 25075
    },
    {
      "entropy": 5.371277475357056,
      "epoch": 2.489082969432314,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004383420957618213,
      "loss": 5.123,
      "mean_token_accuracy": 0.16931429505348206,
      "num_tokens": 54962191.0,
      "step": 25080
    },
    {
      "entropy": 5.340079879760742,
      "epoch": 2.489579198094482,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043831772637061944,
      "loss": 4.9357,
      "mean_token_accuracy": 0.17825116962194443,
      "num_tokens": 54972532.0,
      "step": 25085
    },
    {
      "entropy": 5.410003280639648,
      "epoch": 2.4900754267566496,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043829335292948986,
      "loss": 5.0668,
      "mean_token_accuracy": 0.17164491713047028,
      "num_tokens": 54983212.0,
      "step": 25090
    },
    {
      "entropy": 5.266553020477295,
      "epoch": 2.490571655418817,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043826897543903696,
      "loss": 4.9709,
      "mean_token_accuracy": 0.18454791605472565,
      "num_tokens": 54994094.0,
      "step": 25095
    },
    {
      "entropy": 5.368187808990479,
      "epoch": 2.4910678840809846,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043824459389986534,
      "loss": 5.0675,
      "mean_token_accuracy": 0.17471761852502823,
      "num_tokens": 55004098.0,
      "step": 25100
    },
    {
      "entropy": 5.308345890045166,
      "epoch": 2.491564112743152,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004382202083125795,
      "loss": 4.9579,
      "mean_token_accuracy": 0.18361985683441162,
      "num_tokens": 55013680.0,
      "step": 25105
    },
    {
      "entropy": 5.311703395843506,
      "epoch": 2.4920603414053195,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004381958186777842,
      "loss": 4.9167,
      "mean_token_accuracy": 0.18837100863456727,
      "num_tokens": 55024311.0,
      "step": 25110
    },
    {
      "entropy": 5.36719651222229,
      "epoch": 2.492556570067487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004381714249960843,
      "loss": 5.1843,
      "mean_token_accuracy": 0.16638885885477067,
      "num_tokens": 55036215.0,
      "step": 25115
    },
    {
      "entropy": 5.455293130874634,
      "epoch": 2.4930527987296545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004381470272680846,
      "loss": 5.0295,
      "mean_token_accuracy": 0.17061668932437896,
      "num_tokens": 55046402.0,
      "step": 25120
    },
    {
      "entropy": 5.363403034210205,
      "epoch": 2.493549027391822,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004381226254943902,
      "loss": 4.9775,
      "mean_token_accuracy": 0.17401257306337356,
      "num_tokens": 55056307.0,
      "step": 25125
    },
    {
      "entropy": 5.322064924240112,
      "epoch": 2.49404525605399,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004380982196756062,
      "loss": 4.9967,
      "mean_token_accuracy": 0.18057507872581482,
      "num_tokens": 55066918.0,
      "step": 25130
    },
    {
      "entropy": 5.342610311508179,
      "epoch": 2.4945414847161573,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00043807380981233777,
      "loss": 5.0244,
      "mean_token_accuracy": 0.16964344829320907,
      "num_tokens": 55078166.0,
      "step": 25135
    },
    {
      "entropy": 5.358501243591308,
      "epoch": 2.495037713378325,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004380493959051903,
      "loss": 5.0616,
      "mean_token_accuracy": 0.17545757740736007,
      "num_tokens": 55090919.0,
      "step": 25140
    },
    {
      "entropy": 5.407155323028564,
      "epoch": 2.4955339420404923,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043802497795476916,
      "loss": 5.0711,
      "mean_token_accuracy": 0.17840819656848908,
      "num_tokens": 55100512.0,
      "step": 25145
    },
    {
      "entropy": 5.323906755447387,
      "epoch": 2.4960301707026598,
      "grad_norm": 1.0,
      "learning_rate": 0.0004380005559616798,
      "loss": 5.0008,
      "mean_token_accuracy": 0.1805847927927971,
      "num_tokens": 55111845.0,
      "step": 25150
    },
    {
      "entropy": 5.431775999069214,
      "epoch": 2.4965263993648272,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004379761299265279,
      "loss": 5.1514,
      "mean_token_accuracy": 0.16458204835653306,
      "num_tokens": 55123744.0,
      "step": 25155
    },
    {
      "entropy": 5.4117138385772705,
      "epoch": 2.4970226280269947,
      "grad_norm": 1.125,
      "learning_rate": 0.0004379516998499192,
      "loss": 5.0205,
      "mean_token_accuracy": 0.17804994136095048,
      "num_tokens": 55133789.0,
      "step": 25160
    },
    {
      "entropy": 5.349097633361817,
      "epoch": 2.497518856689162,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004379272657324594,
      "loss": 5.0421,
      "mean_token_accuracy": 0.17784464210271836,
      "num_tokens": 55145325.0,
      "step": 25165
    },
    {
      "entropy": 5.3389636993408205,
      "epoch": 2.49801508535133,
      "grad_norm": 1.0,
      "learning_rate": 0.0004379028275747545,
      "loss": 5.0086,
      "mean_token_accuracy": 0.17576297670602797,
      "num_tokens": 55157639.0,
      "step": 25170
    },
    {
      "entropy": 5.330803728103637,
      "epoch": 2.4985113140134976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043787838537741054,
      "loss": 4.9472,
      "mean_token_accuracy": 0.17743006348609924,
      "num_tokens": 55167796.0,
      "step": 25175
    },
    {
      "entropy": 5.298856735229492,
      "epoch": 2.499007542675665,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004378539391410336,
      "loss": 5.0585,
      "mean_token_accuracy": 0.17527279108762742,
      "num_tokens": 55178384.0,
      "step": 25180
    },
    {
      "entropy": 5.4111532211303714,
      "epoch": 2.4995037713378325,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043782948886622987,
      "loss": 5.0962,
      "mean_token_accuracy": 0.17431144565343856,
      "num_tokens": 55188363.0,
      "step": 25185
    },
    {
      "entropy": 5.438766813278198,
      "epoch": 2.5,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004378050345536057,
      "loss": 5.0695,
      "mean_token_accuracy": 0.1702815055847168,
      "num_tokens": 55198293.0,
      "step": 25190
    },
    {
      "entropy": 5.354202318191528,
      "epoch": 2.5004962286621675,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043778057620376745,
      "loss": 5.0211,
      "mean_token_accuracy": 0.18087466210126876,
      "num_tokens": 55208368.0,
      "step": 25195
    },
    {
      "entropy": 5.3633181095123295,
      "epoch": 2.500992457324335,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043775611381732164,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17419978827238083,
      "num_tokens": 55220075.0,
      "step": 25200
    },
    {
      "entropy": 5.368018674850464,
      "epoch": 2.5014886859865024,
      "grad_norm": 1.078125,
      "learning_rate": 0.000437731647394875,
      "loss": 5.0254,
      "mean_token_accuracy": 0.17707870304584503,
      "num_tokens": 55231060.0,
      "step": 25205
    },
    {
      "entropy": 5.404204845428467,
      "epoch": 2.50198491464867,
      "grad_norm": 1.0,
      "learning_rate": 0.0004377071769370341,
      "loss": 5.036,
      "mean_token_accuracy": 0.17414015382528306,
      "num_tokens": 55242443.0,
      "step": 25210
    },
    {
      "entropy": 5.20925874710083,
      "epoch": 2.5024811433108374,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043768270244440587,
      "loss": 5.0008,
      "mean_token_accuracy": 0.18212016075849533,
      "num_tokens": 55253285.0,
      "step": 25215
    },
    {
      "entropy": 5.302894735336304,
      "epoch": 2.5029773719730053,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043765822391759714,
      "loss": 4.9808,
      "mean_token_accuracy": 0.17842251807451248,
      "num_tokens": 55264694.0,
      "step": 25220
    },
    {
      "entropy": 5.395809364318848,
      "epoch": 2.5034736006351728,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004376337413572149,
      "loss": 5.0106,
      "mean_token_accuracy": 0.17899688929319382,
      "num_tokens": 55275429.0,
      "step": 25225
    },
    {
      "entropy": 5.407653188705444,
      "epoch": 2.5039698292973402,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043760925476386635,
      "loss": 5.0853,
      "mean_token_accuracy": 0.1743750810623169,
      "num_tokens": 55286764.0,
      "step": 25230
    },
    {
      "entropy": 5.278618574142456,
      "epoch": 2.5044660579595077,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004375847641381587,
      "loss": 4.9296,
      "mean_token_accuracy": 0.18614568412303925,
      "num_tokens": 55296284.0,
      "step": 25235
    },
    {
      "entropy": 5.372319221496582,
      "epoch": 2.504962286621675,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004375602694806993,
      "loss": 5.0036,
      "mean_token_accuracy": 0.1809812992811203,
      "num_tokens": 55307684.0,
      "step": 25240
    },
    {
      "entropy": 5.315023422241211,
      "epoch": 2.5054585152838427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043753577079209537,
      "loss": 4.9831,
      "mean_token_accuracy": 0.17840491235256195,
      "num_tokens": 55318202.0,
      "step": 25245
    },
    {
      "entropy": 5.363704347610474,
      "epoch": 2.5059547439460106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043751126807295464,
      "loss": 5.0717,
      "mean_token_accuracy": 0.17588117867708206,
      "num_tokens": 55328827.0,
      "step": 25250
    },
    {
      "entropy": 5.400156164169312,
      "epoch": 2.506450972608178,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043748676132388464,
      "loss": 5.0658,
      "mean_token_accuracy": 0.1675372913479805,
      "num_tokens": 55338171.0,
      "step": 25255
    },
    {
      "entropy": 5.316308975219727,
      "epoch": 2.5069472012703455,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043746225054549316,
      "loss": 5.0142,
      "mean_token_accuracy": 0.1776740789413452,
      "num_tokens": 55349081.0,
      "step": 25260
    },
    {
      "entropy": 5.348902750015259,
      "epoch": 2.507443429932513,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043743773573838785,
      "loss": 4.9863,
      "mean_token_accuracy": 0.17973422706127168,
      "num_tokens": 55359344.0,
      "step": 25265
    },
    {
      "entropy": 5.351339292526245,
      "epoch": 2.5079396585946805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043741321690317673,
      "loss": 4.9106,
      "mean_token_accuracy": 0.18749842494726182,
      "num_tokens": 55369643.0,
      "step": 25270
    },
    {
      "entropy": 5.281532669067383,
      "epoch": 2.508435887256848,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004373886940404677,
      "loss": 4.9288,
      "mean_token_accuracy": 0.18417527973651887,
      "num_tokens": 55380812.0,
      "step": 25275
    },
    {
      "entropy": 5.271239566802978,
      "epoch": 2.5089321159190154,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004373641671508691,
      "loss": 5.006,
      "mean_token_accuracy": 0.18676023483276366,
      "num_tokens": 55393136.0,
      "step": 25280
    },
    {
      "entropy": 5.304402780532837,
      "epoch": 2.509428344581183,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000437339636234989,
      "loss": 4.9195,
      "mean_token_accuracy": 0.17786708623170852,
      "num_tokens": 55404317.0,
      "step": 25285
    },
    {
      "entropy": 5.35689640045166,
      "epoch": 2.5099245732433504,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043731510129343574,
      "loss": 5.0873,
      "mean_token_accuracy": 0.1719956323504448,
      "num_tokens": 55415812.0,
      "step": 25290
    },
    {
      "entropy": 5.407695341110229,
      "epoch": 2.510420801905518,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043729056232681766,
      "loss": 5.0553,
      "mean_token_accuracy": 0.17422572821378707,
      "num_tokens": 55426402.0,
      "step": 25295
    },
    {
      "entropy": 5.401298379898071,
      "epoch": 2.5109170305676853,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004372660193357434,
      "loss": 5.0666,
      "mean_token_accuracy": 0.176849365234375,
      "num_tokens": 55437474.0,
      "step": 25300
    },
    {
      "entropy": 5.357030153274536,
      "epoch": 2.5114132592298533,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004372414723208215,
      "loss": 4.9993,
      "mean_token_accuracy": 0.17850604355335237,
      "num_tokens": 55447587.0,
      "step": 25305
    },
    {
      "entropy": 5.444722604751587,
      "epoch": 2.5119094878920207,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004372169212826606,
      "loss": 5.1797,
      "mean_token_accuracy": 0.16973766088485717,
      "num_tokens": 55460587.0,
      "step": 25310
    },
    {
      "entropy": 5.356863975524902,
      "epoch": 2.512405716554188,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043719236622186976,
      "loss": 5.0027,
      "mean_token_accuracy": 0.1821530282497406,
      "num_tokens": 55471150.0,
      "step": 25315
    },
    {
      "entropy": 5.3341363906860355,
      "epoch": 2.5129019452163557,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004371678071390576,
      "loss": 4.9955,
      "mean_token_accuracy": 0.1771107167005539,
      "num_tokens": 55480368.0,
      "step": 25320
    },
    {
      "entropy": 5.407420253753662,
      "epoch": 2.513398173878523,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043714324403483333,
      "loss": 5.0714,
      "mean_token_accuracy": 0.17170219719409943,
      "num_tokens": 55491954.0,
      "step": 25325
    },
    {
      "entropy": 5.3862639427185055,
      "epoch": 2.5138944025406906,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043711867690980605,
      "loss": 5.0287,
      "mean_token_accuracy": 0.17252536118030548,
      "num_tokens": 55502560.0,
      "step": 25330
    },
    {
      "entropy": 5.329868030548096,
      "epoch": 2.5143906312028586,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043709410576458477,
      "loss": 5.012,
      "mean_token_accuracy": 0.17553719133138657,
      "num_tokens": 55514571.0,
      "step": 25335
    },
    {
      "entropy": 5.385152530670166,
      "epoch": 2.514886859865026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043706953059977907,
      "loss": 5.0464,
      "mean_token_accuracy": 0.17844061702489852,
      "num_tokens": 55525206.0,
      "step": 25340
    },
    {
      "entropy": 5.364814567565918,
      "epoch": 2.5153830885271935,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043704495141599816,
      "loss": 5.0519,
      "mean_token_accuracy": 0.17021530568599702,
      "num_tokens": 55536551.0,
      "step": 25345
    },
    {
      "entropy": 5.356748771667481,
      "epoch": 2.515879317189361,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004370203682138517,
      "loss": 5.0194,
      "mean_token_accuracy": 0.1672179639339447,
      "num_tokens": 55547842.0,
      "step": 25350
    },
    {
      "entropy": 5.424680757522583,
      "epoch": 2.5163755458515285,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004369957809939492,
      "loss": 4.9585,
      "mean_token_accuracy": 0.18766601085662843,
      "num_tokens": 55558248.0,
      "step": 25355
    },
    {
      "entropy": 5.334637403488159,
      "epoch": 2.516871774513696,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004369711897569004,
      "loss": 4.962,
      "mean_token_accuracy": 0.17433681786060334,
      "num_tokens": 55569195.0,
      "step": 25360
    },
    {
      "entropy": 5.347146654129029,
      "epoch": 2.5173680031758634,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004369465945033151,
      "loss": 4.9648,
      "mean_token_accuracy": 0.18254573345184327,
      "num_tokens": 55579649.0,
      "step": 25365
    },
    {
      "entropy": 5.286997556686401,
      "epoch": 2.517864231838031,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043692199523380325,
      "loss": 4.8573,
      "mean_token_accuracy": 0.18348218500614166,
      "num_tokens": 55589550.0,
      "step": 25370
    },
    {
      "entropy": 5.32035346031189,
      "epoch": 2.5183604605001984,
      "grad_norm": 1.0,
      "learning_rate": 0.0004368973919489748,
      "loss": 5.0059,
      "mean_token_accuracy": 0.1764977604150772,
      "num_tokens": 55600673.0,
      "step": 25375
    },
    {
      "entropy": 5.196442651748657,
      "epoch": 2.518856689162366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043687278464943986,
      "loss": 4.9765,
      "mean_token_accuracy": 0.17698698192834855,
      "num_tokens": 55612045.0,
      "step": 25380
    },
    {
      "entropy": 5.334692573547363,
      "epoch": 2.5193529178245333,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004368481733358088,
      "loss": 5.0458,
      "mean_token_accuracy": 0.17574527710676194,
      "num_tokens": 55623666.0,
      "step": 25385
    },
    {
      "entropy": 5.370660591125488,
      "epoch": 2.519849146486701,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004368235580086917,
      "loss": 5.0051,
      "mean_token_accuracy": 0.17973188012838365,
      "num_tokens": 55634236.0,
      "step": 25390
    },
    {
      "entropy": 5.273372936248779,
      "epoch": 2.5203453751488687,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043679893866869904,
      "loss": 4.9742,
      "mean_token_accuracy": 0.1761993944644928,
      "num_tokens": 55645412.0,
      "step": 25395
    },
    {
      "entropy": 5.302120971679687,
      "epoch": 2.520841603811036,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004367743153164414,
      "loss": 4.9698,
      "mean_token_accuracy": 0.1789318412542343,
      "num_tokens": 55655861.0,
      "step": 25400
    },
    {
      "entropy": 5.341976881027222,
      "epoch": 2.5213378324732036,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043674968795252926,
      "loss": 4.9486,
      "mean_token_accuracy": 0.1784791186451912,
      "num_tokens": 55665227.0,
      "step": 25405
    },
    {
      "entropy": 5.322778558731079,
      "epoch": 2.521834061135371,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004367250565775734,
      "loss": 4.9923,
      "mean_token_accuracy": 0.18115394711494445,
      "num_tokens": 55676104.0,
      "step": 25410
    },
    {
      "entropy": 5.332001352310181,
      "epoch": 2.5223302897975386,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004367004211921847,
      "loss": 5.0733,
      "mean_token_accuracy": 0.1731605127453804,
      "num_tokens": 55688222.0,
      "step": 25415
    },
    {
      "entropy": 5.373298406600952,
      "epoch": 2.522826518459706,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043667578179697387,
      "loss": 5.0074,
      "mean_token_accuracy": 0.17902628034353257,
      "num_tokens": 55700138.0,
      "step": 25420
    },
    {
      "entropy": 5.360687589645385,
      "epoch": 2.523322747121874,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004366511383925521,
      "loss": 4.9692,
      "mean_token_accuracy": 0.18138327300548554,
      "num_tokens": 55709444.0,
      "step": 25425
    },
    {
      "entropy": 5.409147071838379,
      "epoch": 2.5238189757840415,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004366264909795305,
      "loss": 5.1189,
      "mean_token_accuracy": 0.17114484310150146,
      "num_tokens": 55720327.0,
      "step": 25430
    },
    {
      "entropy": 5.273277521133423,
      "epoch": 2.524315204446209,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004366018395585201,
      "loss": 4.8853,
      "mean_token_accuracy": 0.18587644249200821,
      "num_tokens": 55730727.0,
      "step": 25435
    },
    {
      "entropy": 5.366585206985474,
      "epoch": 2.5248114331083764,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004365771841301323,
      "loss": 5.0478,
      "mean_token_accuracy": 0.179401496052742,
      "num_tokens": 55741568.0,
      "step": 25440
    },
    {
      "entropy": 5.312329530715942,
      "epoch": 2.525307661770544,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004365525246949786,
      "loss": 4.9881,
      "mean_token_accuracy": 0.17839843779802322,
      "num_tokens": 55752511.0,
      "step": 25445
    },
    {
      "entropy": 5.343187618255615,
      "epoch": 2.5258038904327114,
      "grad_norm": 1.0,
      "learning_rate": 0.00043652786125367035,
      "loss": 4.9946,
      "mean_token_accuracy": 0.1743820369243622,
      "num_tokens": 55763004.0,
      "step": 25450
    },
    {
      "entropy": 5.391305351257325,
      "epoch": 2.526300119094879,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043650319380681915,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1773167684674263,
      "num_tokens": 55773276.0,
      "step": 25455
    },
    {
      "entropy": 5.394506406784058,
      "epoch": 2.5267963477570463,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004364785223550368,
      "loss": 5.0191,
      "mean_token_accuracy": 0.1750926822423935,
      "num_tokens": 55784607.0,
      "step": 25460
    },
    {
      "entropy": 5.4255811214447025,
      "epoch": 2.527292576419214,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043645384689893504,
      "loss": 5.082,
      "mean_token_accuracy": 0.16866173595190048,
      "num_tokens": 55796730.0,
      "step": 25465
    },
    {
      "entropy": 5.3945684909820555,
      "epoch": 2.5277888050813813,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004364291674391258,
      "loss": 5.0475,
      "mean_token_accuracy": 0.16994307339191436,
      "num_tokens": 55806940.0,
      "step": 25470
    },
    {
      "entropy": 5.27814302444458,
      "epoch": 2.528285033743549,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000436404483976221,
      "loss": 4.9759,
      "mean_token_accuracy": 0.18138472586870194,
      "num_tokens": 55818229.0,
      "step": 25475
    },
    {
      "entropy": 5.342270994186402,
      "epoch": 2.5287812624057167,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043637979651083285,
      "loss": 5.0817,
      "mean_token_accuracy": 0.17315515130758286,
      "num_tokens": 55831100.0,
      "step": 25480
    },
    {
      "entropy": 5.449152994155884,
      "epoch": 2.529277491067884,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043635510504357343,
      "loss": 5.0114,
      "mean_token_accuracy": 0.17565738707780837,
      "num_tokens": 55841713.0,
      "step": 25485
    },
    {
      "entropy": 5.3574294090271,
      "epoch": 2.5297737197300516,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043633040957505515,
      "loss": 4.9911,
      "mean_token_accuracy": 0.1706048533320427,
      "num_tokens": 55852737.0,
      "step": 25490
    },
    {
      "entropy": 5.396551513671875,
      "epoch": 2.530269948392219,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004363057101058903,
      "loss": 5.0747,
      "mean_token_accuracy": 0.1749962717294693,
      "num_tokens": 55863471.0,
      "step": 25495
    },
    {
      "entropy": 5.320287132263184,
      "epoch": 2.5307661770543866,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004362810066366915,
      "loss": 4.9596,
      "mean_token_accuracy": 0.1762955293059349,
      "num_tokens": 55874305.0,
      "step": 25500
    },
    {
      "entropy": 5.389797306060791,
      "epoch": 2.531262405716554,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00043625629916807117,
      "loss": 5.0897,
      "mean_token_accuracy": 0.1775917574763298,
      "num_tokens": 55887400.0,
      "step": 25505
    },
    {
      "entropy": 5.336224794387817,
      "epoch": 2.531758634378722,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00043623158770064214,
      "loss": 5.0404,
      "mean_token_accuracy": 0.17568568587303163,
      "num_tokens": 55900052.0,
      "step": 25510
    },
    {
      "entropy": 5.543933296203614,
      "epoch": 2.5322548630408894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004362068722350172,
      "loss": 5.1809,
      "mean_token_accuracy": 0.16670975759625434,
      "num_tokens": 55911609.0,
      "step": 25515
    },
    {
      "entropy": 5.376875305175782,
      "epoch": 2.532751091703057,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043618215277180917,
      "loss": 4.9755,
      "mean_token_accuracy": 0.1797298699617386,
      "num_tokens": 55922873.0,
      "step": 25520
    },
    {
      "entropy": 5.316595935821534,
      "epoch": 2.5332473203652244,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043615742931163106,
      "loss": 5.0184,
      "mean_token_accuracy": 0.17487679421901703,
      "num_tokens": 55932161.0,
      "step": 25525
    },
    {
      "entropy": 5.391117429733276,
      "epoch": 2.533743549027392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000436132701855096,
      "loss": 5.1437,
      "mean_token_accuracy": 0.16799644529819488,
      "num_tokens": 55943601.0,
      "step": 25530
    },
    {
      "entropy": 5.301558303833008,
      "epoch": 2.5342397776895593,
      "grad_norm": 1.0,
      "learning_rate": 0.00043610797040281705,
      "loss": 5.0016,
      "mean_token_accuracy": 0.18092087656259537,
      "num_tokens": 55954981.0,
      "step": 25535
    },
    {
      "entropy": 5.367963409423828,
      "epoch": 2.534736006351727,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004360832349554077,
      "loss": 5.067,
      "mean_token_accuracy": 0.1721960961818695,
      "num_tokens": 55964407.0,
      "step": 25540
    },
    {
      "entropy": 5.332491779327393,
      "epoch": 2.5352322350138943,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004360584955134811,
      "loss": 4.9575,
      "mean_token_accuracy": 0.17816263139247895,
      "num_tokens": 55974634.0,
      "step": 25545
    },
    {
      "entropy": 5.3889251232147215,
      "epoch": 2.5357284636760618,
      "grad_norm": 1.0,
      "learning_rate": 0.00043603375207765103,
      "loss": 5.0439,
      "mean_token_accuracy": 0.16771730780601501,
      "num_tokens": 55986753.0,
      "step": 25550
    },
    {
      "entropy": 5.339296150207519,
      "epoch": 2.5362246923382292,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004360090046485308,
      "loss": 5.0019,
      "mean_token_accuracy": 0.1829095810651779,
      "num_tokens": 55997839.0,
      "step": 25555
    },
    {
      "entropy": 5.400812864303589,
      "epoch": 2.536720921000397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004359842532267342,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18129017651081086,
      "num_tokens": 56008765.0,
      "step": 25560
    },
    {
      "entropy": 5.364564704895019,
      "epoch": 2.5372171496625646,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043595949781287505,
      "loss": 5.0397,
      "mean_token_accuracy": 0.180536687374115,
      "num_tokens": 56020077.0,
      "step": 25565
    },
    {
      "entropy": 5.318691968917847,
      "epoch": 2.537713378324732,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043593473840756715,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18097634017467498,
      "num_tokens": 56031298.0,
      "step": 25570
    },
    {
      "entropy": 5.380969381332397,
      "epoch": 2.5382096069868996,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004359099750114245,
      "loss": 5.0444,
      "mean_token_accuracy": 0.17551231980323792,
      "num_tokens": 56043955.0,
      "step": 25575
    },
    {
      "entropy": 5.323317670822144,
      "epoch": 2.538705835649067,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004358852076250613,
      "loss": 4.9945,
      "mean_token_accuracy": 0.17670984715223312,
      "num_tokens": 56055734.0,
      "step": 25580
    },
    {
      "entropy": 5.421460247039795,
      "epoch": 2.5392020643112345,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043586043624909153,
      "loss": 5.0306,
      "mean_token_accuracy": 0.17833355963230133,
      "num_tokens": 56065472.0,
      "step": 25585
    },
    {
      "entropy": 5.42573299407959,
      "epoch": 2.539698292973402,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004358356608841296,
      "loss": 5.0407,
      "mean_token_accuracy": 0.17816593050956725,
      "num_tokens": 56075691.0,
      "step": 25590
    },
    {
      "entropy": 5.334874963760376,
      "epoch": 2.54019452163557,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043581088153078984,
      "loss": 5.035,
      "mean_token_accuracy": 0.1768227308988571,
      "num_tokens": 56085900.0,
      "step": 25595
    },
    {
      "entropy": 5.393773174285888,
      "epoch": 2.5406907502977374,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043578609818968676,
      "loss": 5.0218,
      "mean_token_accuracy": 0.17801492512226105,
      "num_tokens": 56097675.0,
      "step": 25600
    },
    {
      "entropy": 5.444300317764283,
      "epoch": 2.541186978959905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043576131086143496,
      "loss": 5.0989,
      "mean_token_accuracy": 0.17001542448997498,
      "num_tokens": 56108102.0,
      "step": 25605
    },
    {
      "entropy": 5.276220846176147,
      "epoch": 2.5416832076220723,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043573651954664897,
      "loss": 4.9098,
      "mean_token_accuracy": 0.18415872305631636,
      "num_tokens": 56118489.0,
      "step": 25610
    },
    {
      "entropy": 5.269548988342285,
      "epoch": 2.54217943628424,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043571172424594363,
      "loss": 4.9181,
      "mean_token_accuracy": 0.18525421768426895,
      "num_tokens": 56127949.0,
      "step": 25615
    },
    {
      "entropy": 5.2607392311096195,
      "epoch": 2.5426756649464073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004356869249599339,
      "loss": 4.9564,
      "mean_token_accuracy": 0.17873289287090302,
      "num_tokens": 56138209.0,
      "step": 25620
    },
    {
      "entropy": 5.334210205078125,
      "epoch": 2.5431718936085748,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004356621216892347,
      "loss": 5.05,
      "mean_token_accuracy": 0.18135814666748046,
      "num_tokens": 56149005.0,
      "step": 25625
    },
    {
      "entropy": 5.395855140686035,
      "epoch": 2.5436681222707422,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043563731443446095,
      "loss": 4.9654,
      "mean_token_accuracy": 0.18047747313976287,
      "num_tokens": 56159406.0,
      "step": 25630
    },
    {
      "entropy": 5.3468328475952145,
      "epoch": 2.5441643509329097,
      "grad_norm": 1.0,
      "learning_rate": 0.00043561250319622806,
      "loss": 5.0218,
      "mean_token_accuracy": 0.17628076672554016,
      "num_tokens": 56169998.0,
      "step": 25635
    },
    {
      "entropy": 5.416035556793213,
      "epoch": 2.544660579595077,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043558768797515116,
      "loss": 5.1078,
      "mean_token_accuracy": 0.1709464505314827,
      "num_tokens": 56181096.0,
      "step": 25640
    },
    {
      "entropy": 5.352651071548462,
      "epoch": 2.5451568082572447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043556286877184566,
      "loss": 5.012,
      "mean_token_accuracy": 0.175172159075737,
      "num_tokens": 56191909.0,
      "step": 25645
    },
    {
      "entropy": 5.360966682434082,
      "epoch": 2.5456530369194126,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004355380455869269,
      "loss": 5.0387,
      "mean_token_accuracy": 0.17315662950277327,
      "num_tokens": 56202725.0,
      "step": 25650
    },
    {
      "entropy": 5.373331785202026,
      "epoch": 2.54614926558158,
      "grad_norm": 1.125,
      "learning_rate": 0.0004355132184210106,
      "loss": 5.0085,
      "mean_token_accuracy": 0.1807647556066513,
      "num_tokens": 56213596.0,
      "step": 25655
    },
    {
      "entropy": 5.3203834056854244,
      "epoch": 2.5466454942437475,
      "grad_norm": 1.0,
      "learning_rate": 0.0004354883872747124,
      "loss": 5.0065,
      "mean_token_accuracy": 0.1776125282049179,
      "num_tokens": 56224416.0,
      "step": 25660
    },
    {
      "entropy": 5.316751909255982,
      "epoch": 2.547141722905915,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043546355214864796,
      "loss": 5.0225,
      "mean_token_accuracy": 0.1745956003665924,
      "num_tokens": 56234325.0,
      "step": 25665
    },
    {
      "entropy": 5.369523239135742,
      "epoch": 2.5476379515680825,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004354387130434332,
      "loss": 4.9992,
      "mean_token_accuracy": 0.17702091932296754,
      "num_tokens": 56244178.0,
      "step": 25670
    },
    {
      "entropy": 5.378337764739991,
      "epoch": 2.54813418023025,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004354138699596841,
      "loss": 5.0423,
      "mean_token_accuracy": 0.17619505077600478,
      "num_tokens": 56256036.0,
      "step": 25675
    },
    {
      "entropy": 5.342321252822876,
      "epoch": 2.548630408892418,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004353890228980167,
      "loss": 4.9965,
      "mean_token_accuracy": 0.17922209203243256,
      "num_tokens": 56266203.0,
      "step": 25680
    },
    {
      "entropy": 5.3255023002624515,
      "epoch": 2.5491266375545854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004353641718590471,
      "loss": 5.0204,
      "mean_token_accuracy": 0.17505348920822145,
      "num_tokens": 56276958.0,
      "step": 25685
    },
    {
      "entropy": 5.3345733165740965,
      "epoch": 2.549622866216753,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043533931684339156,
      "loss": 5.0247,
      "mean_token_accuracy": 0.17915854454040528,
      "num_tokens": 56287941.0,
      "step": 25690
    },
    {
      "entropy": 5.292431497573853,
      "epoch": 2.5501190948789203,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004353144578516665,
      "loss": 4.9338,
      "mean_token_accuracy": 0.18261779248714446,
      "num_tokens": 56299390.0,
      "step": 25695
    },
    {
      "entropy": 5.325663232803345,
      "epoch": 2.5506153235410878,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043528959488448835,
      "loss": 4.9855,
      "mean_token_accuracy": 0.17436494678258896,
      "num_tokens": 56310762.0,
      "step": 25700
    },
    {
      "entropy": 5.397273349761963,
      "epoch": 2.5511115522032553,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043526472794247365,
      "loss": 4.9916,
      "mean_token_accuracy": 0.17731835842132568,
      "num_tokens": 56322522.0,
      "step": 25705
    },
    {
      "entropy": 5.344391489028931,
      "epoch": 2.5516077808654227,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043523985702623896,
      "loss": 5.0627,
      "mean_token_accuracy": 0.17083369046449662,
      "num_tokens": 56334124.0,
      "step": 25710
    },
    {
      "entropy": 5.317876100540161,
      "epoch": 2.55210400952759,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043521498213640113,
      "loss": 4.9902,
      "mean_token_accuracy": 0.17726699113845826,
      "num_tokens": 56344761.0,
      "step": 25715
    },
    {
      "entropy": 5.2973840713500975,
      "epoch": 2.5526002381897577,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043519010327357704,
      "loss": 4.9523,
      "mean_token_accuracy": 0.1827906623482704,
      "num_tokens": 56355518.0,
      "step": 25720
    },
    {
      "entropy": 5.318971729278564,
      "epoch": 2.553096466851925,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004351652204383836,
      "loss": 5.0548,
      "mean_token_accuracy": 0.17720222026109694,
      "num_tokens": 56366307.0,
      "step": 25725
    },
    {
      "entropy": 5.353834533691407,
      "epoch": 2.5535926955140926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004351403336314377,
      "loss": 4.9541,
      "mean_token_accuracy": 0.1816214144229889,
      "num_tokens": 56377065.0,
      "step": 25730
    },
    {
      "entropy": 5.3970051288604735,
      "epoch": 2.5540889241762605,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004351154428533566,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18147196024656295,
      "num_tokens": 56387741.0,
      "step": 25735
    },
    {
      "entropy": 5.361698865890503,
      "epoch": 2.554585152838428,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004350905481047576,
      "loss": 5.1919,
      "mean_token_accuracy": 0.16766097247600556,
      "num_tokens": 56398927.0,
      "step": 25740
    },
    {
      "entropy": 5.440524101257324,
      "epoch": 2.5550813815005955,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004350656493862579,
      "loss": 5.127,
      "mean_token_accuracy": 0.17269605547189712,
      "num_tokens": 56408791.0,
      "step": 25745
    },
    {
      "entropy": 5.356230354309082,
      "epoch": 2.555577610162763,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043504074669847517,
      "loss": 4.9517,
      "mean_token_accuracy": 0.1849681556224823,
      "num_tokens": 56419206.0,
      "step": 25750
    },
    {
      "entropy": 5.457614517211914,
      "epoch": 2.5560738388249304,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004350158400420266,
      "loss": 5.1484,
      "mean_token_accuracy": 0.16696566939353943,
      "num_tokens": 56431774.0,
      "step": 25755
    },
    {
      "entropy": 5.397734880447388,
      "epoch": 2.556570067487098,
      "grad_norm": 1.125,
      "learning_rate": 0.00043499092941753003,
      "loss": 5.0527,
      "mean_token_accuracy": 0.17455431818962097,
      "num_tokens": 56442048.0,
      "step": 25760
    },
    {
      "entropy": 5.342829656600952,
      "epoch": 2.557066296149266,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043496601482560327,
      "loss": 5.0207,
      "mean_token_accuracy": 0.1718833863735199,
      "num_tokens": 56453347.0,
      "step": 25765
    },
    {
      "entropy": 5.318601322174072,
      "epoch": 2.5575625248114333,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043494109626686394,
      "loss": 4.9877,
      "mean_token_accuracy": 0.18545232117176055,
      "num_tokens": 56465323.0,
      "step": 25770
    },
    {
      "entropy": 5.331633710861206,
      "epoch": 2.558058753473601,
      "grad_norm": 1.125,
      "learning_rate": 0.00043491617374193017,
      "loss": 5.028,
      "mean_token_accuracy": 0.17558802366256715,
      "num_tokens": 56474704.0,
      "step": 25775
    },
    {
      "entropy": 5.320078277587891,
      "epoch": 2.5585549821357683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004348912472514197,
      "loss": 5.0541,
      "mean_token_accuracy": 0.1803963303565979,
      "num_tokens": 56485028.0,
      "step": 25780
    },
    {
      "entropy": 5.343905830383301,
      "epoch": 2.5590512107979357,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043486631679595097,
      "loss": 4.9359,
      "mean_token_accuracy": 0.18431492000818253,
      "num_tokens": 56496169.0,
      "step": 25785
    },
    {
      "entropy": 5.25052547454834,
      "epoch": 2.559547439460103,
      "grad_norm": 1.125,
      "learning_rate": 0.0004348413823761421,
      "loss": 4.9393,
      "mean_token_accuracy": 0.18284873068332672,
      "num_tokens": 56506324.0,
      "step": 25790
    },
    {
      "entropy": 5.273857688903808,
      "epoch": 2.5600436681222707,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004348164439926112,
      "loss": 4.9226,
      "mean_token_accuracy": 0.18124505132436752,
      "num_tokens": 56516901.0,
      "step": 25795
    },
    {
      "entropy": 5.275978660583496,
      "epoch": 2.560539896784438,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043479150164597695,
      "loss": 4.9098,
      "mean_token_accuracy": 0.18358195722103118,
      "num_tokens": 56527241.0,
      "step": 25800
    },
    {
      "entropy": 5.402507734298706,
      "epoch": 2.5610361254466056,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043476655533685783,
      "loss": 5.1246,
      "mean_token_accuracy": 0.17076703011989594,
      "num_tokens": 56539078.0,
      "step": 25805
    },
    {
      "entropy": 5.307856035232544,
      "epoch": 2.561532354108773,
      "grad_norm": 1.0,
      "learning_rate": 0.00043474160506587234,
      "loss": 4.9527,
      "mean_token_accuracy": 0.17884763926267624,
      "num_tokens": 56551141.0,
      "step": 25810
    },
    {
      "entropy": 5.3343953609466555,
      "epoch": 2.5620285827709406,
      "grad_norm": 1.0,
      "learning_rate": 0.0004347166508336393,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17205391675233841,
      "num_tokens": 56561836.0,
      "step": 25815
    },
    {
      "entropy": 5.362556457519531,
      "epoch": 2.5625248114331085,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043469169264077754,
      "loss": 4.9904,
      "mean_token_accuracy": 0.17636855840682983,
      "num_tokens": 56572699.0,
      "step": 25820
    },
    {
      "entropy": 5.371920108795166,
      "epoch": 2.563021040095276,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004346667304879058,
      "loss": 5.0612,
      "mean_token_accuracy": 0.17566222697496414,
      "num_tokens": 56582296.0,
      "step": 25825
    },
    {
      "entropy": 5.328991937637329,
      "epoch": 2.5635172687574435,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043464176437564317,
      "loss": 5.0051,
      "mean_token_accuracy": 0.17618557512760163,
      "num_tokens": 56592907.0,
      "step": 25830
    },
    {
      "entropy": 5.302520847320556,
      "epoch": 2.564013497419611,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004346167943046089,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18234848529100417,
      "num_tokens": 56604995.0,
      "step": 25835
    },
    {
      "entropy": 5.320268535614014,
      "epoch": 2.5645097260817784,
      "grad_norm": 0.953125,
      "learning_rate": 0.000434591820275422,
      "loss": 4.9621,
      "mean_token_accuracy": 0.17935912013053895,
      "num_tokens": 56614809.0,
      "step": 25840
    },
    {
      "entropy": 5.398555183410645,
      "epoch": 2.565005954743946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004345668422887018,
      "loss": 5.0387,
      "mean_token_accuracy": 0.1729423373937607,
      "num_tokens": 56624409.0,
      "step": 25845
    },
    {
      "entropy": 5.332552909851074,
      "epoch": 2.565502183406114,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004345418603450678,
      "loss": 5.0438,
      "mean_token_accuracy": 0.17377075105905532,
      "num_tokens": 56634596.0,
      "step": 25850
    },
    {
      "entropy": 5.320702648162841,
      "epoch": 2.5659984120682813,
      "grad_norm": 1.0,
      "learning_rate": 0.0004345168744451395,
      "loss": 5.0102,
      "mean_token_accuracy": 0.1785515382885933,
      "num_tokens": 56646710.0,
      "step": 25855
    },
    {
      "entropy": 5.331485366821289,
      "epoch": 2.5664946407304488,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004344918845895364,
      "loss": 4.9964,
      "mean_token_accuracy": 0.17985937893390655,
      "num_tokens": 56658427.0,
      "step": 25860
    },
    {
      "entropy": 5.375764608383179,
      "epoch": 2.5669908693926162,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004344668907788782,
      "loss": 5.0688,
      "mean_token_accuracy": 0.17296691983938217,
      "num_tokens": 56669039.0,
      "step": 25865
    },
    {
      "entropy": 5.32942762374878,
      "epoch": 2.5674870980547837,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004344418930137848,
      "loss": 4.9815,
      "mean_token_accuracy": 0.18216221481561662,
      "num_tokens": 56679063.0,
      "step": 25870
    },
    {
      "entropy": 5.358488512039185,
      "epoch": 2.567983326716951,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004344168912948759,
      "loss": 5.0266,
      "mean_token_accuracy": 0.17888330817222595,
      "num_tokens": 56690142.0,
      "step": 25875
    },
    {
      "entropy": 5.367408514022827,
      "epoch": 2.5684795553791187,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043439188562277174,
      "loss": 5.0466,
      "mean_token_accuracy": 0.17684487402439117,
      "num_tokens": 56700918.0,
      "step": 25880
    },
    {
      "entropy": 5.503118515014648,
      "epoch": 2.568975784041286,
      "grad_norm": 1.125,
      "learning_rate": 0.00043436687599809226,
      "loss": 5.1771,
      "mean_token_accuracy": 0.16807112991809844,
      "num_tokens": 56713520.0,
      "step": 25885
    },
    {
      "entropy": 5.322052621841431,
      "epoch": 2.5694720127034536,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043434186242145763,
      "loss": 5.0451,
      "mean_token_accuracy": 0.17597593516111373,
      "num_tokens": 56723428.0,
      "step": 25890
    },
    {
      "entropy": 5.358119773864746,
      "epoch": 2.569968241365621,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004343168448934881,
      "loss": 4.9846,
      "mean_token_accuracy": 0.17957958877086638,
      "num_tokens": 56734167.0,
      "step": 25895
    },
    {
      "entropy": 5.348465490341186,
      "epoch": 2.5704644700277886,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043429182341480423,
      "loss": 4.9411,
      "mean_token_accuracy": 0.1803985580801964,
      "num_tokens": 56745196.0,
      "step": 25900
    },
    {
      "entropy": 5.320831966400147,
      "epoch": 2.5709606986899565,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004342667979860263,
      "loss": 4.9532,
      "mean_token_accuracy": 0.17729734480381013,
      "num_tokens": 56756166.0,
      "step": 25905
    },
    {
      "entropy": 5.3132609844207765,
      "epoch": 2.571456927352124,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043424176860777495,
      "loss": 4.9784,
      "mean_token_accuracy": 0.17739156484603882,
      "num_tokens": 56766690.0,
      "step": 25910
    },
    {
      "entropy": 5.348473024368286,
      "epoch": 2.5719531560142914,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043421673528067095,
      "loss": 5.0434,
      "mean_token_accuracy": 0.17403752505779266,
      "num_tokens": 56776995.0,
      "step": 25915
    },
    {
      "entropy": 5.4306388854980465,
      "epoch": 2.572449384676459,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043419169800533496,
      "loss": 5.0985,
      "mean_token_accuracy": 0.17362914681434632,
      "num_tokens": 56788509.0,
      "step": 25920
    },
    {
      "entropy": 5.360856771469116,
      "epoch": 2.5729456133386264,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004341666567823879,
      "loss": 5.0724,
      "mean_token_accuracy": 0.16840238869190216,
      "num_tokens": 56799027.0,
      "step": 25925
    },
    {
      "entropy": 5.374106788635254,
      "epoch": 2.573441842000794,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004341416116124508,
      "loss": 5.0536,
      "mean_token_accuracy": 0.1732361286878586,
      "num_tokens": 56809796.0,
      "step": 25930
    },
    {
      "entropy": 5.359841680526733,
      "epoch": 2.5739380706629618,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004341165624961445,
      "loss": 5.0474,
      "mean_token_accuracy": 0.17920070737600327,
      "num_tokens": 56820523.0,
      "step": 25935
    },
    {
      "entropy": 5.3647698879241945,
      "epoch": 2.5744342993251292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043409150943409035,
      "loss": 5.0151,
      "mean_token_accuracy": 0.17312656491994857,
      "num_tokens": 56831668.0,
      "step": 25940
    },
    {
      "entropy": 5.271725654602051,
      "epoch": 2.5749305279872967,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043406645242690967,
      "loss": 4.8948,
      "mean_token_accuracy": 0.18153661787509917,
      "num_tokens": 56841950.0,
      "step": 25945
    },
    {
      "entropy": 5.323250722885132,
      "epoch": 2.575426756649464,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004340413914752237,
      "loss": 5.0301,
      "mean_token_accuracy": 0.1749444529414177,
      "num_tokens": 56852190.0,
      "step": 25950
    },
    {
      "entropy": 5.413641357421875,
      "epoch": 2.5759229853116317,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004340163265796539,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17491190284490585,
      "num_tokens": 56864255.0,
      "step": 25955
    },
    {
      "entropy": 5.373104810714722,
      "epoch": 2.576419213973799,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004339912577408218,
      "loss": 4.9766,
      "mean_token_accuracy": 0.18223365694284438,
      "num_tokens": 56874711.0,
      "step": 25960
    },
    {
      "entropy": 5.2646629333496096,
      "epoch": 2.5769154426359666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004339661849593492,
      "loss": 4.9812,
      "mean_token_accuracy": 0.17649797946214676,
      "num_tokens": 56885790.0,
      "step": 25965
    },
    {
      "entropy": 5.374449825286865,
      "epoch": 2.577411671298134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004339411082358577,
      "loss": 5.0285,
      "mean_token_accuracy": 0.17838821113109588,
      "num_tokens": 56896452.0,
      "step": 25970
    },
    {
      "entropy": 5.398177909851074,
      "epoch": 2.5779078999603016,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043391602757096923,
      "loss": 5.1262,
      "mean_token_accuracy": 0.17672208845615386,
      "num_tokens": 56908216.0,
      "step": 25975
    },
    {
      "entropy": 5.35813684463501,
      "epoch": 2.578404128622469,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043389094296530565,
      "loss": 5.0139,
      "mean_token_accuracy": 0.17012820690870284,
      "num_tokens": 56918346.0,
      "step": 25980
    },
    {
      "entropy": 5.384368515014648,
      "epoch": 2.5789003572846365,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004338658544194891,
      "loss": 5.1408,
      "mean_token_accuracy": 0.16706978976726533,
      "num_tokens": 56930950.0,
      "step": 25985
    },
    {
      "entropy": 5.398607778549194,
      "epoch": 2.5793965859468044,
      "grad_norm": 1.125,
      "learning_rate": 0.00043384076193414176,
      "loss": 5.0544,
      "mean_token_accuracy": 0.17630691081285477,
      "num_tokens": 56940622.0,
      "step": 25990
    },
    {
      "entropy": 5.340560388565064,
      "epoch": 2.579892814608972,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043381566550988564,
      "loss": 5.0019,
      "mean_token_accuracy": 0.18047042936086655,
      "num_tokens": 56952463.0,
      "step": 25995
    },
    {
      "entropy": 5.341096448898315,
      "epoch": 2.5803890432711394,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043379056514734334,
      "loss": 4.9601,
      "mean_token_accuracy": 0.17695607542991637,
      "num_tokens": 56961705.0,
      "step": 26000
    },
    {
      "entropy": 5.361843633651733,
      "epoch": 2.580885271933307,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004337654608471372,
      "loss": 5.0223,
      "mean_token_accuracy": 0.17322562038898467,
      "num_tokens": 56972359.0,
      "step": 26005
    },
    {
      "entropy": 5.359927558898926,
      "epoch": 2.5813815005954743,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043374035260988964,
      "loss": 5.0126,
      "mean_token_accuracy": 0.17880237847566605,
      "num_tokens": 56983342.0,
      "step": 26010
    },
    {
      "entropy": 5.313671493530274,
      "epoch": 2.581877729257642,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043371524043622344,
      "loss": 4.957,
      "mean_token_accuracy": 0.18748963326215745,
      "num_tokens": 56993859.0,
      "step": 26015
    },
    {
      "entropy": 5.401707792282105,
      "epoch": 2.5823739579198093,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004336901243267614,
      "loss": 5.0953,
      "mean_token_accuracy": 0.17503551691770552,
      "num_tokens": 57006631.0,
      "step": 26020
    },
    {
      "entropy": 5.414786767959595,
      "epoch": 2.582870186581977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000433665004282126,
      "loss": 5.0368,
      "mean_token_accuracy": 0.17597126066684723,
      "num_tokens": 57018121.0,
      "step": 26025
    },
    {
      "entropy": 5.326180791854858,
      "epoch": 2.5833664152441447,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043363988030294055,
      "loss": 4.9798,
      "mean_token_accuracy": 0.17860294580459596,
      "num_tokens": 57028127.0,
      "step": 26030
    },
    {
      "entropy": 5.328649950027466,
      "epoch": 2.583862643906312,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004336147523898279,
      "loss": 4.968,
      "mean_token_accuracy": 0.18258248269557953,
      "num_tokens": 57038523.0,
      "step": 26035
    },
    {
      "entropy": 5.288590908050537,
      "epoch": 2.5843588725684796,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004335896205434111,
      "loss": 4.9637,
      "mean_token_accuracy": 0.18038428872823714,
      "num_tokens": 57048641.0,
      "step": 26040
    },
    {
      "entropy": 5.285114049911499,
      "epoch": 2.584855101230647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043356448476431354,
      "loss": 4.9048,
      "mean_token_accuracy": 0.18195340484380723,
      "num_tokens": 57058794.0,
      "step": 26045
    },
    {
      "entropy": 5.383684253692627,
      "epoch": 2.5853513298928146,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004335393450531583,
      "loss": 5.0018,
      "mean_token_accuracy": 0.1775318667292595,
      "num_tokens": 57069814.0,
      "step": 26050
    },
    {
      "entropy": 5.452498912811279,
      "epoch": 2.585847558554982,
      "grad_norm": 1.109375,
      "learning_rate": 0.000433514201410569,
      "loss": 5.0908,
      "mean_token_accuracy": 0.1777885600924492,
      "num_tokens": 57081053.0,
      "step": 26055
    },
    {
      "entropy": 5.284225177764893,
      "epoch": 2.5863437872171495,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043348905383716905,
      "loss": 4.98,
      "mean_token_accuracy": 0.18145263195037842,
      "num_tokens": 57091386.0,
      "step": 26060
    },
    {
      "entropy": 5.338980484008789,
      "epoch": 2.586840015879317,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043346390233358213,
      "loss": 4.962,
      "mean_token_accuracy": 0.17838215082883835,
      "num_tokens": 57102659.0,
      "step": 26065
    },
    {
      "entropy": 5.405575895309449,
      "epoch": 2.5873362445414845,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004334387469004319,
      "loss": 5.0317,
      "mean_token_accuracy": 0.17703465670347213,
      "num_tokens": 57114257.0,
      "step": 26070
    },
    {
      "entropy": 5.334444999694824,
      "epoch": 2.5878324732036524,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004334135875383421,
      "loss": 5.0353,
      "mean_token_accuracy": 0.17945645451545716,
      "num_tokens": 57125037.0,
      "step": 26075
    },
    {
      "entropy": 5.301648235321045,
      "epoch": 2.58832870186582,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004333884242479367,
      "loss": 5.0285,
      "mean_token_accuracy": 0.17319685518741607,
      "num_tokens": 57137581.0,
      "step": 26080
    },
    {
      "entropy": 5.3400341987609865,
      "epoch": 2.5888249305279873,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043336325702983966,
      "loss": 4.9792,
      "mean_token_accuracy": 0.177817265689373,
      "num_tokens": 57148507.0,
      "step": 26085
    },
    {
      "entropy": 5.346319055557251,
      "epoch": 2.589321159190155,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043333808588467504,
      "loss": 4.9417,
      "mean_token_accuracy": 0.18192301094532012,
      "num_tokens": 57159518.0,
      "step": 26090
    },
    {
      "entropy": 5.298037195205689,
      "epoch": 2.5898173878523223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004333129108130671,
      "loss": 4.9689,
      "mean_token_accuracy": 0.18219191282987596,
      "num_tokens": 57170600.0,
      "step": 26095
    },
    {
      "entropy": 5.352877473831176,
      "epoch": 2.5903136165144898,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004332877318156402,
      "loss": 4.9989,
      "mean_token_accuracy": 0.17838962376117706,
      "num_tokens": 57181137.0,
      "step": 26100
    },
    {
      "entropy": 5.330720329284668,
      "epoch": 2.5908098451766572,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043326254889301856,
      "loss": 5.0006,
      "mean_token_accuracy": 0.1762576699256897,
      "num_tokens": 57191975.0,
      "step": 26105
    },
    {
      "entropy": 5.441181945800781,
      "epoch": 2.591306073838825,
      "grad_norm": 1.125,
      "learning_rate": 0.0004332373620458268,
      "loss": 5.0603,
      "mean_token_accuracy": 0.17269552499055862,
      "num_tokens": 57203139.0,
      "step": 26110
    },
    {
      "entropy": 5.36413893699646,
      "epoch": 2.5918023025009926,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043321217127468935,
      "loss": 5.0805,
      "mean_token_accuracy": 0.17039155811071396,
      "num_tokens": 57215542.0,
      "step": 26115
    },
    {
      "entropy": 5.3701701164245605,
      "epoch": 2.59229853116316,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043318697658023097,
      "loss": 4.9998,
      "mean_token_accuracy": 0.1783004507422447,
      "num_tokens": 57226525.0,
      "step": 26120
    },
    {
      "entropy": 5.330853986740112,
      "epoch": 2.5927947598253276,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004331617779630765,
      "loss": 4.9629,
      "mean_token_accuracy": 0.18362227976322174,
      "num_tokens": 57236120.0,
      "step": 26125
    },
    {
      "entropy": 5.240252923965454,
      "epoch": 2.593290988487495,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004331365754238507,
      "loss": 4.9106,
      "mean_token_accuracy": 0.18528916239738463,
      "num_tokens": 57246947.0,
      "step": 26130
    },
    {
      "entropy": 5.302016735076904,
      "epoch": 2.5937872171496625,
      "grad_norm": 1.0,
      "learning_rate": 0.00043311136896317855,
      "loss": 4.9587,
      "mean_token_accuracy": 0.1767682522535324,
      "num_tokens": 57258885.0,
      "step": 26135
    },
    {
      "entropy": 5.376466751098633,
      "epoch": 2.59428344581183,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043308615858168517,
      "loss": 5.0495,
      "mean_token_accuracy": 0.18132032603025436,
      "num_tokens": 57270047.0,
      "step": 26140
    },
    {
      "entropy": 5.405603551864624,
      "epoch": 2.5947796744739975,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004330609442799957,
      "loss": 5.1047,
      "mean_token_accuracy": 0.17591586261987685,
      "num_tokens": 57281405.0,
      "step": 26145
    },
    {
      "entropy": 5.361987638473511,
      "epoch": 2.595275903136165,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004330357260587355,
      "loss": 5.0145,
      "mean_token_accuracy": 0.1803205892443657,
      "num_tokens": 57292880.0,
      "step": 26150
    },
    {
      "entropy": 5.344546318054199,
      "epoch": 2.5957721317983324,
      "grad_norm": 1.0,
      "learning_rate": 0.0004330105039185297,
      "loss": 4.9845,
      "mean_token_accuracy": 0.1795876145362854,
      "num_tokens": 57304544.0,
      "step": 26155
    },
    {
      "entropy": 5.380104875564575,
      "epoch": 2.5962683604605004,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043298527786000393,
      "loss": 4.9746,
      "mean_token_accuracy": 0.1732372149825096,
      "num_tokens": 57315869.0,
      "step": 26160
    },
    {
      "entropy": 5.363059663772583,
      "epoch": 2.596764589122668,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004329600478837837,
      "loss": 5.0234,
      "mean_token_accuracy": 0.17413585931062697,
      "num_tokens": 57327933.0,
      "step": 26165
    },
    {
      "entropy": 5.36707034111023,
      "epoch": 2.5972608177848353,
      "grad_norm": 1.0,
      "learning_rate": 0.00043293481399049474,
      "loss": 4.9949,
      "mean_token_accuracy": 0.17449010014533997,
      "num_tokens": 57338941.0,
      "step": 26170
    },
    {
      "entropy": 5.329374027252197,
      "epoch": 2.597757046447003,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004329095761807626,
      "loss": 5.0174,
      "mean_token_accuracy": 0.17953706085681914,
      "num_tokens": 57350128.0,
      "step": 26175
    },
    {
      "entropy": 5.258301544189453,
      "epoch": 2.5982532751091703,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004328843344552133,
      "loss": 4.9795,
      "mean_token_accuracy": 0.1764381483197212,
      "num_tokens": 57361602.0,
      "step": 26180
    },
    {
      "entropy": 5.441393089294434,
      "epoch": 2.5987495037713377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004328590888144726,
      "loss": 5.0631,
      "mean_token_accuracy": 0.17122764736413956,
      "num_tokens": 57373093.0,
      "step": 26185
    },
    {
      "entropy": 5.4495494842529295,
      "epoch": 2.599245732433505,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004328338392591668,
      "loss": 5.0983,
      "mean_token_accuracy": 0.16849423795938492,
      "num_tokens": 57384799.0,
      "step": 26190
    },
    {
      "entropy": 5.2551929473876955,
      "epoch": 2.599741961095673,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043280858578992185,
      "loss": 4.945,
      "mean_token_accuracy": 0.18278848230838776,
      "num_tokens": 57395703.0,
      "step": 26195
    },
    {
      "entropy": 5.345175743103027,
      "epoch": 2.6002381897578406,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043278332840736397,
      "loss": 5.1241,
      "mean_token_accuracy": 0.17390927970409392,
      "num_tokens": 57406725.0,
      "step": 26200
    },
    {
      "entropy": 5.3578863620758055,
      "epoch": 2.600734418420008,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043275806711211957,
      "loss": 4.9666,
      "mean_token_accuracy": 0.18249986469745635,
      "num_tokens": 57417362.0,
      "step": 26205
    },
    {
      "entropy": 5.360586071014405,
      "epoch": 2.6012306470821756,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043273280190481503,
      "loss": 4.9931,
      "mean_token_accuracy": 0.17432407140731812,
      "num_tokens": 57429178.0,
      "step": 26210
    },
    {
      "entropy": 5.321940660476685,
      "epoch": 2.601726875744343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043270753278607687,
      "loss": 4.9946,
      "mean_token_accuracy": 0.17960734218358992,
      "num_tokens": 57439741.0,
      "step": 26215
    },
    {
      "entropy": 5.327145338058472,
      "epoch": 2.6022231044065105,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043268225975653176,
      "loss": 5.0571,
      "mean_token_accuracy": 0.17687734663486482,
      "num_tokens": 57451754.0,
      "step": 26220
    },
    {
      "entropy": 5.384578371047974,
      "epoch": 2.602719333068678,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004326569828168064,
      "loss": 5.0335,
      "mean_token_accuracy": 0.1744864135980606,
      "num_tokens": 57463396.0,
      "step": 26225
    },
    {
      "entropy": 5.275753831863403,
      "epoch": 2.6032155617308455,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004326317019675275,
      "loss": 4.9978,
      "mean_token_accuracy": 0.18001332581043245,
      "num_tokens": 57474023.0,
      "step": 26230
    },
    {
      "entropy": 5.35338397026062,
      "epoch": 2.603711790393013,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043260641720932214,
      "loss": 5.0032,
      "mean_token_accuracy": 0.1841842621564865,
      "num_tokens": 57484751.0,
      "step": 26235
    },
    {
      "entropy": 5.303620386123657,
      "epoch": 2.6042080190551804,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043258112854281726,
      "loss": 4.9682,
      "mean_token_accuracy": 0.17866575270891188,
      "num_tokens": 57496502.0,
      "step": 26240
    },
    {
      "entropy": 5.310681772232056,
      "epoch": 2.6047042477173483,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004325558359686399,
      "loss": 5.011,
      "mean_token_accuracy": 0.17954657971858978,
      "num_tokens": 57507111.0,
      "step": 26245
    },
    {
      "entropy": 5.3720873355865475,
      "epoch": 2.605200476379516,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043253053948741734,
      "loss": 5.0923,
      "mean_token_accuracy": 0.17080129086971282,
      "num_tokens": 57517844.0,
      "step": 26250
    },
    {
      "entropy": 5.4089820861816404,
      "epoch": 2.6056967050416833,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043250523909977684,
      "loss": 5.062,
      "mean_token_accuracy": 0.17682237774133683,
      "num_tokens": 57528047.0,
      "step": 26255
    },
    {
      "entropy": 5.378644180297852,
      "epoch": 2.6061929337038507,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004324799348063457,
      "loss": 5.0025,
      "mean_token_accuracy": 0.1758698433637619,
      "num_tokens": 57538489.0,
      "step": 26260
    },
    {
      "entropy": 5.295073127746582,
      "epoch": 2.606689162366018,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004324546266077518,
      "loss": 5.0043,
      "mean_token_accuracy": 0.18263305872678756,
      "num_tokens": 57549741.0,
      "step": 26265
    },
    {
      "entropy": 5.293834543228149,
      "epoch": 2.6071853910281857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004324293145046222,
      "loss": 4.9291,
      "mean_token_accuracy": 0.18612911999225618,
      "num_tokens": 57559803.0,
      "step": 26270
    },
    {
      "entropy": 5.373027801513672,
      "epoch": 2.607681619690353,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043240399849758484,
      "loss": 5.0137,
      "mean_token_accuracy": 0.17600320279598236,
      "num_tokens": 57570609.0,
      "step": 26275
    },
    {
      "entropy": 5.450855779647827,
      "epoch": 2.608177848352521,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004323786785872675,
      "loss": 5.0364,
      "mean_token_accuracy": 0.18124636858701706,
      "num_tokens": 57580934.0,
      "step": 26280
    },
    {
      "entropy": 5.3710973262786865,
      "epoch": 2.6086740770146886,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004323533547742981,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1743011951446533,
      "num_tokens": 57594671.0,
      "step": 26285
    },
    {
      "entropy": 5.350666189193726,
      "epoch": 2.609170305676856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004323280270593046,
      "loss": 5.067,
      "mean_token_accuracy": 0.17610003799200058,
      "num_tokens": 57605582.0,
      "step": 26290
    },
    {
      "entropy": 5.354444599151611,
      "epoch": 2.6096665343390235,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043230269544291503,
      "loss": 4.9959,
      "mean_token_accuracy": 0.17640975266695022,
      "num_tokens": 57617185.0,
      "step": 26295
    },
    {
      "entropy": 5.387395334243775,
      "epoch": 2.610162763001191,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004322773599257575,
      "loss": 4.9555,
      "mean_token_accuracy": 0.17840996384620667,
      "num_tokens": 57627205.0,
      "step": 26300
    },
    {
      "entropy": 5.274909543991089,
      "epoch": 2.6106589916633585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043225202050846043,
      "loss": 4.9395,
      "mean_token_accuracy": 0.18275029361248016,
      "num_tokens": 57639496.0,
      "step": 26305
    },
    {
      "entropy": 5.375122785568237,
      "epoch": 2.611155220325526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000432226677191652,
      "loss": 5.0715,
      "mean_token_accuracy": 0.17126695215702056,
      "num_tokens": 57650345.0,
      "step": 26310
    },
    {
      "entropy": 5.311539888381958,
      "epoch": 2.6116514489876934,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043220132997596084,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18066176772117615,
      "num_tokens": 57661409.0,
      "step": 26315
    },
    {
      "entropy": 5.437584972381591,
      "epoch": 2.612147677649861,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004321759788620155,
      "loss": 5.1215,
      "mean_token_accuracy": 0.16815848872065545,
      "num_tokens": 57672639.0,
      "step": 26320
    },
    {
      "entropy": 5.384598064422607,
      "epoch": 2.6126439063120284,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043215062385044446,
      "loss": 5.0968,
      "mean_token_accuracy": 0.17177652865648269,
      "num_tokens": 57682249.0,
      "step": 26325
    },
    {
      "entropy": 5.312732076644897,
      "epoch": 2.613140134974196,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043212526494187665,
      "loss": 4.9267,
      "mean_token_accuracy": 0.18890145421028137,
      "num_tokens": 57692319.0,
      "step": 26330
    },
    {
      "entropy": 5.407537174224854,
      "epoch": 2.6136363636363638,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004320999021369408,
      "loss": 5.1019,
      "mean_token_accuracy": 0.17523859292268754,
      "num_tokens": 57702334.0,
      "step": 26335
    },
    {
      "entropy": 5.339302444458008,
      "epoch": 2.6141325922985312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043207453543626587,
      "loss": 4.9808,
      "mean_token_accuracy": 0.17948753982782364,
      "num_tokens": 57712407.0,
      "step": 26340
    },
    {
      "entropy": 5.279342794418335,
      "epoch": 2.6146288209606987,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043204916484048094,
      "loss": 4.9586,
      "mean_token_accuracy": 0.18243390023708345,
      "num_tokens": 57724252.0,
      "step": 26345
    },
    {
      "entropy": 5.354496955871582,
      "epoch": 2.615125049622866,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043202379035021515,
      "loss": 5.0379,
      "mean_token_accuracy": 0.17275366485118865,
      "num_tokens": 57735033.0,
      "step": 26350
    },
    {
      "entropy": 5.427433347702026,
      "epoch": 2.6156212782850337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004319984119660977,
      "loss": 4.9874,
      "mean_token_accuracy": 0.18056558519601823,
      "num_tokens": 57746182.0,
      "step": 26355
    },
    {
      "entropy": 5.333070278167725,
      "epoch": 2.616117506947201,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043197302968875795,
      "loss": 4.9487,
      "mean_token_accuracy": 0.18047704845666884,
      "num_tokens": 57756479.0,
      "step": 26360
    },
    {
      "entropy": 5.372235345840454,
      "epoch": 2.616613735609369,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004319476435188254,
      "loss": 5.0703,
      "mean_token_accuracy": 0.16771066337823867,
      "num_tokens": 57767773.0,
      "step": 26365
    },
    {
      "entropy": 5.33076958656311,
      "epoch": 2.6171099642715365,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043192225345692933,
      "loss": 4.9787,
      "mean_token_accuracy": 0.1825307786464691,
      "num_tokens": 57777331.0,
      "step": 26370
    },
    {
      "entropy": 5.270195055007934,
      "epoch": 2.617606192933704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043189685950369963,
      "loss": 4.9822,
      "mean_token_accuracy": 0.18194133341312407,
      "num_tokens": 57788538.0,
      "step": 26375
    },
    {
      "entropy": 5.41522536277771,
      "epoch": 2.6181024215958715,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043187146165976577,
      "loss": 5.1262,
      "mean_token_accuracy": 0.17244563847780228,
      "num_tokens": 57799983.0,
      "step": 26380
    },
    {
      "entropy": 5.3747869491577145,
      "epoch": 2.618598650258039,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004318460599257578,
      "loss": 4.9515,
      "mean_token_accuracy": 0.17831269800662994,
      "num_tokens": 57810195.0,
      "step": 26385
    },
    {
      "entropy": 5.3008898258209225,
      "epoch": 2.6190948789202064,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043182065430230545,
      "loss": 4.8979,
      "mean_token_accuracy": 0.187922403216362,
      "num_tokens": 57821930.0,
      "step": 26390
    },
    {
      "entropy": 5.345117425918579,
      "epoch": 2.619591107582374,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004317952447900388,
      "loss": 4.9624,
      "mean_token_accuracy": 0.1771448954939842,
      "num_tokens": 57834272.0,
      "step": 26395
    },
    {
      "entropy": 5.341311502456665,
      "epoch": 2.6200873362445414,
      "grad_norm": 1.0,
      "learning_rate": 0.000431769831389588,
      "loss": 4.9936,
      "mean_token_accuracy": 0.1774161860346794,
      "num_tokens": 57845491.0,
      "step": 26400
    },
    {
      "entropy": 5.323802375793457,
      "epoch": 2.620583564906709,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004317444141015831,
      "loss": 5.0658,
      "mean_token_accuracy": 0.17536287009716034,
      "num_tokens": 57857816.0,
      "step": 26405
    },
    {
      "entropy": 5.3163471698760985,
      "epoch": 2.6210797935688763,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043171899292665457,
      "loss": 5.0173,
      "mean_token_accuracy": 0.17504720389842987,
      "num_tokens": 57867769.0,
      "step": 26410
    },
    {
      "entropy": 5.411194324493408,
      "epoch": 2.621576022231044,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004316935678654327,
      "loss": 5.1029,
      "mean_token_accuracy": 0.1703787311911583,
      "num_tokens": 57878027.0,
      "step": 26415
    },
    {
      "entropy": 5.324813032150269,
      "epoch": 2.6220722508932117,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043166813891854796,
      "loss": 4.9766,
      "mean_token_accuracy": 0.17886045277118684,
      "num_tokens": 57887938.0,
      "step": 26420
    },
    {
      "entropy": 5.364655256271362,
      "epoch": 2.622568479555379,
      "grad_norm": 1.0,
      "learning_rate": 0.000431642706086631,
      "loss": 4.9674,
      "mean_token_accuracy": 0.18400164246559142,
      "num_tokens": 57899375.0,
      "step": 26425
    },
    {
      "entropy": 5.3421204566955565,
      "epoch": 2.6230647082175467,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043161726937031243,
      "loss": 5.0296,
      "mean_token_accuracy": 0.17703863531351088,
      "num_tokens": 57910393.0,
      "step": 26430
    },
    {
      "entropy": 5.336081075668335,
      "epoch": 2.623560936879714,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000431591828770223,
      "loss": 5.0193,
      "mean_token_accuracy": 0.18205724358558656,
      "num_tokens": 57922236.0,
      "step": 26435
    },
    {
      "entropy": 5.339422702789307,
      "epoch": 2.6240571655418816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004315663842869938,
      "loss": 5.0104,
      "mean_token_accuracy": 0.18158810287714006,
      "num_tokens": 57931804.0,
      "step": 26440
    },
    {
      "entropy": 5.382856655120849,
      "epoch": 2.624553394204049,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043154093592125554,
      "loss": 5.0177,
      "mean_token_accuracy": 0.17538420855998993,
      "num_tokens": 57943259.0,
      "step": 26445
    },
    {
      "entropy": 5.26905574798584,
      "epoch": 2.625049622866217,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043151548367363943,
      "loss": 4.8913,
      "mean_token_accuracy": 0.18363696187734604,
      "num_tokens": 57952945.0,
      "step": 26450
    },
    {
      "entropy": 5.314249658584595,
      "epoch": 2.6255458515283845,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043149002754477655,
      "loss": 4.9969,
      "mean_token_accuracy": 0.17759459167718888,
      "num_tokens": 57964030.0,
      "step": 26455
    },
    {
      "entropy": 5.316673278808594,
      "epoch": 2.626042080190552,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004314645675352982,
      "loss": 4.9795,
      "mean_token_accuracy": 0.183671672642231,
      "num_tokens": 57974888.0,
      "step": 26460
    },
    {
      "entropy": 5.372714614868164,
      "epoch": 2.6265383088527194,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004314391036458357,
      "loss": 5.0058,
      "mean_token_accuracy": 0.17882008105516434,
      "num_tokens": 57984922.0,
      "step": 26465
    },
    {
      "entropy": 5.263065814971924,
      "epoch": 2.627034537514887,
      "grad_norm": 1.125,
      "learning_rate": 0.00043141363587702054,
      "loss": 4.9713,
      "mean_token_accuracy": 0.17856837064027786,
      "num_tokens": 57994881.0,
      "step": 26470
    },
    {
      "entropy": 5.330544948577881,
      "epoch": 2.6275307661770544,
      "grad_norm": 1.125,
      "learning_rate": 0.0004313881642294843,
      "loss": 5.0419,
      "mean_token_accuracy": 0.18092069774866104,
      "num_tokens": 58005335.0,
      "step": 26475
    },
    {
      "entropy": 5.325658750534058,
      "epoch": 2.628026994839222,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004313626887038585,
      "loss": 4.9492,
      "mean_token_accuracy": 0.18442580997943878,
      "num_tokens": 58016692.0,
      "step": 26480
    },
    {
      "entropy": 5.371195459365845,
      "epoch": 2.6285232235013893,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000431337209300775,
      "loss": 5.0545,
      "mean_token_accuracy": 0.17585684210062028,
      "num_tokens": 58028282.0,
      "step": 26485
    },
    {
      "entropy": 5.332860660552979,
      "epoch": 2.629019452163557,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004313117260208656,
      "loss": 4.9769,
      "mean_token_accuracy": 0.1815069079399109,
      "num_tokens": 58040108.0,
      "step": 26490
    },
    {
      "entropy": 5.293617916107178,
      "epoch": 2.6295156808257243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004312862388647622,
      "loss": 4.9686,
      "mean_token_accuracy": 0.18326624035835265,
      "num_tokens": 58050156.0,
      "step": 26495
    },
    {
      "entropy": 5.265159940719604,
      "epoch": 2.6300119094878918,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043126074783309687,
      "loss": 4.9765,
      "mean_token_accuracy": 0.17957437485456468,
      "num_tokens": 58060929.0,
      "step": 26500
    },
    {
      "entropy": 5.3437622547149655,
      "epoch": 2.6305081381500597,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004312352529265016,
      "loss": 5.0295,
      "mean_token_accuracy": 0.17196027636528016,
      "num_tokens": 58071684.0,
      "step": 26505
    },
    {
      "entropy": 5.413314247131348,
      "epoch": 2.631004366812227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043120975414560886,
      "loss": 5.0608,
      "mean_token_accuracy": 0.1753399670124054,
      "num_tokens": 58082730.0,
      "step": 26510
    },
    {
      "entropy": 5.41050763130188,
      "epoch": 2.6315005954743946,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043118425149105073,
      "loss": 5.117,
      "mean_token_accuracy": 0.17506113499403,
      "num_tokens": 58095278.0,
      "step": 26515
    },
    {
      "entropy": 5.461518049240112,
      "epoch": 2.631996824136562,
      "grad_norm": 1.0,
      "learning_rate": 0.0004311587449634597,
      "loss": 5.1836,
      "mean_token_accuracy": 0.16989944577217103,
      "num_tokens": 58107816.0,
      "step": 26520
    },
    {
      "entropy": 5.419684743881225,
      "epoch": 2.6324930527987296,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004311332345634683,
      "loss": 5.0681,
      "mean_token_accuracy": 0.16918123811483382,
      "num_tokens": 58117779.0,
      "step": 26525
    },
    {
      "entropy": 5.366522121429443,
      "epoch": 2.632989281460897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043110772029170916,
      "loss": 4.9966,
      "mean_token_accuracy": 0.1854146882891655,
      "num_tokens": 58128665.0,
      "step": 26530
    },
    {
      "entropy": 5.369458103179932,
      "epoch": 2.633485510123065,
      "grad_norm": 1.0,
      "learning_rate": 0.00043108220214881494,
      "loss": 5.0374,
      "mean_token_accuracy": 0.17379398494958878,
      "num_tokens": 58139658.0,
      "step": 26535
    },
    {
      "entropy": 5.435935020446777,
      "epoch": 2.6339817387852325,
      "grad_norm": 1.125,
      "learning_rate": 0.0004310566801354184,
      "loss": 5.0502,
      "mean_token_accuracy": 0.17891518026590347,
      "num_tokens": 58149980.0,
      "step": 26540
    },
    {
      "entropy": 5.326370096206665,
      "epoch": 2.6344779674474,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043103115425215246,
      "loss": 4.9102,
      "mean_token_accuracy": 0.18015470057725907,
      "num_tokens": 58159809.0,
      "step": 26545
    },
    {
      "entropy": 5.273324060440063,
      "epoch": 2.6349741961095674,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043100562449965016,
      "loss": 4.9982,
      "mean_token_accuracy": 0.17651764452457427,
      "num_tokens": 58169912.0,
      "step": 26550
    },
    {
      "entropy": 5.343437242507934,
      "epoch": 2.635470424771735,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043098009087854446,
      "loss": 5.0499,
      "mean_token_accuracy": 0.17375462204217912,
      "num_tokens": 58179949.0,
      "step": 26555
    },
    {
      "entropy": 5.369368267059326,
      "epoch": 2.6359666534339024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043095455338946863,
      "loss": 5.0036,
      "mean_token_accuracy": 0.1801259398460388,
      "num_tokens": 58189847.0,
      "step": 26560
    },
    {
      "entropy": 5.309590148925781,
      "epoch": 2.63646288209607,
      "grad_norm": 1.015625,
      "learning_rate": 0.000430929012033056,
      "loss": 4.9079,
      "mean_token_accuracy": 0.1834378570318222,
      "num_tokens": 58201514.0,
      "step": 26565
    },
    {
      "entropy": 5.319146776199341,
      "epoch": 2.6369591107582373,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043090346680993983,
      "loss": 4.9806,
      "mean_token_accuracy": 0.18762069940567017,
      "num_tokens": 58211940.0,
      "step": 26570
    },
    {
      "entropy": 5.379982137680054,
      "epoch": 2.6374553394204048,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004308779177207536,
      "loss": 5.0203,
      "mean_token_accuracy": 0.17559895664453506,
      "num_tokens": 58222260.0,
      "step": 26575
    },
    {
      "entropy": 5.325366687774658,
      "epoch": 2.6379515680825723,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004308523647661309,
      "loss": 4.9543,
      "mean_token_accuracy": 0.1803962141275406,
      "num_tokens": 58232649.0,
      "step": 26580
    },
    {
      "entropy": 5.309507513046265,
      "epoch": 2.6384477967447397,
      "grad_norm": 1.0,
      "learning_rate": 0.00043082680794670535,
      "loss": 4.9061,
      "mean_token_accuracy": 0.18533823043107986,
      "num_tokens": 58243373.0,
      "step": 26585
    },
    {
      "entropy": 5.265423822402954,
      "epoch": 2.6389440254069076,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043080124726311084,
      "loss": 4.9856,
      "mean_token_accuracy": 0.1832183614373207,
      "num_tokens": 58253863.0,
      "step": 26590
    },
    {
      "entropy": 5.326690769195556,
      "epoch": 2.639440254069075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004307756827159811,
      "loss": 5.0817,
      "mean_token_accuracy": 0.1792024314403534,
      "num_tokens": 58265326.0,
      "step": 26595
    },
    {
      "entropy": 5.410847234725952,
      "epoch": 2.6399364827312426,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043075011430594994,
      "loss": 5.0624,
      "mean_token_accuracy": 0.17806703299283982,
      "num_tokens": 58276006.0,
      "step": 26600
    },
    {
      "entropy": 5.299319791793823,
      "epoch": 2.64043271139341,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043072454203365175,
      "loss": 4.8915,
      "mean_token_accuracy": 0.1859491363167763,
      "num_tokens": 58285932.0,
      "step": 26605
    },
    {
      "entropy": 5.284597301483155,
      "epoch": 2.6409289400555775,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004306989658997204,
      "loss": 4.9769,
      "mean_token_accuracy": 0.18037571758031845,
      "num_tokens": 58297174.0,
      "step": 26610
    },
    {
      "entropy": 5.424824333190918,
      "epoch": 2.641425168717745,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043067338590479023,
      "loss": 5.1103,
      "mean_token_accuracy": 0.17177552729845047,
      "num_tokens": 58308509.0,
      "step": 26615
    },
    {
      "entropy": 5.344420909881592,
      "epoch": 2.641921397379913,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043064780204949543,
      "loss": 5.0289,
      "mean_token_accuracy": 0.17094283401966096,
      "num_tokens": 58318449.0,
      "step": 26620
    },
    {
      "entropy": 5.3539402961730955,
      "epoch": 2.6424176260420804,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004306222143344706,
      "loss": 4.9841,
      "mean_token_accuracy": 0.1789962336421013,
      "num_tokens": 58329759.0,
      "step": 26625
    },
    {
      "entropy": 5.304497528076172,
      "epoch": 2.642913854704248,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004305966227603503,
      "loss": 4.9538,
      "mean_token_accuracy": 0.17618658542633056,
      "num_tokens": 58339363.0,
      "step": 26630
    },
    {
      "entropy": 5.279342174530029,
      "epoch": 2.6434100833664154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004305710273277689,
      "loss": 5.004,
      "mean_token_accuracy": 0.17418816685676575,
      "num_tokens": 58349482.0,
      "step": 26635
    },
    {
      "entropy": 5.356459999084473,
      "epoch": 2.643906312028583,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043054542803736126,
      "loss": 5.03,
      "mean_token_accuracy": 0.17956554591655732,
      "num_tokens": 58359879.0,
      "step": 26640
    },
    {
      "entropy": 5.420480537414551,
      "epoch": 2.6444025406907503,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043051982488976216,
      "loss": 5.0688,
      "mean_token_accuracy": 0.17209936678409576,
      "num_tokens": 58370475.0,
      "step": 26645
    },
    {
      "entropy": 5.361661434173584,
      "epoch": 2.644898769352918,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004304942178856066,
      "loss": 4.9975,
      "mean_token_accuracy": 0.1762124106287956,
      "num_tokens": 58381992.0,
      "step": 26650
    },
    {
      "entropy": 5.302912473678589,
      "epoch": 2.6453949980150853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043046860702552944,
      "loss": 4.9587,
      "mean_token_accuracy": 0.18371070474386214,
      "num_tokens": 58391616.0,
      "step": 26655
    },
    {
      "entropy": 5.377100610733033,
      "epoch": 2.6458912266772527,
      "grad_norm": 1.125,
      "learning_rate": 0.0004304429923101658,
      "loss": 4.9745,
      "mean_token_accuracy": 0.17564579397439956,
      "num_tokens": 58401800.0,
      "step": 26660
    },
    {
      "entropy": 5.301822566986084,
      "epoch": 2.64638745533942,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000430417373740151,
      "loss": 4.9711,
      "mean_token_accuracy": 0.17767702788114548,
      "num_tokens": 58413376.0,
      "step": 26665
    },
    {
      "entropy": 5.214852046966553,
      "epoch": 2.6468836840015877,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004303917513161201,
      "loss": 4.9296,
      "mean_token_accuracy": 0.18637926280498504,
      "num_tokens": 58423772.0,
      "step": 26670
    },
    {
      "entropy": 5.353351974487305,
      "epoch": 2.6473799126637556,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004303661250387088,
      "loss": 5.0397,
      "mean_token_accuracy": 0.17136864215135575,
      "num_tokens": 58434838.0,
      "step": 26675
    },
    {
      "entropy": 5.399013948440552,
      "epoch": 2.647876141325923,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004303404949085522,
      "loss": 5.0536,
      "mean_token_accuracy": 0.1794099196791649,
      "num_tokens": 58445036.0,
      "step": 26680
    },
    {
      "entropy": 5.332508087158203,
      "epoch": 2.6483723699880906,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004303148609262861,
      "loss": 4.9902,
      "mean_token_accuracy": 0.1781172588467598,
      "num_tokens": 58455558.0,
      "step": 26685
    },
    {
      "entropy": 5.315635919570923,
      "epoch": 2.648868598650258,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004302892230925462,
      "loss": 5.0266,
      "mean_token_accuracy": 0.1709106981754303,
      "num_tokens": 58467716.0,
      "step": 26690
    },
    {
      "entropy": 5.31666669845581,
      "epoch": 2.6493648273124255,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043026358140796806,
      "loss": 5.0126,
      "mean_token_accuracy": 0.17554431706666945,
      "num_tokens": 58479421.0,
      "step": 26695
    },
    {
      "entropy": 5.292799949645996,
      "epoch": 2.649861055974593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004302379358731878,
      "loss": 4.9651,
      "mean_token_accuracy": 0.18439426869153977,
      "num_tokens": 58490115.0,
      "step": 26700
    },
    {
      "entropy": 5.320808172225952,
      "epoch": 2.6503572846367605,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004302122864888411,
      "loss": 4.9523,
      "mean_token_accuracy": 0.18280244022607803,
      "num_tokens": 58501711.0,
      "step": 26705
    },
    {
      "entropy": 5.396929311752319,
      "epoch": 2.6508535132989284,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004301866332555643,
      "loss": 5.0056,
      "mean_token_accuracy": 0.18217142224311828,
      "num_tokens": 58513441.0,
      "step": 26710
    },
    {
      "entropy": 5.3338381290435795,
      "epoch": 2.651349741961096,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043016097617399327,
      "loss": 4.9524,
      "mean_token_accuracy": 0.1851884812116623,
      "num_tokens": 58523940.0,
      "step": 26715
    },
    {
      "entropy": 5.307808876037598,
      "epoch": 2.6518459706232633,
      "grad_norm": 1.0,
      "learning_rate": 0.0004301353152447644,
      "loss": 5.0503,
      "mean_token_accuracy": 0.17385032922029495,
      "num_tokens": 58536354.0,
      "step": 26720
    },
    {
      "entropy": 5.355400609970093,
      "epoch": 2.652342199285431,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043010965046851404,
      "loss": 5.0471,
      "mean_token_accuracy": 0.17463019490242004,
      "num_tokens": 58547096.0,
      "step": 26725
    },
    {
      "entropy": 5.339429903030395,
      "epoch": 2.6528384279475983,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043008398184587847,
      "loss": 5.0292,
      "mean_token_accuracy": 0.17443648725748062,
      "num_tokens": 58557951.0,
      "step": 26730
    },
    {
      "entropy": 5.291129207611084,
      "epoch": 2.6533346566097658,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043005830937749447,
      "loss": 4.9375,
      "mean_token_accuracy": 0.18412037640810014,
      "num_tokens": 58569384.0,
      "step": 26735
    },
    {
      "entropy": 5.303788757324218,
      "epoch": 2.6538308852719332,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004300326330639984,
      "loss": 4.9891,
      "mean_token_accuracy": 0.1803268939256668,
      "num_tokens": 58580308.0,
      "step": 26740
    },
    {
      "entropy": 5.307197761535645,
      "epoch": 2.6543271139341007,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004300069529060272,
      "loss": 5.0227,
      "mean_token_accuracy": 0.17215766310691832,
      "num_tokens": 58591780.0,
      "step": 26745
    },
    {
      "entropy": 5.342324924468994,
      "epoch": 2.654823342596268,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042998126890421754,
      "loss": 5.0313,
      "mean_token_accuracy": 0.18283025324344634,
      "num_tokens": 58601477.0,
      "step": 26750
    },
    {
      "entropy": 5.357122850418091,
      "epoch": 2.6553195712584357,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042995558105920627,
      "loss": 5.0078,
      "mean_token_accuracy": 0.18209194242954255,
      "num_tokens": 58612531.0,
      "step": 26755
    },
    {
      "entropy": 5.360575485229492,
      "epoch": 2.6558157999206036,
      "grad_norm": 0.90625,
      "learning_rate": 0.00042992988937163066,
      "loss": 5.0604,
      "mean_token_accuracy": 0.17288485914468765,
      "num_tokens": 58624951.0,
      "step": 26760
    },
    {
      "entropy": 5.341796493530273,
      "epoch": 2.656312028582771,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004299041938421275,
      "loss": 5.0909,
      "mean_token_accuracy": 0.17488720864057541,
      "num_tokens": 58636484.0,
      "step": 26765
    },
    {
      "entropy": 5.378470373153687,
      "epoch": 2.6568082572449385,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042987849447133423,
      "loss": 4.9993,
      "mean_token_accuracy": 0.1754279300570488,
      "num_tokens": 58647386.0,
      "step": 26770
    },
    {
      "entropy": 5.353393411636352,
      "epoch": 2.657304485907106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004298527912598879,
      "loss": 5.0731,
      "mean_token_accuracy": 0.17426415532827377,
      "num_tokens": 58658656.0,
      "step": 26775
    },
    {
      "entropy": 5.344215297698975,
      "epoch": 2.6578007145692735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004298270842084261,
      "loss": 4.965,
      "mean_token_accuracy": 0.17814144790172576,
      "num_tokens": 58669901.0,
      "step": 26780
    },
    {
      "entropy": 5.363067197799682,
      "epoch": 2.658296943231441,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042980137331758626,
      "loss": 5.0874,
      "mean_token_accuracy": 0.1685776486992836,
      "num_tokens": 58680849.0,
      "step": 26785
    },
    {
      "entropy": 5.2992181301116945,
      "epoch": 2.6587931718936084,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042977565858800593,
      "loss": 4.9621,
      "mean_token_accuracy": 0.17603481262922288,
      "num_tokens": 58690642.0,
      "step": 26790
    },
    {
      "entropy": 5.288764429092407,
      "epoch": 2.6592894005557763,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042974994002032275,
      "loss": 4.9197,
      "mean_token_accuracy": 0.1842116132378578,
      "num_tokens": 58700480.0,
      "step": 26795
    },
    {
      "entropy": 5.357825994491577,
      "epoch": 2.659785629217944,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042972421761517456,
      "loss": 5.0535,
      "mean_token_accuracy": 0.16911129355430604,
      "num_tokens": 58712443.0,
      "step": 26800
    },
    {
      "entropy": 5.315442514419556,
      "epoch": 2.6602818578801113,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004296984913731991,
      "loss": 4.9628,
      "mean_token_accuracy": 0.17645656019449235,
      "num_tokens": 58721172.0,
      "step": 26805
    },
    {
      "entropy": 5.284584093093872,
      "epoch": 2.6607780865422788,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004296727612950344,
      "loss": 4.9821,
      "mean_token_accuracy": 0.18121573626995086,
      "num_tokens": 58731060.0,
      "step": 26810
    },
    {
      "entropy": 5.340480852127075,
      "epoch": 2.6612743152044462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004296470273813186,
      "loss": 5.0202,
      "mean_token_accuracy": 0.1793813869357109,
      "num_tokens": 58742189.0,
      "step": 26815
    },
    {
      "entropy": 5.407361745834351,
      "epoch": 2.6617705438666137,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004296212896326896,
      "loss": 5.026,
      "mean_token_accuracy": 0.18154580444097518,
      "num_tokens": 58753617.0,
      "step": 26820
    },
    {
      "entropy": 5.324463129043579,
      "epoch": 2.662266772528781,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042959554804978596,
      "loss": 5.0343,
      "mean_token_accuracy": 0.1815587654709816,
      "num_tokens": 58765299.0,
      "step": 26825
    },
    {
      "entropy": 5.3780364990234375,
      "epoch": 2.6627630011909487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042956980263324576,
      "loss": 4.9641,
      "mean_token_accuracy": 0.17953554391860962,
      "num_tokens": 58775280.0,
      "step": 26830
    },
    {
      "entropy": 5.3699346542358395,
      "epoch": 2.663259229853116,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004295440533837075,
      "loss": 4.9488,
      "mean_token_accuracy": 0.1817614883184433,
      "num_tokens": 58786078.0,
      "step": 26835
    },
    {
      "entropy": 5.315325736999512,
      "epoch": 2.6637554585152836,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042951830030180983,
      "loss": 5.0286,
      "mean_token_accuracy": 0.1787555605173111,
      "num_tokens": 58796665.0,
      "step": 26840
    },
    {
      "entropy": 5.353530216217041,
      "epoch": 2.6642516871774515,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042949254338819123,
      "loss": 5.0401,
      "mean_token_accuracy": 0.1746961370110512,
      "num_tokens": 58808150.0,
      "step": 26845
    },
    {
      "entropy": 5.376213073730469,
      "epoch": 2.664747915839619,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042946678264349036,
      "loss": 4.961,
      "mean_token_accuracy": 0.18173619657754897,
      "num_tokens": 58818278.0,
      "step": 26850
    },
    {
      "entropy": 5.3324113368988035,
      "epoch": 2.6652441445017865,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042944101806834627,
      "loss": 5.0241,
      "mean_token_accuracy": 0.1752706453204155,
      "num_tokens": 58828847.0,
      "step": 26855
    },
    {
      "entropy": 5.296335697174072,
      "epoch": 2.665740373163954,
      "grad_norm": 1.0,
      "learning_rate": 0.0004294152496633977,
      "loss": 4.9792,
      "mean_token_accuracy": 0.18000368177890777,
      "num_tokens": 58839698.0,
      "step": 26860
    },
    {
      "entropy": 5.288365125656128,
      "epoch": 2.6662366018261214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042938947742928357,
      "loss": 4.9698,
      "mean_token_accuracy": 0.1762043282389641,
      "num_tokens": 58850794.0,
      "step": 26865
    },
    {
      "entropy": 5.33875617980957,
      "epoch": 2.666732830488289,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042936370136664314,
      "loss": 5.044,
      "mean_token_accuracy": 0.1731829971075058,
      "num_tokens": 58861590.0,
      "step": 26870
    },
    {
      "entropy": 5.38937349319458,
      "epoch": 2.6672290591504564,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004293379214761156,
      "loss": 5.0871,
      "mean_token_accuracy": 0.17447810918092727,
      "num_tokens": 58873796.0,
      "step": 26875
    },
    {
      "entropy": 5.405997037887573,
      "epoch": 2.6677252878126243,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004293121377583401,
      "loss": 5.0237,
      "mean_token_accuracy": 0.1808563068509102,
      "num_tokens": 58885296.0,
      "step": 26880
    },
    {
      "entropy": 5.356408596038818,
      "epoch": 2.6682215164747918,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004292863502139561,
      "loss": 5.069,
      "mean_token_accuracy": 0.17674961239099501,
      "num_tokens": 58897764.0,
      "step": 26885
    },
    {
      "entropy": 5.366884183883667,
      "epoch": 2.6687177451369593,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004292605588436031,
      "loss": 5.0299,
      "mean_token_accuracy": 0.17171414494514464,
      "num_tokens": 58908964.0,
      "step": 26890
    },
    {
      "entropy": 5.222174596786499,
      "epoch": 2.6692139737991267,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042923476364792063,
      "loss": 4.899,
      "mean_token_accuracy": 0.18245141804218293,
      "num_tokens": 58919152.0,
      "step": 26895
    },
    {
      "entropy": 5.262160873413086,
      "epoch": 2.669710202461294,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004292089646275484,
      "loss": 5.0157,
      "mean_token_accuracy": 0.1776795879006386,
      "num_tokens": 58928849.0,
      "step": 26900
    },
    {
      "entropy": 5.404016923904419,
      "epoch": 2.6702064311234617,
      "grad_norm": 1.0,
      "learning_rate": 0.0004291831617831261,
      "loss": 5.1316,
      "mean_token_accuracy": 0.1705010265111923,
      "num_tokens": 58941560.0,
      "step": 26905
    },
    {
      "entropy": 5.421438884735108,
      "epoch": 2.670702659785629,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042915735511529356,
      "loss": 5.071,
      "mean_token_accuracy": 0.17121021449565887,
      "num_tokens": 58953365.0,
      "step": 26910
    },
    {
      "entropy": 5.373072147369385,
      "epoch": 2.6711988884477966,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004291315446246909,
      "loss": 4.9925,
      "mean_token_accuracy": 0.17828709185123442,
      "num_tokens": 58964602.0,
      "step": 26915
    },
    {
      "entropy": 5.3591203689575195,
      "epoch": 2.671695117109964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042910573031195804,
      "loss": 5.0161,
      "mean_token_accuracy": 0.17970840781927108,
      "num_tokens": 58976154.0,
      "step": 26920
    },
    {
      "entropy": 5.334455776214599,
      "epoch": 2.6721913457721316,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004290799121777351,
      "loss": 5.0378,
      "mean_token_accuracy": 0.17651089280843735,
      "num_tokens": 58985989.0,
      "step": 26925
    },
    {
      "entropy": 5.42476863861084,
      "epoch": 2.672687574434299,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004290540902226624,
      "loss": 5.0958,
      "mean_token_accuracy": 0.16848255395889283,
      "num_tokens": 58997080.0,
      "step": 26930
    },
    {
      "entropy": 5.361442279815674,
      "epoch": 2.673183803096467,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004290282644473802,
      "loss": 5.0126,
      "mean_token_accuracy": 0.1808811753988266,
      "num_tokens": 59008683.0,
      "step": 26935
    },
    {
      "entropy": 5.3305511474609375,
      "epoch": 2.6736800317586344,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004290024348525289,
      "loss": 4.934,
      "mean_token_accuracy": 0.18478200882673262,
      "num_tokens": 59019019.0,
      "step": 26940
    },
    {
      "entropy": 5.315293550491333,
      "epoch": 2.674176260420802,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004289766014387492,
      "loss": 4.9953,
      "mean_token_accuracy": 0.18365231156349182,
      "num_tokens": 59029706.0,
      "step": 26945
    },
    {
      "entropy": 5.290539693832398,
      "epoch": 2.6746724890829694,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004289507642066813,
      "loss": 4.9276,
      "mean_token_accuracy": 0.18377608209848403,
      "num_tokens": 59040208.0,
      "step": 26950
    },
    {
      "entropy": 5.302596616744995,
      "epoch": 2.675168717745137,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004289249231569664,
      "loss": 4.971,
      "mean_token_accuracy": 0.18875319361686707,
      "num_tokens": 59050909.0,
      "step": 26955
    },
    {
      "entropy": 5.260107803344726,
      "epoch": 2.6756649464073043,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042889907829024506,
      "loss": 4.9348,
      "mean_token_accuracy": 0.18286803960800171,
      "num_tokens": 59059834.0,
      "step": 26960
    },
    {
      "entropy": 5.352907466888428,
      "epoch": 2.6761611750694723,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042887322960715813,
      "loss": 5.0261,
      "mean_token_accuracy": 0.17426648437976838,
      "num_tokens": 59071348.0,
      "step": 26965
    },
    {
      "entropy": 5.316970920562744,
      "epoch": 2.6766574037316397,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004288473771083468,
      "loss": 4.9816,
      "mean_token_accuracy": 0.17679497450590134,
      "num_tokens": 59083053.0,
      "step": 26970
    },
    {
      "entropy": 5.358137702941894,
      "epoch": 2.677153632393807,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004288215207944519,
      "loss": 5.0364,
      "mean_token_accuracy": 0.17506891041994094,
      "num_tokens": 59093400.0,
      "step": 26975
    },
    {
      "entropy": 5.365369462966919,
      "epoch": 2.6776498610559747,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042879566066611474,
      "loss": 5.0303,
      "mean_token_accuracy": 0.17651850134134292,
      "num_tokens": 59103653.0,
      "step": 26980
    },
    {
      "entropy": 5.386867904663086,
      "epoch": 2.678146089718142,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004287697967239767,
      "loss": 5.039,
      "mean_token_accuracy": 0.17601328641176223,
      "num_tokens": 59116660.0,
      "step": 26985
    },
    {
      "entropy": 5.269305896759033,
      "epoch": 2.6786423183803096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000428743928968679,
      "loss": 4.9178,
      "mean_token_accuracy": 0.18236164152622222,
      "num_tokens": 59127424.0,
      "step": 26990
    },
    {
      "entropy": 5.332809782028198,
      "epoch": 2.679138547042477,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042871805740086317,
      "loss": 5.0277,
      "mean_token_accuracy": 0.17971001118421553,
      "num_tokens": 59137566.0,
      "step": 26995
    },
    {
      "entropy": 5.2890458583831785,
      "epoch": 2.6796347757046446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004286921820211707,
      "loss": 4.9237,
      "mean_token_accuracy": 0.18563691824674605,
      "num_tokens": 59148404.0,
      "step": 27000
    },
    {
      "epoch": 2.6796347757046446,
      "eval_entropy": 5.178363583891713,
      "eval_loss": 5.163365364074707,
      "eval_mean_token_accuracy": 0.17659405117553992,
      "eval_num_tokens": 59148404.0,
      "eval_runtime": 26.7469,
      "eval_samples_per_second": 1258.8,
      "eval_steps_per_second": 157.364,
      "step": 27000
    },
    {
      "entropy": 5.288960886001587,
      "epoch": 2.680131004366812,
      "grad_norm": 1.0,
      "learning_rate": 0.00042866630283024327,
      "loss": 4.9963,
      "mean_token_accuracy": 0.17510005980730056,
      "num_tokens": 59158097.0,
      "step": 27005
    },
    {
      "entropy": 5.276083183288574,
      "epoch": 2.6806272330289795,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004286404198287227,
      "loss": 4.8564,
      "mean_token_accuracy": 0.19332584142684936,
      "num_tokens": 59168940.0,
      "step": 27010
    },
    {
      "entropy": 5.336492395401001,
      "epoch": 2.681123461691147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004286145330172508,
      "loss": 4.9978,
      "mean_token_accuracy": 0.17555222660303116,
      "num_tokens": 59180231.0,
      "step": 27015
    },
    {
      "entropy": 5.407161903381348,
      "epoch": 2.681619690353315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004285886423964695,
      "loss": 5.0878,
      "mean_token_accuracy": 0.17224086672067643,
      "num_tokens": 59191565.0,
      "step": 27020
    },
    {
      "entropy": 5.337383651733399,
      "epoch": 2.6821159190154824,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004285627479670208,
      "loss": 5.0873,
      "mean_token_accuracy": 0.1754372239112854,
      "num_tokens": 59203394.0,
      "step": 27025
    },
    {
      "entropy": 5.246579027175903,
      "epoch": 2.68261214767765,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004285368497295468,
      "loss": 4.9409,
      "mean_token_accuracy": 0.17624286711215972,
      "num_tokens": 59213987.0,
      "step": 27030
    },
    {
      "entropy": 5.317457151412964,
      "epoch": 2.6831083763398174,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004285109476846898,
      "loss": 4.9705,
      "mean_token_accuracy": 0.18032313138246536,
      "num_tokens": 59225825.0,
      "step": 27035
    },
    {
      "entropy": 5.274135160446167,
      "epoch": 2.683604605001985,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004284850418330921,
      "loss": 4.9449,
      "mean_token_accuracy": 0.18662504553794862,
      "num_tokens": 59235978.0,
      "step": 27040
    },
    {
      "entropy": 5.394292402267456,
      "epoch": 2.6841008336641523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042845913217539613,
      "loss": 5.1414,
      "mean_token_accuracy": 0.16766363382339478,
      "num_tokens": 59247505.0,
      "step": 27045
    },
    {
      "entropy": 5.322884368896484,
      "epoch": 2.6845970623263202,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042843321871224426,
      "loss": 4.9239,
      "mean_token_accuracy": 0.18042558431625366,
      "num_tokens": 59258401.0,
      "step": 27050
    },
    {
      "entropy": 5.353007507324219,
      "epoch": 2.6850932909884877,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042840730144427927,
      "loss": 4.998,
      "mean_token_accuracy": 0.18175841718912125,
      "num_tokens": 59268068.0,
      "step": 27055
    },
    {
      "entropy": 5.301691198348999,
      "epoch": 2.685589519650655,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042838138037214375,
      "loss": 4.9962,
      "mean_token_accuracy": 0.17663196176290513,
      "num_tokens": 59279567.0,
      "step": 27060
    },
    {
      "entropy": 5.290392065048218,
      "epoch": 2.6860857483128227,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042835545549648054,
      "loss": 4.9327,
      "mean_token_accuracy": 0.18494328409433364,
      "num_tokens": 59289862.0,
      "step": 27065
    },
    {
      "entropy": 5.326085948944092,
      "epoch": 2.68658197697499,
      "grad_norm": 1.0,
      "learning_rate": 0.00042832952681793235,
      "loss": 4.9598,
      "mean_token_accuracy": 0.18604735732078553,
      "num_tokens": 59301304.0,
      "step": 27070
    },
    {
      "entropy": 5.304173231124878,
      "epoch": 2.6870782056371576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042830359433714243,
      "loss": 4.9911,
      "mean_token_accuracy": 0.18432389497756957,
      "num_tokens": 59312184.0,
      "step": 27075
    },
    {
      "entropy": 5.282737874984742,
      "epoch": 2.687574434299325,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042827765805475367,
      "loss": 4.8696,
      "mean_token_accuracy": 0.18102579116821288,
      "num_tokens": 59322688.0,
      "step": 27080
    },
    {
      "entropy": 5.274728059768677,
      "epoch": 2.6880706629614926,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042825171797140926,
      "loss": 4.959,
      "mean_token_accuracy": 0.1850545346736908,
      "num_tokens": 59333349.0,
      "step": 27085
    },
    {
      "entropy": 5.291688728332519,
      "epoch": 2.68856689162366,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004282257740877525,
      "loss": 4.9364,
      "mean_token_accuracy": 0.18712426126003265,
      "num_tokens": 59344473.0,
      "step": 27090
    },
    {
      "entropy": 5.340813684463501,
      "epoch": 2.6890631202858275,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004281998264044267,
      "loss": 5.0375,
      "mean_token_accuracy": 0.1809038758277893,
      "num_tokens": 59355993.0,
      "step": 27095
    },
    {
      "entropy": 5.384326505661011,
      "epoch": 2.689559348947995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004281738749220753,
      "loss": 5.0651,
      "mean_token_accuracy": 0.17431778013706206,
      "num_tokens": 59367182.0,
      "step": 27100
    },
    {
      "entropy": 5.417451000213623,
      "epoch": 2.690055577610163,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042814791964134184,
      "loss": 5.0491,
      "mean_token_accuracy": 0.17447710335254668,
      "num_tokens": 59379235.0,
      "step": 27105
    },
    {
      "entropy": 5.291768264770508,
      "epoch": 2.6905518062723304,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042812196056287,
      "loss": 4.9747,
      "mean_token_accuracy": 0.18832896202802657,
      "num_tokens": 59388514.0,
      "step": 27110
    },
    {
      "entropy": 5.275520658493042,
      "epoch": 2.691048034934498,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042809599768730357,
      "loss": 5.0199,
      "mean_token_accuracy": 0.18060334473848344,
      "num_tokens": 59399496.0,
      "step": 27115
    },
    {
      "entropy": 5.303924179077148,
      "epoch": 2.6915442635966653,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004280700310152861,
      "loss": 4.9841,
      "mean_token_accuracy": 0.18563860654830933,
      "num_tokens": 59411986.0,
      "step": 27120
    },
    {
      "entropy": 5.435899257659912,
      "epoch": 2.692040492258833,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004280440605474619,
      "loss": 5.1901,
      "mean_token_accuracy": 0.1697033241391182,
      "num_tokens": 59424232.0,
      "step": 27125
    },
    {
      "entropy": 5.43027024269104,
      "epoch": 2.6925367209210003,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004280180862844747,
      "loss": 5.0723,
      "mean_token_accuracy": 0.1723134309053421,
      "num_tokens": 59435201.0,
      "step": 27130
    },
    {
      "entropy": 5.370019149780274,
      "epoch": 2.693032949583168,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042799210822696865,
      "loss": 5.078,
      "mean_token_accuracy": 0.17135122269392014,
      "num_tokens": 59445552.0,
      "step": 27135
    },
    {
      "entropy": 5.4057127952575685,
      "epoch": 2.6935291782453357,
      "grad_norm": 1.125,
      "learning_rate": 0.000427966126375588,
      "loss": 5.0693,
      "mean_token_accuracy": 0.16819885969161988,
      "num_tokens": 59455634.0,
      "step": 27140
    },
    {
      "entropy": 5.381214141845703,
      "epoch": 2.694025406907503,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004279401407309771,
      "loss": 5.0276,
      "mean_token_accuracy": 0.17774800956249237,
      "num_tokens": 59467955.0,
      "step": 27145
    },
    {
      "entropy": 5.354256010055542,
      "epoch": 2.6945216355696706,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004279141512937802,
      "loss": 5.0222,
      "mean_token_accuracy": 0.17854488790035247,
      "num_tokens": 59479566.0,
      "step": 27150
    },
    {
      "entropy": 5.266553640365601,
      "epoch": 2.695017864231838,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004278881580646418,
      "loss": 4.978,
      "mean_token_accuracy": 0.18513259291648865,
      "num_tokens": 59489967.0,
      "step": 27155
    },
    {
      "entropy": 5.349713754653931,
      "epoch": 2.6955140928940056,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004278621610442066,
      "loss": 5.0058,
      "mean_token_accuracy": 0.1784725859761238,
      "num_tokens": 59501640.0,
      "step": 27160
    },
    {
      "entropy": 5.340087032318115,
      "epoch": 2.696010321556173,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004278361602331192,
      "loss": 4.9606,
      "mean_token_accuracy": 0.18998000770807266,
      "num_tokens": 59512767.0,
      "step": 27165
    },
    {
      "entropy": 5.320263433456421,
      "epoch": 2.6965065502183405,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042781015563202434,
      "loss": 4.9255,
      "mean_token_accuracy": 0.18831563591957093,
      "num_tokens": 59522188.0,
      "step": 27170
    },
    {
      "entropy": 5.320263767242432,
      "epoch": 2.697002778880508,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004277841472415669,
      "loss": 4.9692,
      "mean_token_accuracy": 0.18499271124601363,
      "num_tokens": 59534871.0,
      "step": 27175
    },
    {
      "entropy": 5.417339277267456,
      "epoch": 2.6974990075426755,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004277581350623918,
      "loss": 5.0945,
      "mean_token_accuracy": 0.16667983531951905,
      "num_tokens": 59546858.0,
      "step": 27180
    },
    {
      "entropy": 5.3005327701568605,
      "epoch": 2.697995236204843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004277321190951441,
      "loss": 4.9327,
      "mean_token_accuracy": 0.17894740104675294,
      "num_tokens": 59557022.0,
      "step": 27185
    },
    {
      "entropy": 5.294617700576782,
      "epoch": 2.698491464867011,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042770609934046904,
      "loss": 4.9875,
      "mean_token_accuracy": 0.18264349102973937,
      "num_tokens": 59567564.0,
      "step": 27190
    },
    {
      "entropy": 5.436119794845581,
      "epoch": 2.6989876935291783,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042768007579901176,
      "loss": 5.1102,
      "mean_token_accuracy": 0.16319828927516938,
      "num_tokens": 59579122.0,
      "step": 27195
    },
    {
      "entropy": 5.421905279159546,
      "epoch": 2.699483922191346,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042765404847141755,
      "loss": 5.0678,
      "mean_token_accuracy": 0.18066777884960175,
      "num_tokens": 59590477.0,
      "step": 27200
    },
    {
      "entropy": 5.327125597000122,
      "epoch": 2.6999801508535133,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004276280173583319,
      "loss": 4.9752,
      "mean_token_accuracy": 0.178664730489254,
      "num_tokens": 59601217.0,
      "step": 27205
    },
    {
      "entropy": 5.293893146514892,
      "epoch": 2.7004763795156808,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042760198246040025,
      "loss": 4.9834,
      "mean_token_accuracy": 0.18428637087345123,
      "num_tokens": 59612271.0,
      "step": 27210
    },
    {
      "entropy": 5.375889015197754,
      "epoch": 2.7009726081778482,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004275759437782683,
      "loss": 5.0803,
      "mean_token_accuracy": 0.16979404240846635,
      "num_tokens": 59623948.0,
      "step": 27215
    },
    {
      "entropy": 5.284844875335693,
      "epoch": 2.701468836840016,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042754990131258175,
      "loss": 4.9308,
      "mean_token_accuracy": 0.18534862101078034,
      "num_tokens": 59635720.0,
      "step": 27220
    },
    {
      "entropy": 5.329910564422607,
      "epoch": 2.7019650655021836,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004275238550639863,
      "loss": 5.0218,
      "mean_token_accuracy": 0.1783503234386444,
      "num_tokens": 59645171.0,
      "step": 27225
    },
    {
      "entropy": 5.398351716995239,
      "epoch": 2.702461294164351,
      "grad_norm": 1.046875,
      "learning_rate": 0.000427497805033128,
      "loss": 5.1053,
      "mean_token_accuracy": 0.17011975347995759,
      "num_tokens": 59655266.0,
      "step": 27230
    },
    {
      "entropy": 5.305723714828491,
      "epoch": 2.7029575228265186,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004274717512206527,
      "loss": 4.9941,
      "mean_token_accuracy": 0.180736443400383,
      "num_tokens": 59667055.0,
      "step": 27235
    },
    {
      "entropy": 5.307472991943359,
      "epoch": 2.703453751488686,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004274456936272065,
      "loss": 4.9664,
      "mean_token_accuracy": 0.1813073441386223,
      "num_tokens": 59678621.0,
      "step": 27240
    },
    {
      "entropy": 5.283454895019531,
      "epoch": 2.7039499801508535,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042741963225343554,
      "loss": 4.9527,
      "mean_token_accuracy": 0.18495094627141953,
      "num_tokens": 59690514.0,
      "step": 27245
    },
    {
      "entropy": 5.3977439403533936,
      "epoch": 2.704446208813021,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004273935670999861,
      "loss": 5.031,
      "mean_token_accuracy": 0.1746712416410446,
      "num_tokens": 59701271.0,
      "step": 27250
    },
    {
      "entropy": 5.333916759490966,
      "epoch": 2.7049424374751885,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042736749816750476,
      "loss": 5.0802,
      "mean_token_accuracy": 0.1706086963415146,
      "num_tokens": 59712787.0,
      "step": 27255
    },
    {
      "entropy": 5.268269872665405,
      "epoch": 2.705438666137356,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004273414254566376,
      "loss": 4.9284,
      "mean_token_accuracy": 0.18609073758125305,
      "num_tokens": 59722787.0,
      "step": 27260
    },
    {
      "entropy": 5.354815912246704,
      "epoch": 2.7059348947995234,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004273153489680315,
      "loss": 5.0073,
      "mean_token_accuracy": 0.18467934280633927,
      "num_tokens": 59733302.0,
      "step": 27265
    },
    {
      "entropy": 5.297034072875976,
      "epoch": 2.706431123461691,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042728926870233287,
      "loss": 5.0039,
      "mean_token_accuracy": 0.1746957615017891,
      "num_tokens": 59743290.0,
      "step": 27270
    },
    {
      "entropy": 5.392978811264038,
      "epoch": 2.706927352123859,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042726318466018855,
      "loss": 5.0682,
      "mean_token_accuracy": 0.17573204636573792,
      "num_tokens": 59754145.0,
      "step": 27275
    },
    {
      "entropy": 5.362620544433594,
      "epoch": 2.7074235807860263,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004272370968422454,
      "loss": 5.0242,
      "mean_token_accuracy": 0.18052285611629487,
      "num_tokens": 59764075.0,
      "step": 27280
    },
    {
      "entropy": 5.359824562072754,
      "epoch": 2.7079198094481938,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042721100524915017,
      "loss": 5.0464,
      "mean_token_accuracy": 0.17544921785593032,
      "num_tokens": 59775495.0,
      "step": 27285
    },
    {
      "entropy": 5.277297592163086,
      "epoch": 2.7084160381103612,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042718490988155014,
      "loss": 4.8924,
      "mean_token_accuracy": 0.19995022118091582,
      "num_tokens": 59787006.0,
      "step": 27290
    },
    {
      "entropy": 5.376462030410766,
      "epoch": 2.7089122667725287,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042715881074009216,
      "loss": 5.0326,
      "mean_token_accuracy": 0.17698290646076204,
      "num_tokens": 59797680.0,
      "step": 27295
    },
    {
      "entropy": 5.322022247314453,
      "epoch": 2.709408495434696,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042713270782542364,
      "loss": 4.9992,
      "mean_token_accuracy": 0.17850448936223984,
      "num_tokens": 59809666.0,
      "step": 27300
    },
    {
      "entropy": 5.283144855499268,
      "epoch": 2.7099047240968637,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004271066011381917,
      "loss": 4.9514,
      "mean_token_accuracy": 0.1826021194458008,
      "num_tokens": 59819889.0,
      "step": 27305
    },
    {
      "entropy": 5.2844658374786375,
      "epoch": 2.7104009527590316,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004270804906790438,
      "loss": 4.9089,
      "mean_token_accuracy": 0.18591534346342087,
      "num_tokens": 59830344.0,
      "step": 27310
    },
    {
      "entropy": 5.328588771820068,
      "epoch": 2.710897181421199,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004270543764486275,
      "loss": 4.9708,
      "mean_token_accuracy": 0.18362710922956466,
      "num_tokens": 59843498.0,
      "step": 27315
    },
    {
      "entropy": 5.3780770778656,
      "epoch": 2.7113934100833665,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004270282584475903,
      "loss": 5.0296,
      "mean_token_accuracy": 0.1783589854836464,
      "num_tokens": 59854633.0,
      "step": 27320
    },
    {
      "entropy": 5.345932245254517,
      "epoch": 2.711889638745534,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042700213667657983,
      "loss": 5.103,
      "mean_token_accuracy": 0.17464209347963333,
      "num_tokens": 59865925.0,
      "step": 27325
    },
    {
      "entropy": 5.329665088653565,
      "epoch": 2.7123858674077015,
      "grad_norm": 1.109375,
      "learning_rate": 0.000426976011136244,
      "loss": 4.9733,
      "mean_token_accuracy": 0.18086551278829574,
      "num_tokens": 59877196.0,
      "step": 27330
    },
    {
      "entropy": 5.38898549079895,
      "epoch": 2.712882096069869,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042694988182723047,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17274332344532012,
      "num_tokens": 59887411.0,
      "step": 27335
    },
    {
      "entropy": 5.326571273803711,
      "epoch": 2.7133783247320364,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042692374875018735,
      "loss": 5.0167,
      "mean_token_accuracy": 0.18638352751731874,
      "num_tokens": 59899123.0,
      "step": 27340
    },
    {
      "entropy": 5.278571891784668,
      "epoch": 2.713874553394204,
      "grad_norm": 1.0,
      "learning_rate": 0.00042689761190576263,
      "loss": 4.9274,
      "mean_token_accuracy": 0.18905171304941176,
      "num_tokens": 59910326.0,
      "step": 27345
    },
    {
      "entropy": 5.353512144088745,
      "epoch": 2.7143707820563714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042687147129460447,
      "loss": 5.0545,
      "mean_token_accuracy": 0.17368259131908417,
      "num_tokens": 59920722.0,
      "step": 27350
    },
    {
      "entropy": 5.359633493423462,
      "epoch": 2.714867010718539,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042684532691736104,
      "loss": 5.0233,
      "mean_token_accuracy": 0.1760413810610771,
      "num_tokens": 59932568.0,
      "step": 27355
    },
    {
      "entropy": 5.3590597152709964,
      "epoch": 2.715363239380707,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042681917877468076,
      "loss": 5.0773,
      "mean_token_accuracy": 0.17003913521766661,
      "num_tokens": 59943805.0,
      "step": 27360
    },
    {
      "entropy": 5.358906602859497,
      "epoch": 2.7158594680428743,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042679302686721187,
      "loss": 4.9555,
      "mean_token_accuracy": 0.18631673604249954,
      "num_tokens": 59954455.0,
      "step": 27365
    },
    {
      "entropy": 5.338893985748291,
      "epoch": 2.7163556967050417,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004267668711956031,
      "loss": 4.9574,
      "mean_token_accuracy": 0.18204647898674012,
      "num_tokens": 59966722.0,
      "step": 27370
    },
    {
      "entropy": 5.311344289779663,
      "epoch": 2.716851925367209,
      "grad_norm": 1.0,
      "learning_rate": 0.00042674071176050296,
      "loss": 4.9688,
      "mean_token_accuracy": 0.17849921584129333,
      "num_tokens": 59976702.0,
      "step": 27375
    },
    {
      "entropy": 5.307222652435303,
      "epoch": 2.7173481540293767,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042671454856256,
      "loss": 4.9592,
      "mean_token_accuracy": 0.182059383392334,
      "num_tokens": 59987038.0,
      "step": 27380
    },
    {
      "entropy": 5.411672306060791,
      "epoch": 2.717844382691544,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004266883816024233,
      "loss": 5.0141,
      "mean_token_accuracy": 0.1788013055920601,
      "num_tokens": 59996755.0,
      "step": 27385
    },
    {
      "entropy": 5.340665340423584,
      "epoch": 2.7183406113537116,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004266622108807415,
      "loss": 4.9598,
      "mean_token_accuracy": 0.18180500864982604,
      "num_tokens": 60006419.0,
      "step": 27390
    },
    {
      "entropy": 5.3446649551391605,
      "epoch": 2.7188368400158796,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004266360363981637,
      "loss": 5.0559,
      "mean_token_accuracy": 0.17826861888170242,
      "num_tokens": 60017793.0,
      "step": 27395
    },
    {
      "entropy": 5.291457557678223,
      "epoch": 2.719333068678047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042660985815533905,
      "loss": 4.946,
      "mean_token_accuracy": 0.17991572767496108,
      "num_tokens": 60029392.0,
      "step": 27400
    },
    {
      "entropy": 5.344898414611817,
      "epoch": 2.7198292973402145,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042658367615291644,
      "loss": 5.0701,
      "mean_token_accuracy": 0.18217568844556808,
      "num_tokens": 60040467.0,
      "step": 27405
    },
    {
      "entropy": 5.349403667449951,
      "epoch": 2.720325526002382,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004265574903915454,
      "loss": 5.0043,
      "mean_token_accuracy": 0.1755836382508278,
      "num_tokens": 60051583.0,
      "step": 27410
    },
    {
      "entropy": 5.339745378494262,
      "epoch": 2.7208217546645495,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004265313008718751,
      "loss": 4.9961,
      "mean_token_accuracy": 0.1775031343102455,
      "num_tokens": 60062218.0,
      "step": 27415
    },
    {
      "entropy": 5.344811296463012,
      "epoch": 2.721317983326717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042650510759455516,
      "loss": 4.9992,
      "mean_token_accuracy": 0.18334992378950118,
      "num_tokens": 60074455.0,
      "step": 27420
    },
    {
      "entropy": 5.362440299987793,
      "epoch": 2.7218142119888844,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042647891056023495,
      "loss": 4.9528,
      "mean_token_accuracy": 0.17783569544553757,
      "num_tokens": 60085442.0,
      "step": 27425
    },
    {
      "entropy": 5.358032321929931,
      "epoch": 2.722310440651052,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042645270976956426,
      "loss": 5.0485,
      "mean_token_accuracy": 0.1778809294104576,
      "num_tokens": 60096441.0,
      "step": 27430
    },
    {
      "entropy": 5.413398170471192,
      "epoch": 2.7228066693132194,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042642650522319253,
      "loss": 5.0036,
      "mean_token_accuracy": 0.18267711997032166,
      "num_tokens": 60107399.0,
      "step": 27435
    },
    {
      "entropy": 5.35535888671875,
      "epoch": 2.723302897975387,
      "grad_norm": 1.125,
      "learning_rate": 0.0004264002969217699,
      "loss": 4.9753,
      "mean_token_accuracy": 0.18476110845804214,
      "num_tokens": 60117568.0,
      "step": 27440
    },
    {
      "entropy": 5.33410530090332,
      "epoch": 2.7237991266375547,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004263740848659461,
      "loss": 5.0377,
      "mean_token_accuracy": 0.17395889461040498,
      "num_tokens": 60127214.0,
      "step": 27445
    },
    {
      "entropy": 5.2852858543396,
      "epoch": 2.724295355299722,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042634786905637123,
      "loss": 4.9835,
      "mean_token_accuracy": 0.18120017200708388,
      "num_tokens": 60138167.0,
      "step": 27450
    },
    {
      "entropy": 5.307426595687867,
      "epoch": 2.7247915839618897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004263216494936953,
      "loss": 4.9527,
      "mean_token_accuracy": 0.18027831614017487,
      "num_tokens": 60148841.0,
      "step": 27455
    },
    {
      "entropy": 5.342232513427734,
      "epoch": 2.725287812624057,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042629542617856855,
      "loss": 5.0417,
      "mean_token_accuracy": 0.1803953915834427,
      "num_tokens": 60159512.0,
      "step": 27460
    },
    {
      "entropy": 5.318543386459351,
      "epoch": 2.7257840412862246,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004262691991116412,
      "loss": 4.9914,
      "mean_token_accuracy": 0.17348575890064238,
      "num_tokens": 60170463.0,
      "step": 27465
    },
    {
      "entropy": 5.275872039794922,
      "epoch": 2.726280269948392,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004262429682935637,
      "loss": 4.8529,
      "mean_token_accuracy": 0.1883625417947769,
      "num_tokens": 60180431.0,
      "step": 27470
    },
    {
      "entropy": 5.33149733543396,
      "epoch": 2.7267764986105596,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004262167337249865,
      "loss": 5.002,
      "mean_token_accuracy": 0.1780652418732643,
      "num_tokens": 60190410.0,
      "step": 27475
    },
    {
      "entropy": 5.302311706542969,
      "epoch": 2.7272727272727275,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004261904954065601,
      "loss": 4.9409,
      "mean_token_accuracy": 0.17905097603797912,
      "num_tokens": 60201909.0,
      "step": 27480
    },
    {
      "entropy": 5.363467121124268,
      "epoch": 2.727768955934895,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004261642533389352,
      "loss": 4.9447,
      "mean_token_accuracy": 0.1810883477330208,
      "num_tokens": 60212577.0,
      "step": 27485
    },
    {
      "entropy": 5.326886463165283,
      "epoch": 2.7282651845970625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042613800752276264,
      "loss": 5.0569,
      "mean_token_accuracy": 0.1744513139128685,
      "num_tokens": 60222615.0,
      "step": 27490
    },
    {
      "entropy": 5.308686399459839,
      "epoch": 2.72876141325923,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042611175795869306,
      "loss": 5.0708,
      "mean_token_accuracy": 0.17694258838891982,
      "num_tokens": 60234338.0,
      "step": 27495
    },
    {
      "entropy": 5.379582309722901,
      "epoch": 2.7292576419213974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004260855046473775,
      "loss": 4.965,
      "mean_token_accuracy": 0.18285751938819886,
      "num_tokens": 60244781.0,
      "step": 27500
    },
    {
      "entropy": 5.364536094665527,
      "epoch": 2.729753870583565,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042605924758946706,
      "loss": 5.033,
      "mean_token_accuracy": 0.17440435737371446,
      "num_tokens": 60255150.0,
      "step": 27505
    },
    {
      "entropy": 5.305346441268921,
      "epoch": 2.7302500992457324,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042603298678561277,
      "loss": 4.9843,
      "mean_token_accuracy": 0.17913282066583633,
      "num_tokens": 60266073.0,
      "step": 27510
    },
    {
      "entropy": 5.3612313747406,
      "epoch": 2.7307463279079,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004260067222364659,
      "loss": 5.0205,
      "mean_token_accuracy": 0.1731000065803528,
      "num_tokens": 60277520.0,
      "step": 27515
    },
    {
      "entropy": 5.363452529907226,
      "epoch": 2.7312425565700673,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042598045394267757,
      "loss": 4.9946,
      "mean_token_accuracy": 0.18091044425964356,
      "num_tokens": 60288530.0,
      "step": 27520
    },
    {
      "entropy": 5.329304218292236,
      "epoch": 2.731738785232235,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042595418190489944,
      "loss": 4.9372,
      "mean_token_accuracy": 0.1828804299235344,
      "num_tokens": 60299926.0,
      "step": 27525
    },
    {
      "entropy": 5.39128360748291,
      "epoch": 2.7322350138944023,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00042592790612378277,
      "loss": 5.1175,
      "mean_token_accuracy": 0.17327685058116912,
      "num_tokens": 60312129.0,
      "step": 27530
    },
    {
      "entropy": 5.368721914291382,
      "epoch": 2.73273124255657,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042590162659997933,
      "loss": 5.0192,
      "mean_token_accuracy": 0.1777412325143814,
      "num_tokens": 60322405.0,
      "step": 27535
    },
    {
      "entropy": 5.380045986175537,
      "epoch": 2.7332274712187377,
      "grad_norm": 1.0,
      "learning_rate": 0.0004258753433341407,
      "loss": 5.0062,
      "mean_token_accuracy": 0.17702761441469192,
      "num_tokens": 60333112.0,
      "step": 27540
    },
    {
      "entropy": 5.359342956542969,
      "epoch": 2.733723699880905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042584905632691867,
      "loss": 5.0057,
      "mean_token_accuracy": 0.17801157683134078,
      "num_tokens": 60343500.0,
      "step": 27545
    },
    {
      "entropy": 5.321665239334107,
      "epoch": 2.7342199285430726,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004258227655789651,
      "loss": 4.9847,
      "mean_token_accuracy": 0.17799243330955505,
      "num_tokens": 60354711.0,
      "step": 27550
    },
    {
      "entropy": 5.350671100616455,
      "epoch": 2.73471615720524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004257964710909318,
      "loss": 4.9871,
      "mean_token_accuracy": 0.17643335610628127,
      "num_tokens": 60365064.0,
      "step": 27555
    },
    {
      "entropy": 5.342635297775269,
      "epoch": 2.7352123858674076,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042577017286347116,
      "loss": 5.0353,
      "mean_token_accuracy": 0.17585411965847014,
      "num_tokens": 60376425.0,
      "step": 27560
    },
    {
      "entropy": 5.348534107208252,
      "epoch": 2.7357086145295755,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042574387089723504,
      "loss": 4.9105,
      "mean_token_accuracy": 0.1843389794230461,
      "num_tokens": 60386745.0,
      "step": 27565
    },
    {
      "entropy": 5.336801242828369,
      "epoch": 2.736204843191743,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042571756519287567,
      "loss": 4.9667,
      "mean_token_accuracy": 0.18202827423810958,
      "num_tokens": 60397038.0,
      "step": 27570
    },
    {
      "entropy": 5.287863874435425,
      "epoch": 2.7367010718539104,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042569125575104557,
      "loss": 4.9757,
      "mean_token_accuracy": 0.18170148879289627,
      "num_tokens": 60406955.0,
      "step": 27575
    },
    {
      "entropy": 5.337188625335694,
      "epoch": 2.737197300516078,
      "grad_norm": 1.0,
      "learning_rate": 0.00042566494257239695,
      "loss": 4.991,
      "mean_token_accuracy": 0.18001007288694382,
      "num_tokens": 60418245.0,
      "step": 27580
    },
    {
      "entropy": 5.3631024837493895,
      "epoch": 2.7376935291782454,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004256386256575824,
      "loss": 5.0075,
      "mean_token_accuracy": 0.17218142747879028,
      "num_tokens": 60429310.0,
      "step": 27585
    },
    {
      "entropy": 5.379766273498535,
      "epoch": 2.738189757840413,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004256123050072546,
      "loss": 5.0431,
      "mean_token_accuracy": 0.1805417999625206,
      "num_tokens": 60441355.0,
      "step": 27590
    },
    {
      "entropy": 5.46537070274353,
      "epoch": 2.7386859865025803,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00042558598062206616,
      "loss": 5.1828,
      "mean_token_accuracy": 0.17151687741279603,
      "num_tokens": 60452853.0,
      "step": 27595
    },
    {
      "entropy": 5.359786891937256,
      "epoch": 2.739182215164748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042555965250266984,
      "loss": 4.9898,
      "mean_token_accuracy": 0.1755588635802269,
      "num_tokens": 60462934.0,
      "step": 27600
    },
    {
      "entropy": 5.350796937942505,
      "epoch": 2.7396784438269153,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042553332064971866,
      "loss": 5.0681,
      "mean_token_accuracy": 0.17223575264215468,
      "num_tokens": 60475575.0,
      "step": 27605
    },
    {
      "entropy": 5.248402690887451,
      "epoch": 2.7401746724890828,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042550698506386536,
      "loss": 4.8604,
      "mean_token_accuracy": 0.1909021556377411,
      "num_tokens": 60486495.0,
      "step": 27610
    },
    {
      "entropy": 5.300896263122558,
      "epoch": 2.7406709011512502,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042548064574576323,
      "loss": 4.9775,
      "mean_token_accuracy": 0.1890899956226349,
      "num_tokens": 60496806.0,
      "step": 27615
    },
    {
      "entropy": 5.318427896499633,
      "epoch": 2.741167129813418,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004254543026960654,
      "loss": 4.9839,
      "mean_token_accuracy": 0.1817610129714012,
      "num_tokens": 60508905.0,
      "step": 27620
    },
    {
      "entropy": 5.263612699508667,
      "epoch": 2.7416633584755856,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004254279559154249,
      "loss": 4.94,
      "mean_token_accuracy": 0.17960686534643172,
      "num_tokens": 60519603.0,
      "step": 27625
    },
    {
      "entropy": 5.351475524902344,
      "epoch": 2.742159587137753,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042540160540449536,
      "loss": 5.0122,
      "mean_token_accuracy": 0.17944026738405228,
      "num_tokens": 60530977.0,
      "step": 27630
    },
    {
      "entropy": 5.361991691589355,
      "epoch": 2.7426558157999206,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004253752511639301,
      "loss": 5.028,
      "mean_token_accuracy": 0.17199240177869796,
      "num_tokens": 60541354.0,
      "step": 27635
    },
    {
      "entropy": 5.300813627243042,
      "epoch": 2.743152044462088,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004253488931943826,
      "loss": 4.9943,
      "mean_token_accuracy": 0.1843712568283081,
      "num_tokens": 60552531.0,
      "step": 27640
    },
    {
      "entropy": 5.283671474456787,
      "epoch": 2.7436482731242555,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042532253149650653,
      "loss": 4.9655,
      "mean_token_accuracy": 0.18699568659067153,
      "num_tokens": 60562552.0,
      "step": 27645
    },
    {
      "entropy": 5.354524850845337,
      "epoch": 2.7441445017864234,
      "grad_norm": 1.0,
      "learning_rate": 0.0004252961660709556,
      "loss": 5.0078,
      "mean_token_accuracy": 0.17231801748275757,
      "num_tokens": 60572849.0,
      "step": 27650
    },
    {
      "entropy": 5.328178358078003,
      "epoch": 2.744640730448591,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042526979691838353,
      "loss": 4.9549,
      "mean_token_accuracy": 0.18682256788015367,
      "num_tokens": 60584095.0,
      "step": 27655
    },
    {
      "entropy": 5.374096250534057,
      "epoch": 2.7451369591107584,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042524342403944446,
      "loss": 5.0308,
      "mean_token_accuracy": 0.17703590542078018,
      "num_tokens": 60594353.0,
      "step": 27660
    },
    {
      "entropy": 5.391670989990234,
      "epoch": 2.745633187772926,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042521704743479213,
      "loss": 5.0285,
      "mean_token_accuracy": 0.1754138022661209,
      "num_tokens": 60604558.0,
      "step": 27665
    },
    {
      "entropy": 5.284481620788574,
      "epoch": 2.7461294164350933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042519066710508064,
      "loss": 4.9691,
      "mean_token_accuracy": 0.17866220474243164,
      "num_tokens": 60615670.0,
      "step": 27670
    },
    {
      "entropy": 5.26500129699707,
      "epoch": 2.746625645097261,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042516428305096435,
      "loss": 4.9051,
      "mean_token_accuracy": 0.18857925534248351,
      "num_tokens": 60626452.0,
      "step": 27675
    },
    {
      "entropy": 5.326062774658203,
      "epoch": 2.7471218737594283,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004251378952730973,
      "loss": 4.9987,
      "mean_token_accuracy": 0.17969779372215272,
      "num_tokens": 60637122.0,
      "step": 27680
    },
    {
      "entropy": 5.346644592285156,
      "epoch": 2.7476181024215958,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004251115037721341,
      "loss": 4.9789,
      "mean_token_accuracy": 0.18890277296304703,
      "num_tokens": 60648835.0,
      "step": 27685
    },
    {
      "entropy": 5.3319464206695555,
      "epoch": 2.7481143310837632,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042508510854872905,
      "loss": 5.0345,
      "mean_token_accuracy": 0.17477318793535232,
      "num_tokens": 60658537.0,
      "step": 27690
    },
    {
      "entropy": 5.40693507194519,
      "epoch": 2.7486105597459307,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042505870960353665,
      "loss": 5.0397,
      "mean_token_accuracy": 0.17926022112369538,
      "num_tokens": 60670450.0,
      "step": 27695
    },
    {
      "entropy": 5.430488586425781,
      "epoch": 2.749106788408098,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004250323069372116,
      "loss": 5.0072,
      "mean_token_accuracy": 0.18400170058012008,
      "num_tokens": 60680539.0,
      "step": 27700
    },
    {
      "entropy": 5.355594253540039,
      "epoch": 2.749603017070266,
      "grad_norm": 1.125,
      "learning_rate": 0.00042500590055040866,
      "loss": 5.0943,
      "mean_token_accuracy": 0.17577212154865265,
      "num_tokens": 60693106.0,
      "step": 27705
    },
    {
      "entropy": 5.44789342880249,
      "epoch": 2.7500992457324336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042497949044378267,
      "loss": 5.0618,
      "mean_token_accuracy": 0.17521715462207793,
      "num_tokens": 60703509.0,
      "step": 27710
    },
    {
      "entropy": 5.359287261962891,
      "epoch": 2.750595474394601,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042495307661798837,
      "loss": 5.0289,
      "mean_token_accuracy": 0.17814191579818725,
      "num_tokens": 60714113.0,
      "step": 27715
    },
    {
      "entropy": 5.408002662658691,
      "epoch": 2.7510917030567685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042492665907368094,
      "loss": 5.0902,
      "mean_token_accuracy": 0.16891328543424605,
      "num_tokens": 60726826.0,
      "step": 27720
    },
    {
      "entropy": 5.343953084945679,
      "epoch": 2.751587931718936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042490023781151546,
      "loss": 4.9959,
      "mean_token_accuracy": 0.17881002575159072,
      "num_tokens": 60738443.0,
      "step": 27725
    },
    {
      "entropy": 5.317821979522705,
      "epoch": 2.7520841603811035,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042487381283214706,
      "loss": 4.9598,
      "mean_token_accuracy": 0.18111031353473664,
      "num_tokens": 60748865.0,
      "step": 27730
    },
    {
      "entropy": 5.346115684509277,
      "epoch": 2.7525803890432714,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004248473841362311,
      "loss": 5.0066,
      "mean_token_accuracy": 0.18398686349391938,
      "num_tokens": 60760001.0,
      "step": 27735
    },
    {
      "entropy": 5.402475261688233,
      "epoch": 2.753076617705439,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042482095172442286,
      "loss": 5.0457,
      "mean_token_accuracy": 0.18001634180545806,
      "num_tokens": 60772003.0,
      "step": 27740
    },
    {
      "entropy": 5.310573387145996,
      "epoch": 2.7535728463676064,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004247945155973778,
      "loss": 4.9292,
      "mean_token_accuracy": 0.18787006735801698,
      "num_tokens": 60783167.0,
      "step": 27745
    },
    {
      "entropy": 5.36251540184021,
      "epoch": 2.754069075029774,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004247680757557516,
      "loss": 5.0338,
      "mean_token_accuracy": 0.17869070321321487,
      "num_tokens": 60793917.0,
      "step": 27750
    },
    {
      "entropy": 5.429286670684815,
      "epoch": 2.7545653036919413,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042474163220019987,
      "loss": 5.1183,
      "mean_token_accuracy": 0.17107173651456833,
      "num_tokens": 60805166.0,
      "step": 27755
    },
    {
      "entropy": 5.4346918106079105,
      "epoch": 2.7550615323541088,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042471518493137836,
      "loss": 5.0561,
      "mean_token_accuracy": 0.1774685874581337,
      "num_tokens": 60816984.0,
      "step": 27760
    },
    {
      "entropy": 5.309739351272583,
      "epoch": 2.7555577610162763,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004246887339499427,
      "loss": 4.9552,
      "mean_token_accuracy": 0.18301518112421036,
      "num_tokens": 60829481.0,
      "step": 27765
    },
    {
      "entropy": 5.337366151809692,
      "epoch": 2.7560539896784437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042466227925654917,
      "loss": 4.9522,
      "mean_token_accuracy": 0.18468558937311172,
      "num_tokens": 60839376.0,
      "step": 27770
    },
    {
      "entropy": 5.307725048065185,
      "epoch": 2.756550218340611,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004246358208518535,
      "loss": 4.9961,
      "mean_token_accuracy": 0.1784382775425911,
      "num_tokens": 60850383.0,
      "step": 27775
    },
    {
      "entropy": 5.309764099121094,
      "epoch": 2.7570464470027787,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042460935873651196,
      "loss": 4.9561,
      "mean_token_accuracy": 0.1796943798661232,
      "num_tokens": 60861249.0,
      "step": 27780
    },
    {
      "entropy": 5.3388160228729244,
      "epoch": 2.757542675664946,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042458289291118066,
      "loss": 5.0085,
      "mean_token_accuracy": 0.1781307652592659,
      "num_tokens": 60872220.0,
      "step": 27785
    },
    {
      "entropy": 5.3364886283874515,
      "epoch": 2.758038904327114,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042455642337651595,
      "loss": 4.9866,
      "mean_token_accuracy": 0.17707824110984802,
      "num_tokens": 60882006.0,
      "step": 27790
    },
    {
      "entropy": 5.359373664855957,
      "epoch": 2.7585351329892815,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004245299501331743,
      "loss": 4.9193,
      "mean_token_accuracy": 0.1893056869506836,
      "num_tokens": 60891732.0,
      "step": 27795
    },
    {
      "entropy": 5.340056848526001,
      "epoch": 2.759031361651449,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042450347318181194,
      "loss": 5.0391,
      "mean_token_accuracy": 0.17900743037462236,
      "num_tokens": 60902990.0,
      "step": 27800
    },
    {
      "entropy": 5.3340143203735355,
      "epoch": 2.7595275903136165,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042447699252308566,
      "loss": 4.979,
      "mean_token_accuracy": 0.1796618655323982,
      "num_tokens": 60914673.0,
      "step": 27805
    },
    {
      "entropy": 5.320575714111328,
      "epoch": 2.760023818975784,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004244505081576521,
      "loss": 4.9573,
      "mean_token_accuracy": 0.1843356266617775,
      "num_tokens": 60925872.0,
      "step": 27810
    },
    {
      "entropy": 5.357184934616089,
      "epoch": 2.7605200476379514,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042442402008616793,
      "loss": 5.0471,
      "mean_token_accuracy": 0.17973460704088212,
      "num_tokens": 60936589.0,
      "step": 27815
    },
    {
      "entropy": 5.356748008728028,
      "epoch": 2.7610162763001194,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042439752830929,
      "loss": 4.9879,
      "mean_token_accuracy": 0.17817127853631973,
      "num_tokens": 60947213.0,
      "step": 27820
    },
    {
      "entropy": 5.407035207748413,
      "epoch": 2.761512504962287,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004243710328276754,
      "loss": 5.0138,
      "mean_token_accuracy": 0.17339904308319093,
      "num_tokens": 60958615.0,
      "step": 27825
    },
    {
      "entropy": 5.311867666244507,
      "epoch": 2.7620087336244543,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000424344533641981,
      "loss": 4.939,
      "mean_token_accuracy": 0.18388899564743041,
      "num_tokens": 60968830.0,
      "step": 27830
    },
    {
      "entropy": 5.294661808013916,
      "epoch": 2.762504962286622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000424318030752864,
      "loss": 4.9129,
      "mean_token_accuracy": 0.1835325539112091,
      "num_tokens": 60979744.0,
      "step": 27835
    },
    {
      "entropy": 5.338097810745239,
      "epoch": 2.7630011909487893,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004242915241609815,
      "loss": 5.0049,
      "mean_token_accuracy": 0.17534390687942505,
      "num_tokens": 60990291.0,
      "step": 27840
    },
    {
      "entropy": 5.33810658454895,
      "epoch": 2.7634974196109567,
      "grad_norm": 1.0,
      "learning_rate": 0.000424265013866991,
      "loss": 4.9791,
      "mean_token_accuracy": 0.18277374804019927,
      "num_tokens": 61002170.0,
      "step": 27845
    },
    {
      "entropy": 5.347066974639892,
      "epoch": 2.763993648273124,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004242384998715497,
      "loss": 5.0343,
      "mean_token_accuracy": 0.17830309718847276,
      "num_tokens": 61012639.0,
      "step": 27850
    },
    {
      "entropy": 5.395547199249267,
      "epoch": 2.7644898769352917,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004242119821753153,
      "loss": 4.9857,
      "mean_token_accuracy": 0.17935121506452562,
      "num_tokens": 61022657.0,
      "step": 27855
    },
    {
      "entropy": 5.308938217163086,
      "epoch": 2.764986105597459,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004241854607789452,
      "loss": 4.9404,
      "mean_token_accuracy": 0.18261522948741912,
      "num_tokens": 61033108.0,
      "step": 27860
    },
    {
      "entropy": 5.274685192108154,
      "epoch": 2.7654823342596266,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042415893568309714,
      "loss": 4.9744,
      "mean_token_accuracy": 0.17960318624973298,
      "num_tokens": 61043651.0,
      "step": 27865
    },
    {
      "entropy": 5.284752988815308,
      "epoch": 2.765978562921794,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042413240688842884,
      "loss": 4.9212,
      "mean_token_accuracy": 0.18622350841760635,
      "num_tokens": 61053234.0,
      "step": 27870
    },
    {
      "entropy": 5.294169282913208,
      "epoch": 2.766474791583962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004241058743955982,
      "loss": 4.9657,
      "mean_token_accuracy": 0.17877672165632247,
      "num_tokens": 61065178.0,
      "step": 27875
    },
    {
      "entropy": 5.328287935256958,
      "epoch": 2.7669710202461295,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004240793382052633,
      "loss": 4.9755,
      "mean_token_accuracy": 0.18450323194265367,
      "num_tokens": 61075709.0,
      "step": 27880
    },
    {
      "entropy": 5.359931182861328,
      "epoch": 2.767467248908297,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000424052798318082,
      "loss": 5.0354,
      "mean_token_accuracy": 0.17342312186956405,
      "num_tokens": 61086880.0,
      "step": 27885
    },
    {
      "entropy": 5.300610494613648,
      "epoch": 2.7679634775704645,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004240262547347124,
      "loss": 4.966,
      "mean_token_accuracy": 0.19046102911233903,
      "num_tokens": 61097008.0,
      "step": 27890
    },
    {
      "entropy": 5.307938146591186,
      "epoch": 2.768459706232632,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042399970745581283,
      "loss": 4.9611,
      "mean_token_accuracy": 0.18145473301410675,
      "num_tokens": 61107844.0,
      "step": 27895
    },
    {
      "entropy": 5.251866006851197,
      "epoch": 2.7689559348947994,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042397315648204164,
      "loss": 4.9264,
      "mean_token_accuracy": 0.19362008571624756,
      "num_tokens": 61118652.0,
      "step": 27900
    },
    {
      "entropy": 5.359127759933472,
      "epoch": 2.7694521635569673,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042394660181405716,
      "loss": 5.0616,
      "mean_token_accuracy": 0.17622987180948257,
      "num_tokens": 61129143.0,
      "step": 27905
    },
    {
      "entropy": 5.354335498809815,
      "epoch": 2.769948392219135,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004239200434525179,
      "loss": 5.0001,
      "mean_token_accuracy": 0.17506590932607652,
      "num_tokens": 61140623.0,
      "step": 27910
    },
    {
      "entropy": 5.29187331199646,
      "epoch": 2.7704446208813023,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004238934813980824,
      "loss": 4.9634,
      "mean_token_accuracy": 0.17673904597759246,
      "num_tokens": 61150998.0,
      "step": 27915
    },
    {
      "entropy": 5.344198751449585,
      "epoch": 2.7709408495434698,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042386691565140947,
      "loss": 5.0038,
      "mean_token_accuracy": 0.17981629371643065,
      "num_tokens": 61162086.0,
      "step": 27920
    },
    {
      "entropy": 5.352705812454223,
      "epoch": 2.7714370782056372,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042384034621315773,
      "loss": 4.9634,
      "mean_token_accuracy": 0.17970454096794128,
      "num_tokens": 61171806.0,
      "step": 27925
    },
    {
      "entropy": 5.28999810218811,
      "epoch": 2.7719333068678047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004238137730839862,
      "loss": 4.9533,
      "mean_token_accuracy": 0.1774725779891014,
      "num_tokens": 61182056.0,
      "step": 27930
    },
    {
      "entropy": 5.312408971786499,
      "epoch": 2.772429535529972,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004237871962645538,
      "loss": 4.9887,
      "mean_token_accuracy": 0.18202342242002487,
      "num_tokens": 61193231.0,
      "step": 27935
    },
    {
      "entropy": 5.296475744247436,
      "epoch": 2.7729257641921397,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042376061575551946,
      "loss": 4.9647,
      "mean_token_accuracy": 0.1804387614130974,
      "num_tokens": 61203785.0,
      "step": 27940
    },
    {
      "entropy": 5.311031341552734,
      "epoch": 2.773421992854307,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004237340315575425,
      "loss": 4.9804,
      "mean_token_accuracy": 0.18496800363063812,
      "num_tokens": 61215244.0,
      "step": 27945
    },
    {
      "entropy": 5.356261491775513,
      "epoch": 2.7739182215164746,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004237074436712819,
      "loss": 4.9511,
      "mean_token_accuracy": 0.17590915858745576,
      "num_tokens": 61227147.0,
      "step": 27950
    },
    {
      "entropy": 5.370459127426147,
      "epoch": 2.774414450178642,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042368085209739715,
      "loss": 5.0545,
      "mean_token_accuracy": 0.17443457543849944,
      "num_tokens": 61237687.0,
      "step": 27955
    },
    {
      "entropy": 5.251413440704345,
      "epoch": 2.77491067884081,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004236542568365478,
      "loss": 4.9247,
      "mean_token_accuracy": 0.18578029572963714,
      "num_tokens": 61247700.0,
      "step": 27960
    },
    {
      "entropy": 5.325287532806397,
      "epoch": 2.7754069075029775,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042362765788939307,
      "loss": 4.9266,
      "mean_token_accuracy": 0.1761413872241974,
      "num_tokens": 61258407.0,
      "step": 27965
    },
    {
      "entropy": 5.386866474151612,
      "epoch": 2.775903136165145,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004236010552565927,
      "loss": 5.0943,
      "mean_token_accuracy": 0.16848712265491486,
      "num_tokens": 61269065.0,
      "step": 27970
    },
    {
      "entropy": 5.4075380802154545,
      "epoch": 2.7763993648273124,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042357444893880645,
      "loss": 4.9933,
      "mean_token_accuracy": 0.1799297198653221,
      "num_tokens": 61280767.0,
      "step": 27975
    },
    {
      "entropy": 5.27960171699524,
      "epoch": 2.77689559348948,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042354783893669385,
      "loss": 4.9321,
      "mean_token_accuracy": 0.17867035269737244,
      "num_tokens": 61291967.0,
      "step": 27980
    },
    {
      "entropy": 5.341236019134522,
      "epoch": 2.7773918221516474,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004235212252509151,
      "loss": 5.0877,
      "mean_token_accuracy": 0.17374467253684997,
      "num_tokens": 61303572.0,
      "step": 27985
    },
    {
      "entropy": 5.37512993812561,
      "epoch": 2.777888050813815,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042349460788212995,
      "loss": 5.0246,
      "mean_token_accuracy": 0.17837724834680557,
      "num_tokens": 61315674.0,
      "step": 27990
    },
    {
      "entropy": 5.366075229644776,
      "epoch": 2.7783842794759828,
      "grad_norm": 1.0,
      "learning_rate": 0.00042346798683099844,
      "loss": 4.9447,
      "mean_token_accuracy": 0.18368050307035447,
      "num_tokens": 61326334.0,
      "step": 27995
    },
    {
      "entropy": 5.337068271636963,
      "epoch": 2.7788805081381502,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004234413620981809,
      "loss": 5.0678,
      "mean_token_accuracy": 0.1787543073296547,
      "num_tokens": 61337505.0,
      "step": 28000
    },
    {
      "entropy": 5.331257581710815,
      "epoch": 2.7793767368003177,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042341473368433735,
      "loss": 5.0429,
      "mean_token_accuracy": 0.17821891605854034,
      "num_tokens": 61348451.0,
      "step": 28005
    },
    {
      "entropy": 5.402016448974609,
      "epoch": 2.779872965462485,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004233881015901282,
      "loss": 5.1665,
      "mean_token_accuracy": 0.16999365091323854,
      "num_tokens": 61358356.0,
      "step": 28010
    },
    {
      "entropy": 5.444169235229492,
      "epoch": 2.7803691941246527,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004233614658162139,
      "loss": 5.0096,
      "mean_token_accuracy": 0.18060506284236907,
      "num_tokens": 61370996.0,
      "step": 28015
    },
    {
      "entropy": 5.413422918319702,
      "epoch": 2.78086542278682,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042333482636325504,
      "loss": 5.0554,
      "mean_token_accuracy": 0.17386234700679778,
      "num_tokens": 61383928.0,
      "step": 28020
    },
    {
      "entropy": 5.331725263595581,
      "epoch": 2.7813616514489876,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000423308183231912,
      "loss": 5.0139,
      "mean_token_accuracy": 0.17836616188287735,
      "num_tokens": 61394277.0,
      "step": 28025
    },
    {
      "entropy": 5.312063837051392,
      "epoch": 2.781857880111155,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042328153642284567,
      "loss": 4.9892,
      "mean_token_accuracy": 0.1831934005022049,
      "num_tokens": 61405928.0,
      "step": 28030
    },
    {
      "entropy": 5.378446865081787,
      "epoch": 2.7823541087733226,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004232548859367167,
      "loss": 5.0536,
      "mean_token_accuracy": 0.1728468731045723,
      "num_tokens": 61417893.0,
      "step": 28035
    },
    {
      "entropy": 5.322105264663696,
      "epoch": 2.78285033743549,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000423228231774186,
      "loss": 5.0042,
      "mean_token_accuracy": 0.18051295578479767,
      "num_tokens": 61428099.0,
      "step": 28040
    },
    {
      "entropy": 5.327148580551148,
      "epoch": 2.783346566097658,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004232015739359146,
      "loss": 4.9293,
      "mean_token_accuracy": 0.18717676699161528,
      "num_tokens": 61437636.0,
      "step": 28045
    },
    {
      "entropy": 5.29964771270752,
      "epoch": 2.7838427947598254,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042317491242256346,
      "loss": 4.9121,
      "mean_token_accuracy": 0.1790805608034134,
      "num_tokens": 61448038.0,
      "step": 28050
    },
    {
      "entropy": 5.312914276123047,
      "epoch": 2.784339023421993,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004231482472347938,
      "loss": 4.9774,
      "mean_token_accuracy": 0.17872326523065568,
      "num_tokens": 61460258.0,
      "step": 28055
    },
    {
      "entropy": 5.335671138763428,
      "epoch": 2.7848352520841604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004231215783732668,
      "loss": 4.9999,
      "mean_token_accuracy": 0.1798801138997078,
      "num_tokens": 61472274.0,
      "step": 28060
    },
    {
      "entropy": 5.32075605392456,
      "epoch": 2.785331480746328,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042309490583864395,
      "loss": 4.9921,
      "mean_token_accuracy": 0.18316084742546082,
      "num_tokens": 61482975.0,
      "step": 28065
    },
    {
      "entropy": 5.348207712173462,
      "epoch": 2.7858277094084953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004230682296315865,
      "loss": 5.0165,
      "mean_token_accuracy": 0.176599682867527,
      "num_tokens": 61494223.0,
      "step": 28070
    },
    {
      "entropy": 5.401851320266724,
      "epoch": 2.786323938070663,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000423041549752756,
      "loss": 5.0547,
      "mean_token_accuracy": 0.17810311615467073,
      "num_tokens": 61505981.0,
      "step": 28075
    },
    {
      "entropy": 5.248749494552612,
      "epoch": 2.7868201667328307,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004230148662028141,
      "loss": 4.9488,
      "mean_token_accuracy": 0.18333244770765306,
      "num_tokens": 61516410.0,
      "step": 28080
    },
    {
      "entropy": 5.30927357673645,
      "epoch": 2.787316395394998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042298817898242237,
      "loss": 4.9507,
      "mean_token_accuracy": 0.17872238606214524,
      "num_tokens": 61526551.0,
      "step": 28085
    },
    {
      "entropy": 5.377071714401245,
      "epoch": 2.7878126240571657,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042296148809224275,
      "loss": 5.0902,
      "mean_token_accuracy": 0.17342455238103865,
      "num_tokens": 61538214.0,
      "step": 28090
    },
    {
      "entropy": 5.351520919799805,
      "epoch": 2.788308852719333,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004229347935329371,
      "loss": 5.0538,
      "mean_token_accuracy": 0.17335594147443772,
      "num_tokens": 61549726.0,
      "step": 28095
    },
    {
      "entropy": 5.425281620025634,
      "epoch": 2.7888050813815006,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042290809530516724,
      "loss": 5.1089,
      "mean_token_accuracy": 0.1688452810049057,
      "num_tokens": 61560070.0,
      "step": 28100
    },
    {
      "entropy": 5.442900371551514,
      "epoch": 2.789301310043668,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004228813934095954,
      "loss": 5.046,
      "mean_token_accuracy": 0.17862680405378342,
      "num_tokens": 61571348.0,
      "step": 28105
    },
    {
      "entropy": 5.320029020309448,
      "epoch": 2.7897975387058356,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042285468784688363,
      "loss": 4.9304,
      "mean_token_accuracy": 0.18339848518371582,
      "num_tokens": 61581981.0,
      "step": 28110
    },
    {
      "entropy": 5.312740802764893,
      "epoch": 2.790293767368003,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042282797861769423,
      "loss": 5.0276,
      "mean_token_accuracy": 0.17877482026815414,
      "num_tokens": 61591248.0,
      "step": 28115
    },
    {
      "entropy": 5.339161396026611,
      "epoch": 2.7907899960301705,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004228012657226895,
      "loss": 5.032,
      "mean_token_accuracy": 0.17686053216457367,
      "num_tokens": 61602432.0,
      "step": 28120
    },
    {
      "entropy": 5.265822029113769,
      "epoch": 2.791286224692338,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042277454916253186,
      "loss": 4.9923,
      "mean_token_accuracy": 0.17896666824817659,
      "num_tokens": 61613418.0,
      "step": 28125
    },
    {
      "entropy": 5.319204998016358,
      "epoch": 2.791782453354506,
      "grad_norm": 1.125,
      "learning_rate": 0.0004227478289378838,
      "loss": 4.9694,
      "mean_token_accuracy": 0.1742994323372841,
      "num_tokens": 61623783.0,
      "step": 28130
    },
    {
      "entropy": 5.386015272140503,
      "epoch": 2.7922786820166734,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042272110504940795,
      "loss": 5.0672,
      "mean_token_accuracy": 0.17399631142616273,
      "num_tokens": 61633838.0,
      "step": 28135
    },
    {
      "entropy": 5.329862356185913,
      "epoch": 2.792774910678841,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042269437749776703,
      "loss": 4.9453,
      "mean_token_accuracy": 0.18401684165000914,
      "num_tokens": 61645982.0,
      "step": 28140
    },
    {
      "entropy": 5.277101612091064,
      "epoch": 2.7932711393410083,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004226676462836238,
      "loss": 4.9113,
      "mean_token_accuracy": 0.1838517367839813,
      "num_tokens": 61656030.0,
      "step": 28145
    },
    {
      "entropy": 5.2639843940734865,
      "epoch": 2.793767368003176,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004226409114076411,
      "loss": 4.9253,
      "mean_token_accuracy": 0.18220307528972626,
      "num_tokens": 61666746.0,
      "step": 28150
    },
    {
      "entropy": 5.3628674983978275,
      "epoch": 2.7942635966653433,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042261417287048196,
      "loss": 5.1052,
      "mean_token_accuracy": 0.17522049099206924,
      "num_tokens": 61677681.0,
      "step": 28155
    },
    {
      "entropy": 5.420082378387451,
      "epoch": 2.7947598253275108,
      "grad_norm": 1.0,
      "learning_rate": 0.0004225874306728094,
      "loss": 5.0691,
      "mean_token_accuracy": 0.17240481078624725,
      "num_tokens": 61689564.0,
      "step": 28160
    },
    {
      "entropy": 5.353487730026245,
      "epoch": 2.7952560539896787,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042256068481528666,
      "loss": 4.9695,
      "mean_token_accuracy": 0.18217552751302718,
      "num_tokens": 61701074.0,
      "step": 28165
    },
    {
      "entropy": 5.286141872406006,
      "epoch": 2.795752282651846,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004225339352985768,
      "loss": 4.9594,
      "mean_token_accuracy": 0.1816261366009712,
      "num_tokens": 61712203.0,
      "step": 28170
    },
    {
      "entropy": 5.293099689483642,
      "epoch": 2.7962485113140136,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042250718212334334,
      "loss": 4.9208,
      "mean_token_accuracy": 0.18002975434064866,
      "num_tokens": 61723362.0,
      "step": 28175
    },
    {
      "entropy": 5.322131872177124,
      "epoch": 2.796744739976181,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042248042529024956,
      "loss": 4.9415,
      "mean_token_accuracy": 0.1840262532234192,
      "num_tokens": 61733454.0,
      "step": 28180
    },
    {
      "entropy": 5.406666803359985,
      "epoch": 2.7972409686383486,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042245366479995894,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17306478470563888,
      "num_tokens": 61745014.0,
      "step": 28185
    },
    {
      "entropy": 5.282044363021851,
      "epoch": 2.797737197300516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042242690065313535,
      "loss": 5.0078,
      "mean_token_accuracy": 0.1779974579811096,
      "num_tokens": 61755334.0,
      "step": 28190
    },
    {
      "entropy": 5.244341421127319,
      "epoch": 2.7982334259626835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042240013285044215,
      "loss": 4.9211,
      "mean_token_accuracy": 0.18371790945529937,
      "num_tokens": 61766024.0,
      "step": 28195
    },
    {
      "entropy": 5.35431056022644,
      "epoch": 2.798729654624851,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00042237336139254336,
      "loss": 4.9877,
      "mean_token_accuracy": 0.18821009248495102,
      "num_tokens": 61778228.0,
      "step": 28200
    },
    {
      "entropy": 5.388200759887695,
      "epoch": 2.7992258832870185,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042234658628010274,
      "loss": 5.0291,
      "mean_token_accuracy": 0.17172736376523973,
      "num_tokens": 61789072.0,
      "step": 28205
    },
    {
      "entropy": 5.333230257034302,
      "epoch": 2.799722111949186,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042231980751378435,
      "loss": 4.9744,
      "mean_token_accuracy": 0.18652440160512923,
      "num_tokens": 61799503.0,
      "step": 28210
    },
    {
      "entropy": 5.316104459762573,
      "epoch": 2.8002183406113534,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042229302509425205,
      "loss": 4.9307,
      "mean_token_accuracy": 0.18451789319515227,
      "num_tokens": 61811063.0,
      "step": 28215
    },
    {
      "entropy": 5.32419638633728,
      "epoch": 2.8007145692735214,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004222662390221702,
      "loss": 5.041,
      "mean_token_accuracy": 0.17430002093315125,
      "num_tokens": 61822832.0,
      "step": 28220
    },
    {
      "entropy": 5.347461175918579,
      "epoch": 2.801210797935689,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004222394492982029,
      "loss": 5.0197,
      "mean_token_accuracy": 0.17383750677108764,
      "num_tokens": 61832456.0,
      "step": 28225
    },
    {
      "entropy": 5.341380405426025,
      "epoch": 2.8017070265978563,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004222126559230146,
      "loss": 4.9464,
      "mean_token_accuracy": 0.18413645625114441,
      "num_tokens": 61843043.0,
      "step": 28230
    },
    {
      "entropy": 5.285549068450928,
      "epoch": 2.802203255260024,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004221858588972696,
      "loss": 5.0058,
      "mean_token_accuracy": 0.17677100896835327,
      "num_tokens": 61853249.0,
      "step": 28235
    },
    {
      "entropy": 5.261868524551391,
      "epoch": 2.8026994839221913,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042215905822163245,
      "loss": 4.9303,
      "mean_token_accuracy": 0.1883348450064659,
      "num_tokens": 61863842.0,
      "step": 28240
    },
    {
      "entropy": 5.309464597702027,
      "epoch": 2.8031957125843587,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004221322538967678,
      "loss": 5.0085,
      "mean_token_accuracy": 0.18028521686792373,
      "num_tokens": 61876924.0,
      "step": 28245
    },
    {
      "entropy": 5.409192514419556,
      "epoch": 2.8036919412465267,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004221054459233403,
      "loss": 5.0688,
      "mean_token_accuracy": 0.17630613297224046,
      "num_tokens": 61886766.0,
      "step": 28250
    },
    {
      "entropy": 5.405657052993774,
      "epoch": 2.804188169908694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042207863430201466,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18303633630275726,
      "num_tokens": 61896538.0,
      "step": 28255
    },
    {
      "entropy": 5.301174545288086,
      "epoch": 2.8046843985708616,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042205181903345586,
      "loss": 4.9745,
      "mean_token_accuracy": 0.1738964021205902,
      "num_tokens": 61907328.0,
      "step": 28260
    },
    {
      "entropy": 5.355307960510254,
      "epoch": 2.805180627233029,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004220250001183288,
      "loss": 5.0197,
      "mean_token_accuracy": 0.1724324882030487,
      "num_tokens": 61918797.0,
      "step": 28265
    },
    {
      "entropy": 5.384719467163086,
      "epoch": 2.8056768558951966,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004219981775572986,
      "loss": 5.0365,
      "mean_token_accuracy": 0.1806182712316513,
      "num_tokens": 61928755.0,
      "step": 28270
    },
    {
      "entropy": 5.3331231594085695,
      "epoch": 2.806173084557364,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004219713513510303,
      "loss": 5.0164,
      "mean_token_accuracy": 0.1750733286142349,
      "num_tokens": 61939993.0,
      "step": 28275
    },
    {
      "entropy": 5.280187177658081,
      "epoch": 2.8066693132195315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042194452150018917,
      "loss": 4.8747,
      "mean_token_accuracy": 0.1938151940703392,
      "num_tokens": 61950982.0,
      "step": 28280
    },
    {
      "entropy": 5.358029127120972,
      "epoch": 2.807165541881699,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042191768800544055,
      "loss": 4.9675,
      "mean_token_accuracy": 0.17528049349784852,
      "num_tokens": 61961786.0,
      "step": 28285
    },
    {
      "entropy": 5.343008184432984,
      "epoch": 2.8076617705438665,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042189085086744986,
      "loss": 4.9644,
      "mean_token_accuracy": 0.18765897750854493,
      "num_tokens": 61972548.0,
      "step": 28290
    },
    {
      "entropy": 5.309186506271362,
      "epoch": 2.808157999206034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004218640100868826,
      "loss": 4.9456,
      "mean_token_accuracy": 0.18425666391849518,
      "num_tokens": 61981822.0,
      "step": 28295
    },
    {
      "entropy": 5.348912715911865,
      "epoch": 2.8086542278682014,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042183716566440433,
      "loss": 5.0812,
      "mean_token_accuracy": 0.16871012151241302,
      "num_tokens": 61992997.0,
      "step": 28300
    },
    {
      "entropy": 5.360820245742798,
      "epoch": 2.8091504565303693,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004218103176006808,
      "loss": 5.0096,
      "mean_token_accuracy": 0.17764952927827835,
      "num_tokens": 62004391.0,
      "step": 28305
    },
    {
      "entropy": 5.3346266746521,
      "epoch": 2.809646685192537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004217834658963777,
      "loss": 4.9832,
      "mean_token_accuracy": 0.1842896267771721,
      "num_tokens": 62014099.0,
      "step": 28310
    },
    {
      "entropy": 5.3511176109313965,
      "epoch": 2.8101429138547043,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042175661055216093,
      "loss": 4.9483,
      "mean_token_accuracy": 0.18184658735990525,
      "num_tokens": 62023872.0,
      "step": 28315
    },
    {
      "entropy": 5.3753588676452635,
      "epoch": 2.8106391425168717,
      "grad_norm": 1.125,
      "learning_rate": 0.0004217297515686965,
      "loss": 4.9452,
      "mean_token_accuracy": 0.1809636950492859,
      "num_tokens": 62035566.0,
      "step": 28320
    },
    {
      "entropy": 5.314786386489868,
      "epoch": 2.811135371179039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004217028889466503,
      "loss": 4.9793,
      "mean_token_accuracy": 0.17864156812429427,
      "num_tokens": 62046742.0,
      "step": 28325
    },
    {
      "entropy": 5.296075773239136,
      "epoch": 2.8116315998412067,
      "grad_norm": 1.125,
      "learning_rate": 0.0004216760226866887,
      "loss": 4.9378,
      "mean_token_accuracy": 0.1850101187825203,
      "num_tokens": 62056545.0,
      "step": 28330
    },
    {
      "entropy": 5.347076416015625,
      "epoch": 2.8121278285033746,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042164915278947775,
      "loss": 5.0541,
      "mean_token_accuracy": 0.1730403020977974,
      "num_tokens": 62066414.0,
      "step": 28335
    },
    {
      "entropy": 5.265140676498413,
      "epoch": 2.812624057165542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004216222792556838,
      "loss": 4.9518,
      "mean_token_accuracy": 0.17865752577781677,
      "num_tokens": 62078406.0,
      "step": 28340
    },
    {
      "entropy": 5.325258827209472,
      "epoch": 2.8131202858277096,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004215954020859732,
      "loss": 4.9931,
      "mean_token_accuracy": 0.1801100492477417,
      "num_tokens": 62089516.0,
      "step": 28345
    },
    {
      "entropy": 5.372987127304077,
      "epoch": 2.813616514489877,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042156852128101264,
      "loss": 4.9572,
      "mean_token_accuracy": 0.1844891145825386,
      "num_tokens": 62099287.0,
      "step": 28350
    },
    {
      "entropy": 5.29378604888916,
      "epoch": 2.8141127431520445,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042154163684146855,
      "loss": 4.9287,
      "mean_token_accuracy": 0.18595947623252868,
      "num_tokens": 62111110.0,
      "step": 28355
    },
    {
      "entropy": 5.452878522872925,
      "epoch": 2.814608971814212,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004215147487680076,
      "loss": 5.1059,
      "mean_token_accuracy": 0.17028351724147797,
      "num_tokens": 62123614.0,
      "step": 28360
    },
    {
      "entropy": 5.348181676864624,
      "epoch": 2.8151052004763795,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042148785706129657,
      "loss": 5.0499,
      "mean_token_accuracy": 0.17662809938192367,
      "num_tokens": 62133682.0,
      "step": 28365
    },
    {
      "entropy": 5.286633777618408,
      "epoch": 2.815601429138547,
      "grad_norm": 1.125,
      "learning_rate": 0.0004214609617220024,
      "loss": 4.9634,
      "mean_token_accuracy": 0.1847042813897133,
      "num_tokens": 62144802.0,
      "step": 28370
    },
    {
      "entropy": 5.33760552406311,
      "epoch": 2.8160976578007144,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042143406275079186,
      "loss": 4.9395,
      "mean_token_accuracy": 0.1777850866317749,
      "num_tokens": 62154205.0,
      "step": 28375
    },
    {
      "entropy": 5.308980989456177,
      "epoch": 2.816593886462882,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004214071601483322,
      "loss": 4.8706,
      "mean_token_accuracy": 0.19300298690795897,
      "num_tokens": 62164648.0,
      "step": 28380
    },
    {
      "entropy": 5.307291746139526,
      "epoch": 2.8170901151250494,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004213802539152905,
      "loss": 4.961,
      "mean_token_accuracy": 0.18389626145362853,
      "num_tokens": 62176073.0,
      "step": 28385
    },
    {
      "entropy": 5.369343900680542,
      "epoch": 2.8175863437872173,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042135334405233384,
      "loss": 5.0084,
      "mean_token_accuracy": 0.17695395797491073,
      "num_tokens": 62187350.0,
      "step": 28390
    },
    {
      "entropy": 5.337830018997193,
      "epoch": 2.8180825724493848,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004213264305601296,
      "loss": 4.9311,
      "mean_token_accuracy": 0.18000833243131636,
      "num_tokens": 62197729.0,
      "step": 28395
    },
    {
      "entropy": 5.290300130844116,
      "epoch": 2.8185788011115522,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042129951343934527,
      "loss": 4.9212,
      "mean_token_accuracy": 0.18515839725732802,
      "num_tokens": 62209047.0,
      "step": 28400
    },
    {
      "entropy": 5.30870509147644,
      "epoch": 2.8190750297737197,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004212725926906482,
      "loss": 4.9578,
      "mean_token_accuracy": 0.18230582922697067,
      "num_tokens": 62218859.0,
      "step": 28405
    },
    {
      "entropy": 5.279337930679321,
      "epoch": 2.819571258435887,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042124566831470604,
      "loss": 4.9681,
      "mean_token_accuracy": 0.1803799480199814,
      "num_tokens": 62230396.0,
      "step": 28410
    },
    {
      "entropy": 5.325257396697998,
      "epoch": 2.8200674870980547,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004212187403121864,
      "loss": 4.9458,
      "mean_token_accuracy": 0.18602275848388672,
      "num_tokens": 62241790.0,
      "step": 28415
    },
    {
      "entropy": 5.314579200744629,
      "epoch": 2.8205637157602226,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042119180868375707,
      "loss": 4.9331,
      "mean_token_accuracy": 0.17874497622251512,
      "num_tokens": 62253820.0,
      "step": 28420
    },
    {
      "entropy": 5.2990094184875485,
      "epoch": 2.82105994442239,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004211648734300859,
      "loss": 4.9626,
      "mean_token_accuracy": 0.18174932897090912,
      "num_tokens": 62263875.0,
      "step": 28425
    },
    {
      "entropy": 5.332892179489136,
      "epoch": 2.8215561730845575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042113793455184084,
      "loss": 5.0181,
      "mean_token_accuracy": 0.17616894096136093,
      "num_tokens": 62275390.0,
      "step": 28430
    },
    {
      "entropy": 5.328651380538941,
      "epoch": 2.822052401746725,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004211109920496899,
      "loss": 5.038,
      "mean_token_accuracy": 0.17267679721117019,
      "num_tokens": 62286969.0,
      "step": 28435
    },
    {
      "entropy": 5.3543251037597654,
      "epoch": 2.8225486304088925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042108404592430126,
      "loss": 4.9821,
      "mean_token_accuracy": 0.17534731775522233,
      "num_tokens": 62297810.0,
      "step": 28440
    },
    {
      "entropy": 5.317121744155884,
      "epoch": 2.82304485907106,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000421057096176343,
      "loss": 4.9714,
      "mean_token_accuracy": 0.18414086997509002,
      "num_tokens": 62309474.0,
      "step": 28445
    },
    {
      "entropy": 5.2782172679901125,
      "epoch": 2.8235410877332274,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004210301428064835,
      "loss": 5.0182,
      "mean_token_accuracy": 0.17552927136421204,
      "num_tokens": 62320022.0,
      "step": 28450
    },
    {
      "entropy": 5.386905527114868,
      "epoch": 2.824037316395395,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004210031858153911,
      "loss": 5.0012,
      "mean_token_accuracy": 0.18529821634292604,
      "num_tokens": 62330450.0,
      "step": 28455
    },
    {
      "entropy": 5.34531888961792,
      "epoch": 2.8245335450575624,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042097622520373434,
      "loss": 5.001,
      "mean_token_accuracy": 0.18125528544187547,
      "num_tokens": 62341419.0,
      "step": 28460
    },
    {
      "entropy": 5.338266277313233,
      "epoch": 2.82502977371973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004209492609721817,
      "loss": 5.0355,
      "mean_token_accuracy": 0.17591184377670288,
      "num_tokens": 62353312.0,
      "step": 28465
    },
    {
      "entropy": 5.34740138053894,
      "epoch": 2.8255260023818973,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004209222931214019,
      "loss": 4.9952,
      "mean_token_accuracy": 0.1823977917432785,
      "num_tokens": 62364838.0,
      "step": 28470
    },
    {
      "entropy": 5.304411935806274,
      "epoch": 2.8260222310440652,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042089532165206367,
      "loss": 4.898,
      "mean_token_accuracy": 0.1921560749411583,
      "num_tokens": 62376443.0,
      "step": 28475
    },
    {
      "entropy": 5.380113649368286,
      "epoch": 2.8265184597062327,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004208683465648358,
      "loss": 5.0827,
      "mean_token_accuracy": 0.17383290976285934,
      "num_tokens": 62387263.0,
      "step": 28480
    },
    {
      "entropy": 5.324349737167358,
      "epoch": 2.8270146883684,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042084136786038737,
      "loss": 4.9083,
      "mean_token_accuracy": 0.18108803927898406,
      "num_tokens": 62398929.0,
      "step": 28485
    },
    {
      "entropy": 5.326780652999878,
      "epoch": 2.8275109170305677,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004208143855393872,
      "loss": 5.0237,
      "mean_token_accuracy": 0.17776139825582504,
      "num_tokens": 62409516.0,
      "step": 28490
    },
    {
      "entropy": 5.3600239753723145,
      "epoch": 2.828007145692735,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042078739960250446,
      "loss": 4.9809,
      "mean_token_accuracy": 0.18170999437570573,
      "num_tokens": 62421856.0,
      "step": 28495
    },
    {
      "entropy": 5.354677343368531,
      "epoch": 2.8285033743549026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042076041005040835,
      "loss": 4.9392,
      "mean_token_accuracy": 0.17897017896175385,
      "num_tokens": 62432340.0,
      "step": 28500
    },
    {
      "entropy": 5.3026224136352536,
      "epoch": 2.8289996030170705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004207334168837682,
      "loss": 5.0055,
      "mean_token_accuracy": 0.1730238080024719,
      "num_tokens": 62442652.0,
      "step": 28505
    },
    {
      "entropy": 5.348048067092895,
      "epoch": 2.829495831679238,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004207064201032533,
      "loss": 4.9142,
      "mean_token_accuracy": 0.1889326199889183,
      "num_tokens": 62454503.0,
      "step": 28510
    },
    {
      "entropy": 5.275537538528442,
      "epoch": 2.8299920603414055,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042067941970953316,
      "loss": 4.9252,
      "mean_token_accuracy": 0.1816031217575073,
      "num_tokens": 62464858.0,
      "step": 28515
    },
    {
      "entropy": 5.367615604400635,
      "epoch": 2.830488289003573,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042065241570327734,
      "loss": 5.0088,
      "mean_token_accuracy": 0.1780984118580818,
      "num_tokens": 62475153.0,
      "step": 28520
    },
    {
      "entropy": 5.347154808044434,
      "epoch": 2.8309845176657404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004206254080851555,
      "loss": 5.0433,
      "mean_token_accuracy": 0.17556291818618774,
      "num_tokens": 62487883.0,
      "step": 28525
    },
    {
      "entropy": 5.326135873794556,
      "epoch": 2.831480746327908,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042059839685583726,
      "loss": 4.9625,
      "mean_token_accuracy": 0.18209240287542344,
      "num_tokens": 62499905.0,
      "step": 28530
    },
    {
      "entropy": 5.369292688369751,
      "epoch": 2.8319769749900754,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042057138201599247,
      "loss": 5.0238,
      "mean_token_accuracy": 0.17580707669258117,
      "num_tokens": 62511403.0,
      "step": 28535
    },
    {
      "entropy": 5.3417102813720705,
      "epoch": 2.832473203652243,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042054436356629113,
      "loss": 4.9722,
      "mean_token_accuracy": 0.1872175633907318,
      "num_tokens": 62522543.0,
      "step": 28540
    },
    {
      "entropy": 5.28001971244812,
      "epoch": 2.8329694323144103,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004205173415074032,
      "loss": 4.9712,
      "mean_token_accuracy": 0.1794374480843544,
      "num_tokens": 62533053.0,
      "step": 28545
    },
    {
      "entropy": 5.348807144165039,
      "epoch": 2.833465660976578,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004204903158399989,
      "loss": 5.0654,
      "mean_token_accuracy": 0.1723019078373909,
      "num_tokens": 62543699.0,
      "step": 28550
    },
    {
      "entropy": 5.331068277359009,
      "epoch": 2.8339618896387453,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004204632865647481,
      "loss": 5.0061,
      "mean_token_accuracy": 0.18357067555189133,
      "num_tokens": 62554121.0,
      "step": 28555
    },
    {
      "entropy": 5.323069763183594,
      "epoch": 2.834458118300913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004204362536823213,
      "loss": 4.9374,
      "mean_token_accuracy": 0.18329390287399291,
      "num_tokens": 62564010.0,
      "step": 28560
    },
    {
      "entropy": 5.315274572372436,
      "epoch": 2.8349543469630807,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042040921719338877,
      "loss": 4.9135,
      "mean_token_accuracy": 0.18982397317886351,
      "num_tokens": 62574451.0,
      "step": 28565
    },
    {
      "entropy": 5.323036956787109,
      "epoch": 2.835450575625248,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000420382177098621,
      "loss": 4.9886,
      "mean_token_accuracy": 0.17836477607488632,
      "num_tokens": 62585298.0,
      "step": 28570
    },
    {
      "entropy": 5.3201014518737795,
      "epoch": 2.8359468042874156,
      "grad_norm": 1.0,
      "learning_rate": 0.00042035513339868855,
      "loss": 5.0527,
      "mean_token_accuracy": 0.17610400170087814,
      "num_tokens": 62596248.0,
      "step": 28575
    },
    {
      "entropy": 5.3747447490692135,
      "epoch": 2.836443032949583,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004203280860942619,
      "loss": 5.0165,
      "mean_token_accuracy": 0.17364106327295303,
      "num_tokens": 62607233.0,
      "step": 28580
    },
    {
      "entropy": 5.344322061538696,
      "epoch": 2.8369392616117506,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004203010351860119,
      "loss": 4.9257,
      "mean_token_accuracy": 0.18842774033546447,
      "num_tokens": 62618564.0,
      "step": 28585
    },
    {
      "entropy": 5.299536180496216,
      "epoch": 2.837435490273918,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004202739806746094,
      "loss": 5.0067,
      "mean_token_accuracy": 0.17533232867717743,
      "num_tokens": 62630049.0,
      "step": 28590
    },
    {
      "entropy": 5.3506951332092285,
      "epoch": 2.837931718936086,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004202469225607252,
      "loss": 5.0593,
      "mean_token_accuracy": 0.17071030884981156,
      "num_tokens": 62640887.0,
      "step": 28595
    },
    {
      "entropy": 5.361901044845581,
      "epoch": 2.8384279475982535,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004202198608450302,
      "loss": 5.0067,
      "mean_token_accuracy": 0.17504161298274995,
      "num_tokens": 62651816.0,
      "step": 28600
    },
    {
      "entropy": 5.315093612670898,
      "epoch": 2.838924176260421,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004201927955281957,
      "loss": 4.9583,
      "mean_token_accuracy": 0.18272241353988647,
      "num_tokens": 62662601.0,
      "step": 28605
    },
    {
      "entropy": 5.30598669052124,
      "epoch": 2.8394204049225884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004201657266108926,
      "loss": 4.9699,
      "mean_token_accuracy": 0.18645646125078202,
      "num_tokens": 62673661.0,
      "step": 28610
    },
    {
      "entropy": 5.299740362167358,
      "epoch": 2.839916633584756,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042013865409379244,
      "loss": 4.9378,
      "mean_token_accuracy": 0.18396537750959396,
      "num_tokens": 62683984.0,
      "step": 28615
    },
    {
      "entropy": 5.366539859771729,
      "epoch": 2.8404128622469234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004201115779775663,
      "loss": 5.002,
      "mean_token_accuracy": 0.17824754118919373,
      "num_tokens": 62695338.0,
      "step": 28620
    },
    {
      "entropy": 5.389687252044678,
      "epoch": 2.840909090909091,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004200844982628857,
      "loss": 5.0423,
      "mean_token_accuracy": 0.17737592607736588,
      "num_tokens": 62706341.0,
      "step": 28625
    },
    {
      "entropy": 5.3766717433929445,
      "epoch": 2.8414053195712583,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004200574149504222,
      "loss": 5.0347,
      "mean_token_accuracy": 0.18338865488767625,
      "num_tokens": 62717895.0,
      "step": 28630
    },
    {
      "entropy": 5.3138671875,
      "epoch": 2.841901548233426,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004200303280408474,
      "loss": 4.9899,
      "mean_token_accuracy": 0.17860433608293533,
      "num_tokens": 62729745.0,
      "step": 28635
    },
    {
      "entropy": 5.33027024269104,
      "epoch": 2.8423977768955933,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042000323753483296,
      "loss": 4.9843,
      "mean_token_accuracy": 0.1862964540719986,
      "num_tokens": 62740298.0,
      "step": 28640
    },
    {
      "entropy": 5.3177018642425535,
      "epoch": 2.842894005557761,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041997614343305067,
      "loss": 4.9568,
      "mean_token_accuracy": 0.18641080856323242,
      "num_tokens": 62751292.0,
      "step": 28645
    },
    {
      "entropy": 5.3603846549987795,
      "epoch": 2.8433902342199286,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041994904573617247,
      "loss": 5.0796,
      "mean_token_accuracy": 0.17370579838752748,
      "num_tokens": 62761789.0,
      "step": 28650
    },
    {
      "entropy": 5.335276889801025,
      "epoch": 2.843886462882096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041992194444487026,
      "loss": 4.9698,
      "mean_token_accuracy": 0.18337650299072267,
      "num_tokens": 62773317.0,
      "step": 28655
    },
    {
      "entropy": 5.316733646392822,
      "epoch": 2.8443826915442636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041989483955981607,
      "loss": 4.9383,
      "mean_token_accuracy": 0.18867828845977783,
      "num_tokens": 62783780.0,
      "step": 28660
    },
    {
      "entropy": 5.344378423690796,
      "epoch": 2.844878920206431,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041986773108168216,
      "loss": 4.9667,
      "mean_token_accuracy": 0.17634293287992478,
      "num_tokens": 62794535.0,
      "step": 28665
    },
    {
      "entropy": 5.291914939880371,
      "epoch": 2.8453751488685985,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041984061901114057,
      "loss": 4.9524,
      "mean_token_accuracy": 0.17850553542375563,
      "num_tokens": 62804410.0,
      "step": 28670
    },
    {
      "entropy": 5.321073818206787,
      "epoch": 2.845871377530766,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004198135033488638,
      "loss": 4.9388,
      "mean_token_accuracy": 0.18938398808240892,
      "num_tokens": 62814693.0,
      "step": 28675
    },
    {
      "entropy": 5.3087911128997805,
      "epoch": 2.846367606192934,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041978638409552417,
      "loss": 4.9544,
      "mean_token_accuracy": 0.18652634024620057,
      "num_tokens": 62826114.0,
      "step": 28680
    },
    {
      "entropy": 5.286051797866821,
      "epoch": 2.8468638348551014,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004197592612517942,
      "loss": 5.0048,
      "mean_token_accuracy": 0.18087669014930724,
      "num_tokens": 62838130.0,
      "step": 28685
    },
    {
      "entropy": 5.4439493179321286,
      "epoch": 2.847360063517269,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004197321348183465,
      "loss": 5.0458,
      "mean_token_accuracy": 0.1740319013595581,
      "num_tokens": 62849290.0,
      "step": 28690
    },
    {
      "entropy": 5.42426028251648,
      "epoch": 2.8478562921794364,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041970500479585364,
      "loss": 5.1215,
      "mean_token_accuracy": 0.1718273863196373,
      "num_tokens": 62860888.0,
      "step": 28695
    },
    {
      "entropy": 5.386332082748413,
      "epoch": 2.848352520841604,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041967787118498854,
      "loss": 5.0206,
      "mean_token_accuracy": 0.18257911056280135,
      "num_tokens": 62870231.0,
      "step": 28700
    },
    {
      "entropy": 5.319769334793091,
      "epoch": 2.8488487495037713,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004196507339864241,
      "loss": 4.9266,
      "mean_token_accuracy": 0.18139995634555817,
      "num_tokens": 62881147.0,
      "step": 28705
    },
    {
      "entropy": 5.298535108566284,
      "epoch": 2.849344978165939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000419623593200833,
      "loss": 4.9519,
      "mean_token_accuracy": 0.18597617149353027,
      "num_tokens": 62891272.0,
      "step": 28710
    },
    {
      "entropy": 5.261644268035889,
      "epoch": 2.8498412068281063,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041959644882888844,
      "loss": 4.9333,
      "mean_token_accuracy": 0.18193840682506562,
      "num_tokens": 62902376.0,
      "step": 28715
    },
    {
      "entropy": 5.300510454177856,
      "epoch": 2.8503374354902737,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041956930087126357,
      "loss": 4.9394,
      "mean_token_accuracy": 0.18938708901405335,
      "num_tokens": 62913811.0,
      "step": 28720
    },
    {
      "entropy": 5.331725978851319,
      "epoch": 2.850833664152441,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041954214932863156,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18056388199329376,
      "num_tokens": 62926001.0,
      "step": 28725
    },
    {
      "entropy": 5.306574296951294,
      "epoch": 2.851329892814609,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004195149942016657,
      "loss": 4.9631,
      "mean_token_accuracy": 0.1789489731192589,
      "num_tokens": 62936183.0,
      "step": 28730
    },
    {
      "entropy": 5.29832968711853,
      "epoch": 2.8518261214767766,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004194878354910394,
      "loss": 4.9377,
      "mean_token_accuracy": 0.18582096248865126,
      "num_tokens": 62947427.0,
      "step": 28735
    },
    {
      "entropy": 5.370442008972168,
      "epoch": 2.852322350138944,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041946067319742605,
      "loss": 5.0251,
      "mean_token_accuracy": 0.1780203565955162,
      "num_tokens": 62958630.0,
      "step": 28740
    },
    {
      "entropy": 5.301569318771362,
      "epoch": 2.8528185788011116,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041943350732149935,
      "loss": 4.9417,
      "mean_token_accuracy": 0.18061609268188478,
      "num_tokens": 62968856.0,
      "step": 28745
    },
    {
      "entropy": 5.323074245452881,
      "epoch": 2.853314807463279,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041940633786393295,
      "loss": 4.8653,
      "mean_token_accuracy": 0.18698527216911315,
      "num_tokens": 62978393.0,
      "step": 28750
    },
    {
      "entropy": 5.277708244323731,
      "epoch": 2.8538110361254465,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041937916482540045,
      "loss": 4.9989,
      "mean_token_accuracy": 0.18233876675367355,
      "num_tokens": 62989786.0,
      "step": 28755
    },
    {
      "entropy": 5.306382083892823,
      "epoch": 2.854307264787614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041935198820657575,
      "loss": 4.9841,
      "mean_token_accuracy": 0.18499622941017152,
      "num_tokens": 63000021.0,
      "step": 28760
    },
    {
      "entropy": 5.3530580520629885,
      "epoch": 2.854803493449782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004193248080081329,
      "loss": 4.9961,
      "mean_token_accuracy": 0.17429332733154296,
      "num_tokens": 63011398.0,
      "step": 28765
    },
    {
      "entropy": 5.3369709014892575,
      "epoch": 2.8552997221119494,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004192976242307457,
      "loss": 4.9562,
      "mean_token_accuracy": 0.17797858268022537,
      "num_tokens": 63021155.0,
      "step": 28770
    },
    {
      "entropy": 5.335000610351562,
      "epoch": 2.855795950774117,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004192704368750884,
      "loss": 4.9992,
      "mean_token_accuracy": 0.1822325825691223,
      "num_tokens": 63031510.0,
      "step": 28775
    },
    {
      "entropy": 5.413480281829834,
      "epoch": 2.8562921794362843,
      "grad_norm": 1.125,
      "learning_rate": 0.0004192432459418351,
      "loss": 5.0001,
      "mean_token_accuracy": 0.17438175678253173,
      "num_tokens": 63041978.0,
      "step": 28780
    },
    {
      "entropy": 5.390053558349609,
      "epoch": 2.856788408098452,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004192160514316602,
      "loss": 5.0391,
      "mean_token_accuracy": 0.17801456302404403,
      "num_tokens": 63052194.0,
      "step": 28785
    },
    {
      "entropy": 5.2590227127075195,
      "epoch": 2.8572846367606193,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004191888533452379,
      "loss": 4.9002,
      "mean_token_accuracy": 0.18612256795167922,
      "num_tokens": 63062144.0,
      "step": 28790
    },
    {
      "entropy": 5.269210243225098,
      "epoch": 2.8577808654227868,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004191616516832428,
      "loss": 5.0031,
      "mean_token_accuracy": 0.17344932854175568,
      "num_tokens": 63072403.0,
      "step": 28795
    },
    {
      "entropy": 5.307751321792603,
      "epoch": 2.8582770940849542,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004191344464463494,
      "loss": 4.9392,
      "mean_token_accuracy": 0.18377008885145188,
      "num_tokens": 63084799.0,
      "step": 28800
    },
    {
      "entropy": 5.365159702301026,
      "epoch": 2.8587733227471217,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041910723763523224,
      "loss": 5.0326,
      "mean_token_accuracy": 0.18690432459115983,
      "num_tokens": 63095437.0,
      "step": 28805
    },
    {
      "entropy": 5.346981143951416,
      "epoch": 2.859269551409289,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004190800252505661,
      "loss": 5.0513,
      "mean_token_accuracy": 0.17484375685453415,
      "num_tokens": 63106743.0,
      "step": 28810
    },
    {
      "entropy": 5.317916631698608,
      "epoch": 2.8597657800714567,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004190528092930259,
      "loss": 4.9532,
      "mean_token_accuracy": 0.18433489948511123,
      "num_tokens": 63120525.0,
      "step": 28815
    },
    {
      "entropy": 5.351857328414917,
      "epoch": 2.8602620087336246,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004190255897632863,
      "loss": 4.9748,
      "mean_token_accuracy": 0.18190867602825164,
      "num_tokens": 63131510.0,
      "step": 28820
    },
    {
      "entropy": 5.401505613327027,
      "epoch": 2.860758237395792,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041899836666202254,
      "loss": 5.0341,
      "mean_token_accuracy": 0.18591573387384414,
      "num_tokens": 63143094.0,
      "step": 28825
    },
    {
      "entropy": 5.258724308013916,
      "epoch": 2.8612544660579595,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004189711399899095,
      "loss": 4.9472,
      "mean_token_accuracy": 0.18559656292200089,
      "num_tokens": 63155159.0,
      "step": 28830
    },
    {
      "entropy": 5.270547199249267,
      "epoch": 2.861750694720127,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004189439097476224,
      "loss": 4.9935,
      "mean_token_accuracy": 0.18833862245082855,
      "num_tokens": 63167132.0,
      "step": 28835
    },
    {
      "entropy": 5.313321876525879,
      "epoch": 2.8622469233822945,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004189166759358366,
      "loss": 4.9931,
      "mean_token_accuracy": 0.1842258393764496,
      "num_tokens": 63176864.0,
      "step": 28840
    },
    {
      "entropy": 5.380536460876465,
      "epoch": 2.862743152044462,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041888943855522727,
      "loss": 4.9765,
      "mean_token_accuracy": 0.18127398937940598,
      "num_tokens": 63188003.0,
      "step": 28845
    },
    {
      "entropy": 5.291923379898071,
      "epoch": 2.86323938070663,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041886219760646997,
      "loss": 4.9121,
      "mean_token_accuracy": 0.18079456686973572,
      "num_tokens": 63197634.0,
      "step": 28850
    },
    {
      "entropy": 5.250195550918579,
      "epoch": 2.8637356093687973,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004188349530902402,
      "loss": 4.8868,
      "mean_token_accuracy": 0.18869269490242005,
      "num_tokens": 63207441.0,
      "step": 28855
    },
    {
      "entropy": 5.2829657077789305,
      "epoch": 2.864231838030965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041880770500721346,
      "loss": 4.9463,
      "mean_token_accuracy": 0.18215555399656297,
      "num_tokens": 63217541.0,
      "step": 28860
    },
    {
      "entropy": 5.269743394851685,
      "epoch": 2.8647280666931323,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004187804533580655,
      "loss": 4.8981,
      "mean_token_accuracy": 0.19140853583812714,
      "num_tokens": 63228391.0,
      "step": 28865
    },
    {
      "entropy": 5.317827320098877,
      "epoch": 2.8652242953552998,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004187531981434721,
      "loss": 4.9453,
      "mean_token_accuracy": 0.18625912964344024,
      "num_tokens": 63238462.0,
      "step": 28870
    },
    {
      "entropy": 5.3214695930480955,
      "epoch": 2.8657205240174672,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041872593936410923,
      "loss": 5.0057,
      "mean_token_accuracy": 0.17980576753616334,
      "num_tokens": 63249305.0,
      "step": 28875
    },
    {
      "entropy": 5.390914964675903,
      "epoch": 2.8662167526796347,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041869867702065266,
      "loss": 5.0228,
      "mean_token_accuracy": 0.18009461611509323,
      "num_tokens": 63260832.0,
      "step": 28880
    },
    {
      "entropy": 5.313502359390259,
      "epoch": 2.866712981341802,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004186714111137787,
      "loss": 4.9881,
      "mean_token_accuracy": 0.1750674545764923,
      "num_tokens": 63273689.0,
      "step": 28885
    },
    {
      "entropy": 5.40730128288269,
      "epoch": 2.8672092100039697,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004186441416441632,
      "loss": 5.1296,
      "mean_token_accuracy": 0.17074090987443924,
      "num_tokens": 63284912.0,
      "step": 28890
    },
    {
      "entropy": 5.402767610549927,
      "epoch": 2.867705438666137,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004186168686124825,
      "loss": 5.0256,
      "mean_token_accuracy": 0.17077893614768982,
      "num_tokens": 63296428.0,
      "step": 28895
    },
    {
      "entropy": 5.316618347167969,
      "epoch": 2.8682016673283046,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041858959201941307,
      "loss": 4.9404,
      "mean_token_accuracy": 0.1884831517934799,
      "num_tokens": 63307509.0,
      "step": 28900
    },
    {
      "entropy": 5.368279027938843,
      "epoch": 2.8686978959904725,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041856231186563107,
      "loss": 5.0702,
      "mean_token_accuracy": 0.1736469730734825,
      "num_tokens": 63319194.0,
      "step": 28905
    },
    {
      "entropy": 5.292404413223267,
      "epoch": 2.86919412465264,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041853502815181306,
      "loss": 4.8934,
      "mean_token_accuracy": 0.18540094196796417,
      "num_tokens": 63329122.0,
      "step": 28910
    },
    {
      "entropy": 5.33369460105896,
      "epoch": 2.8696903533148075,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004185077408786357,
      "loss": 4.9873,
      "mean_token_accuracy": 0.1796829655766487,
      "num_tokens": 63340304.0,
      "step": 28915
    },
    {
      "entropy": 5.361721134185791,
      "epoch": 2.870186581976975,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004184804500467756,
      "loss": 4.9752,
      "mean_token_accuracy": 0.1792677253484726,
      "num_tokens": 63351029.0,
      "step": 28920
    },
    {
      "entropy": 5.310054540634155,
      "epoch": 2.8706828106391424,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041845315565690945,
      "loss": 4.9147,
      "mean_token_accuracy": 0.1805896505713463,
      "num_tokens": 63361339.0,
      "step": 28925
    },
    {
      "entropy": 5.380335474014283,
      "epoch": 2.87117903930131,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041842585770971417,
      "loss": 4.9771,
      "mean_token_accuracy": 0.18444427400827407,
      "num_tokens": 63373431.0,
      "step": 28930
    },
    {
      "entropy": 5.285171318054199,
      "epoch": 2.871675267963478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004183985562058667,
      "loss": 4.9396,
      "mean_token_accuracy": 0.18243106305599213,
      "num_tokens": 63384096.0,
      "step": 28935
    },
    {
      "entropy": 5.359778022766113,
      "epoch": 2.8721714966256453,
      "grad_norm": 1.046875,
      "learning_rate": 0.000418371251146044,
      "loss": 5.008,
      "mean_token_accuracy": 0.17886601239442826,
      "num_tokens": 63394852.0,
      "step": 28940
    },
    {
      "entropy": 5.351619100570678,
      "epoch": 2.872667725287813,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004183439425309232,
      "loss": 4.9631,
      "mean_token_accuracy": 0.18507122099399567,
      "num_tokens": 63405575.0,
      "step": 28945
    },
    {
      "entropy": 5.350751209259033,
      "epoch": 2.8731639539499803,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004183166303611816,
      "loss": 5.0223,
      "mean_token_accuracy": 0.18359985649585725,
      "num_tokens": 63416677.0,
      "step": 28950
    },
    {
      "entropy": 5.33743462562561,
      "epoch": 2.8736601826121477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004182893146374963,
      "loss": 5.0812,
      "mean_token_accuracy": 0.17758579403162003,
      "num_tokens": 63428613.0,
      "step": 28955
    },
    {
      "entropy": 5.35674262046814,
      "epoch": 2.874156411274315,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041826199536054476,
      "loss": 5.0151,
      "mean_token_accuracy": 0.17985695451498032,
      "num_tokens": 63439321.0,
      "step": 28960
    },
    {
      "entropy": 5.389108800888062,
      "epoch": 2.8746526399364827,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004182346725310045,
      "loss": 5.0842,
      "mean_token_accuracy": 0.17556892931461335,
      "num_tokens": 63451279.0,
      "step": 28965
    },
    {
      "entropy": 5.390461015701294,
      "epoch": 2.87514886859865,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041820734614955295,
      "loss": 4.9862,
      "mean_token_accuracy": 0.18234420120716094,
      "num_tokens": 63460801.0,
      "step": 28970
    },
    {
      "entropy": 5.31269383430481,
      "epoch": 2.8756450972608176,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004181800162168678,
      "loss": 4.9661,
      "mean_token_accuracy": 0.17931545227766038,
      "num_tokens": 63470232.0,
      "step": 28975
    },
    {
      "entropy": 5.31996865272522,
      "epoch": 2.876141325922985,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004181526827336268,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18269616663455962,
      "num_tokens": 63480247.0,
      "step": 28980
    },
    {
      "entropy": 5.32126898765564,
      "epoch": 2.8766375545851526,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004181253457005077,
      "loss": 4.9809,
      "mean_token_accuracy": 0.17480241358280182,
      "num_tokens": 63492104.0,
      "step": 28985
    },
    {
      "entropy": 5.338863945007324,
      "epoch": 2.8771337832473205,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041809800511818853,
      "loss": 4.99,
      "mean_token_accuracy": 0.18438510447740555,
      "num_tokens": 63502784.0,
      "step": 28990
    },
    {
      "entropy": 5.391718578338623,
      "epoch": 2.877630011909488,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041807066098734704,
      "loss": 5.058,
      "mean_token_accuracy": 0.1773552864789963,
      "num_tokens": 63514150.0,
      "step": 28995
    },
    {
      "entropy": 5.367831993103027,
      "epoch": 2.8781262405716554,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004180433133086615,
      "loss": 5.0158,
      "mean_token_accuracy": 0.18280314058065414,
      "num_tokens": 63526682.0,
      "step": 29000
    },
    {
      "entropy": 5.359856367111206,
      "epoch": 2.878622469233823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004180159620828101,
      "loss": 5.0169,
      "mean_token_accuracy": 0.1761864259839058,
      "num_tokens": 63537448.0,
      "step": 29005
    },
    {
      "entropy": 5.313276720046997,
      "epoch": 2.8791186978959904,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004179886073104709,
      "loss": 4.9349,
      "mean_token_accuracy": 0.18224285542964935,
      "num_tokens": 63547489.0,
      "step": 29010
    },
    {
      "entropy": 5.319760274887085,
      "epoch": 2.879614926558158,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004179612489923224,
      "loss": 4.944,
      "mean_token_accuracy": 0.1896757110953331,
      "num_tokens": 63558324.0,
      "step": 29015
    },
    {
      "entropy": 5.339535713195801,
      "epoch": 2.880111155220326,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000417933887129043,
      "loss": 4.9564,
      "mean_token_accuracy": 0.18310637176036834,
      "num_tokens": 63568630.0,
      "step": 29020
    },
    {
      "entropy": 5.258089685440064,
      "epoch": 2.8806073838824933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004179065217213111,
      "loss": 4.9057,
      "mean_token_accuracy": 0.18073288947343827,
      "num_tokens": 63578068.0,
      "step": 29025
    },
    {
      "entropy": 5.310549736022949,
      "epoch": 2.8811036125446607,
      "grad_norm": 1.0,
      "learning_rate": 0.00041787915276980556,
      "loss": 5.0272,
      "mean_token_accuracy": 0.17643349319696428,
      "num_tokens": 63589103.0,
      "step": 29030
    },
    {
      "entropy": 5.32974944114685,
      "epoch": 2.881599841206828,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041785178027520486,
      "loss": 4.9631,
      "mean_token_accuracy": 0.18235402256250383,
      "num_tokens": 63600305.0,
      "step": 29035
    },
    {
      "entropy": 5.419808435440063,
      "epoch": 2.8820960698689957,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00041782440423818775,
      "loss": 5.037,
      "mean_token_accuracy": 0.18012840300798416,
      "num_tokens": 63611910.0,
      "step": 29040
    },
    {
      "entropy": 5.353421449661255,
      "epoch": 2.882592298531163,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004177970246594332,
      "loss": 4.9221,
      "mean_token_accuracy": 0.18235496431589127,
      "num_tokens": 63622674.0,
      "step": 29045
    },
    {
      "entropy": 5.346738004684449,
      "epoch": 2.8830885271933306,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041776964153962017,
      "loss": 5.0319,
      "mean_token_accuracy": 0.1720239743590355,
      "num_tokens": 63633274.0,
      "step": 29050
    },
    {
      "entropy": 5.358149003982544,
      "epoch": 2.883584755855498,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041774225487942765,
      "loss": 4.9779,
      "mean_token_accuracy": 0.18668194413185119,
      "num_tokens": 63644738.0,
      "step": 29055
    },
    {
      "entropy": 5.338395404815674,
      "epoch": 2.8840809845176656,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004177148646795348,
      "loss": 5.0181,
      "mean_token_accuracy": 0.17970823496580124,
      "num_tokens": 63655468.0,
      "step": 29060
    },
    {
      "entropy": 5.295706939697266,
      "epoch": 2.884577213179833,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041768747094062084,
      "loss": 4.8708,
      "mean_token_accuracy": 0.1846926987171173,
      "num_tokens": 63667675.0,
      "step": 29065
    },
    {
      "entropy": 5.317497682571411,
      "epoch": 2.8850734418420005,
      "grad_norm": 0.890625,
      "learning_rate": 0.00041766007366336494,
      "loss": 5.0143,
      "mean_token_accuracy": 0.1810229167342186,
      "num_tokens": 63679251.0,
      "step": 29070
    },
    {
      "entropy": 5.264214420318604,
      "epoch": 2.8855696705041685,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041763267284844675,
      "loss": 4.9378,
      "mean_token_accuracy": 0.18634056597948073,
      "num_tokens": 63690809.0,
      "step": 29075
    },
    {
      "entropy": 5.320691680908203,
      "epoch": 2.886065899166336,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004176052684965456,
      "loss": 4.9261,
      "mean_token_accuracy": 0.18037503212690353,
      "num_tokens": 63702434.0,
      "step": 29080
    },
    {
      "entropy": 5.357242631912231,
      "epoch": 2.8865621278285034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004175778606083411,
      "loss": 5.0451,
      "mean_token_accuracy": 0.17739060670137405,
      "num_tokens": 63713173.0,
      "step": 29085
    },
    {
      "entropy": 5.36808180809021,
      "epoch": 2.887058356490671,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004175504491845128,
      "loss": 5.0151,
      "mean_token_accuracy": 0.1768234297633171,
      "num_tokens": 63723956.0,
      "step": 29090
    },
    {
      "entropy": 5.33939037322998,
      "epoch": 2.8875545851528384,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041752303422574067,
      "loss": 5.039,
      "mean_token_accuracy": 0.17711836844682693,
      "num_tokens": 63735280.0,
      "step": 29095
    },
    {
      "entropy": 5.41951813697815,
      "epoch": 2.888050813815006,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004174956157327043,
      "loss": 5.0104,
      "mean_token_accuracy": 0.1700599119067192,
      "num_tokens": 63746075.0,
      "step": 29100
    },
    {
      "entropy": 5.350586080551148,
      "epoch": 2.8885470424771738,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004174681937060837,
      "loss": 4.9803,
      "mean_token_accuracy": 0.18471972346305848,
      "num_tokens": 63758152.0,
      "step": 29105
    },
    {
      "entropy": 5.309057331085205,
      "epoch": 2.8890432711393412,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041744076814655893,
      "loss": 5.0102,
      "mean_token_accuracy": 0.18190837651491165,
      "num_tokens": 63768855.0,
      "step": 29110
    },
    {
      "entropy": 5.341919708251953,
      "epoch": 2.8895394998015087,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041741333905481,
      "loss": 5.0124,
      "mean_token_accuracy": 0.17464365363121032,
      "num_tokens": 63780258.0,
      "step": 29115
    },
    {
      "entropy": 5.384697580337525,
      "epoch": 2.890035728463676,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004173859064315172,
      "loss": 5.0664,
      "mean_token_accuracy": 0.17613840103149414,
      "num_tokens": 63791959.0,
      "step": 29120
    },
    {
      "entropy": 5.368592929840088,
      "epoch": 2.8905319571258437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041735847027736067,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18132029175758363,
      "num_tokens": 63802859.0,
      "step": 29125
    },
    {
      "entropy": 5.315375947952271,
      "epoch": 2.891028185788011,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004173310305930209,
      "loss": 4.9471,
      "mean_token_accuracy": 0.17688386738300324,
      "num_tokens": 63814275.0,
      "step": 29130
    },
    {
      "entropy": 5.402803516387939,
      "epoch": 2.8915244144501786,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041730358737917823,
      "loss": 4.9996,
      "mean_token_accuracy": 0.1818391725420952,
      "num_tokens": 63824633.0,
      "step": 29135
    },
    {
      "entropy": 5.356333541870117,
      "epoch": 2.892020643112346,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041727614063651313,
      "loss": 4.9904,
      "mean_token_accuracy": 0.18013789504766464,
      "num_tokens": 63836478.0,
      "step": 29140
    },
    {
      "entropy": 5.368706941604614,
      "epoch": 2.8925168717745136,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004172486903657065,
      "loss": 5.0162,
      "mean_token_accuracy": 0.17755720913410186,
      "num_tokens": 63847917.0,
      "step": 29145
    },
    {
      "entropy": 5.3055493354797365,
      "epoch": 2.893013100436681,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004172212365674388,
      "loss": 5.0129,
      "mean_token_accuracy": 0.18282997757196426,
      "num_tokens": 63857258.0,
      "step": 29150
    },
    {
      "entropy": 5.323853063583374,
      "epoch": 2.8935093290988485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041719377924239083,
      "loss": 5.0561,
      "mean_token_accuracy": 0.176475365459919,
      "num_tokens": 63868627.0,
      "step": 29155
    },
    {
      "entropy": 5.395700550079345,
      "epoch": 2.8940055577610164,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041716631839124356,
      "loss": 5.0453,
      "mean_token_accuracy": 0.18252562135457992,
      "num_tokens": 63879954.0,
      "step": 29160
    },
    {
      "entropy": 5.460035800933838,
      "epoch": 2.894501786423184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000417138854014678,
      "loss": 5.1246,
      "mean_token_accuracy": 0.17249011769890785,
      "num_tokens": 63891147.0,
      "step": 29165
    },
    {
      "entropy": 5.307288789749146,
      "epoch": 2.8949980150853514,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004171113861133751,
      "loss": 4.961,
      "mean_token_accuracy": 0.1775546446442604,
      "num_tokens": 63902380.0,
      "step": 29170
    },
    {
      "entropy": 5.311139965057373,
      "epoch": 2.895494243747519,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004170839146880161,
      "loss": 5.0138,
      "mean_token_accuracy": 0.18462066501379013,
      "num_tokens": 63912557.0,
      "step": 29175
    },
    {
      "entropy": 5.3519495010375975,
      "epoch": 2.8959904724096863,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004170564397392821,
      "loss": 5.032,
      "mean_token_accuracy": 0.1745504766702652,
      "num_tokens": 63923959.0,
      "step": 29180
    },
    {
      "entropy": 5.358626747131348,
      "epoch": 2.896486701071854,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004170289612678545,
      "loss": 4.9515,
      "mean_token_accuracy": 0.18897261768579482,
      "num_tokens": 63935600.0,
      "step": 29185
    },
    {
      "entropy": 5.366093492507934,
      "epoch": 2.8969829297340213,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041700147927441477,
      "loss": 5.1203,
      "mean_token_accuracy": 0.17345927208662032,
      "num_tokens": 63946371.0,
      "step": 29190
    },
    {
      "entropy": 5.361727094650268,
      "epoch": 2.897479158396189,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041697399375964435,
      "loss": 5.0613,
      "mean_token_accuracy": 0.17407326698303222,
      "num_tokens": 63956019.0,
      "step": 29195
    },
    {
      "entropy": 5.365759944915771,
      "epoch": 2.8979753870583567,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041694650472422467,
      "loss": 4.9993,
      "mean_token_accuracy": 0.1751288577914238,
      "num_tokens": 63967330.0,
      "step": 29200
    },
    {
      "entropy": 5.323833894729614,
      "epoch": 2.898471615720524,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041691901216883763,
      "loss": 4.9729,
      "mean_token_accuracy": 0.18482379466295243,
      "num_tokens": 63978952.0,
      "step": 29205
    },
    {
      "entropy": 5.318281841278076,
      "epoch": 2.8989678443826916,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000416891516094165,
      "loss": 5.0017,
      "mean_token_accuracy": 0.1772097811102867,
      "num_tokens": 63989913.0,
      "step": 29210
    },
    {
      "entropy": 5.245384454727173,
      "epoch": 2.899464073044859,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004168640165008884,
      "loss": 4.9072,
      "mean_token_accuracy": 0.18863601982593536,
      "num_tokens": 64000033.0,
      "step": 29215
    },
    {
      "entropy": 5.33118634223938,
      "epoch": 2.8999603017070266,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041683651338968994,
      "loss": 4.9846,
      "mean_token_accuracy": 0.17771048694849015,
      "num_tokens": 64010746.0,
      "step": 29220
    },
    {
      "entropy": 5.352051162719727,
      "epoch": 2.900456530369194,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041680900676125147,
      "loss": 4.9642,
      "mean_token_accuracy": 0.18008633702993393,
      "num_tokens": 64021179.0,
      "step": 29225
    },
    {
      "entropy": 5.36590542793274,
      "epoch": 2.9009527590313615,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041678149661625525,
      "loss": 5.0139,
      "mean_token_accuracy": 0.17436424791812896,
      "num_tokens": 64032953.0,
      "step": 29230
    },
    {
      "entropy": 5.315649843215942,
      "epoch": 2.901448987693529,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004167539829553835,
      "loss": 4.859,
      "mean_token_accuracy": 0.19560223072767258,
      "num_tokens": 64042548.0,
      "step": 29235
    },
    {
      "entropy": 5.268677234649658,
      "epoch": 2.9019452163556965,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041672646577931835,
      "loss": 4.9247,
      "mean_token_accuracy": 0.18310151994228363,
      "num_tokens": 64053254.0,
      "step": 29240
    },
    {
      "entropy": 5.270452785491943,
      "epoch": 2.9024414450178644,
      "grad_norm": 0.90625,
      "learning_rate": 0.00041669894508874225,
      "loss": 4.9783,
      "mean_token_accuracy": 0.17453655004501342,
      "num_tokens": 64063307.0,
      "step": 29245
    },
    {
      "entropy": 5.3248037815094,
      "epoch": 2.902937673680032,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041667142088433766,
      "loss": 4.9764,
      "mean_token_accuracy": 0.18568291813135146,
      "num_tokens": 64074274.0,
      "step": 29250
    },
    {
      "entropy": 5.203045988082886,
      "epoch": 2.9034339023421993,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041664389316678713,
      "loss": 4.8444,
      "mean_token_accuracy": 0.19932950884103776,
      "num_tokens": 64084243.0,
      "step": 29255
    },
    {
      "entropy": 5.319317483901978,
      "epoch": 2.903930131004367,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004166163619367733,
      "loss": 4.9381,
      "mean_token_accuracy": 0.18613807260990142,
      "num_tokens": 64095899.0,
      "step": 29260
    },
    {
      "entropy": 5.301117706298828,
      "epoch": 2.9044263596665343,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041658882719497876,
      "loss": 4.8813,
      "mean_token_accuracy": 0.18109363913536072,
      "num_tokens": 64107408.0,
      "step": 29265
    },
    {
      "entropy": 5.3749391555786135,
      "epoch": 2.9049225883287018,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004165612889420864,
      "loss": 5.0343,
      "mean_token_accuracy": 0.17790033519268036,
      "num_tokens": 64118935.0,
      "step": 29270
    },
    {
      "entropy": 5.339265918731689,
      "epoch": 2.9054188169908692,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004165337471787792,
      "loss": 4.9901,
      "mean_token_accuracy": 0.1823713168501854,
      "num_tokens": 64129085.0,
      "step": 29275
    },
    {
      "entropy": 5.245773220062256,
      "epoch": 2.905915045653037,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041650620190574004,
      "loss": 4.8914,
      "mean_token_accuracy": 0.18996512591838838,
      "num_tokens": 64139605.0,
      "step": 29280
    },
    {
      "entropy": 5.358142471313476,
      "epoch": 2.9064112743152046,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041647865312365186,
      "loss": 5.0288,
      "mean_token_accuracy": 0.17983788549900054,
      "num_tokens": 64151171.0,
      "step": 29285
    },
    {
      "entropy": 5.39275598526001,
      "epoch": 2.906907502977372,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041645110083319805,
      "loss": 5.0119,
      "mean_token_accuracy": 0.17245693057775496,
      "num_tokens": 64162417.0,
      "step": 29290
    },
    {
      "entropy": 5.401563596725464,
      "epoch": 2.9074037316395396,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041642354503506175,
      "loss": 4.9896,
      "mean_token_accuracy": 0.17450710386037827,
      "num_tokens": 64172755.0,
      "step": 29295
    },
    {
      "entropy": 5.347092866897583,
      "epoch": 2.907899960301707,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041639598572992625,
      "loss": 4.9818,
      "mean_token_accuracy": 0.17739432007074357,
      "num_tokens": 64184270.0,
      "step": 29300
    },
    {
      "entropy": 5.3803246974945065,
      "epoch": 2.9083961889638745,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041636842291847495,
      "loss": 5.0489,
      "mean_token_accuracy": 0.17380618155002595,
      "num_tokens": 64195400.0,
      "step": 29305
    },
    {
      "entropy": 5.331266117095947,
      "epoch": 2.908892417626042,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041634085660139145,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18300728350877762,
      "num_tokens": 64205101.0,
      "step": 29310
    },
    {
      "entropy": 5.276601314544678,
      "epoch": 2.9093886462882095,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004163132867793592,
      "loss": 4.8942,
      "mean_token_accuracy": 0.18357529789209365,
      "num_tokens": 64216047.0,
      "step": 29315
    },
    {
      "entropy": 5.306755828857422,
      "epoch": 2.909884874950377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000416285713453062,
      "loss": 4.9781,
      "mean_token_accuracy": 0.18375121802091599,
      "num_tokens": 64226195.0,
      "step": 29320
    },
    {
      "entropy": 5.331162071228027,
      "epoch": 2.9103811036125444,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004162581366231835,
      "loss": 4.9705,
      "mean_token_accuracy": 0.18514053225517274,
      "num_tokens": 64236405.0,
      "step": 29325
    },
    {
      "entropy": 5.298386812210083,
      "epoch": 2.9108773322747123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041623055629040764,
      "loss": 4.9336,
      "mean_token_accuracy": 0.182055826485157,
      "num_tokens": 64246989.0,
      "step": 29330
    },
    {
      "entropy": 5.278272914886474,
      "epoch": 2.91137356093688,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004162029724554183,
      "loss": 4.9515,
      "mean_token_accuracy": 0.17964468151330948,
      "num_tokens": 64257928.0,
      "step": 29335
    },
    {
      "entropy": 5.324585390090943,
      "epoch": 2.9118697895990473,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041617538511889946,
      "loss": 4.9847,
      "mean_token_accuracy": 0.1804472878575325,
      "num_tokens": 64268985.0,
      "step": 29340
    },
    {
      "entropy": 5.478241682052612,
      "epoch": 2.9123660182612148,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004161477942815353,
      "loss": 5.1393,
      "mean_token_accuracy": 0.17161781936883927,
      "num_tokens": 64279907.0,
      "step": 29345
    },
    {
      "entropy": 5.40117883682251,
      "epoch": 2.9128622469233822,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041612019994401,
      "loss": 4.9848,
      "mean_token_accuracy": 0.17496951371431352,
      "num_tokens": 64291496.0,
      "step": 29350
    },
    {
      "entropy": 5.304608726501465,
      "epoch": 2.9133584755855497,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004160926021070078,
      "loss": 4.9651,
      "mean_token_accuracy": 0.18839789032936097,
      "num_tokens": 64302160.0,
      "step": 29355
    },
    {
      "entropy": 5.362702512741089,
      "epoch": 2.913854704247717,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004160650007712132,
      "loss": 5.0255,
      "mean_token_accuracy": 0.18021430522203447,
      "num_tokens": 64313671.0,
      "step": 29360
    },
    {
      "entropy": 5.327708005905151,
      "epoch": 2.914350932909885,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041603739593731046,
      "loss": 5.0132,
      "mean_token_accuracy": 0.1768110916018486,
      "num_tokens": 64325376.0,
      "step": 29365
    },
    {
      "entropy": 5.359637880325318,
      "epoch": 2.9148471615720526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004160097876059842,
      "loss": 4.9952,
      "mean_token_accuracy": 0.1808785155415535,
      "num_tokens": 64336048.0,
      "step": 29370
    },
    {
      "entropy": 5.395231628417969,
      "epoch": 2.91534339023422,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004159821757779191,
      "loss": 5.0262,
      "mean_token_accuracy": 0.17413267940282823,
      "num_tokens": 64346833.0,
      "step": 29375
    },
    {
      "entropy": 5.331878185272217,
      "epoch": 2.9158396188963875,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041595456045379975,
      "loss": 4.9725,
      "mean_token_accuracy": 0.17915357351303102,
      "num_tokens": 64359656.0,
      "step": 29380
    },
    {
      "entropy": 5.288068103790283,
      "epoch": 2.916335847558555,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004159269416343111,
      "loss": 4.9224,
      "mean_token_accuracy": 0.1837301656603813,
      "num_tokens": 64371130.0,
      "step": 29385
    },
    {
      "entropy": 5.254090261459351,
      "epoch": 2.9168320762207225,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000415899319320138,
      "loss": 4.9659,
      "mean_token_accuracy": 0.18586190938949584,
      "num_tokens": 64381826.0,
      "step": 29390
    },
    {
      "entropy": 5.380335950851441,
      "epoch": 2.91732830488289,
      "grad_norm": 1.0,
      "learning_rate": 0.00041587169351196536,
      "loss": 5.0837,
      "mean_token_accuracy": 0.17130287289619445,
      "num_tokens": 64392725.0,
      "step": 29395
    },
    {
      "entropy": 5.402438974380493,
      "epoch": 2.9178245335450574,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041584406421047825,
      "loss": 4.9715,
      "mean_token_accuracy": 0.18290422707796097,
      "num_tokens": 64402587.0,
      "step": 29400
    },
    {
      "entropy": 5.359978151321411,
      "epoch": 2.918320762207225,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041581643141636184,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18210332691669465,
      "num_tokens": 64414692.0,
      "step": 29405
    },
    {
      "entropy": 5.334445476531982,
      "epoch": 2.9188169908693924,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004157887951303014,
      "loss": 5.0383,
      "mean_token_accuracy": 0.17257201820611953,
      "num_tokens": 64427278.0,
      "step": 29410
    },
    {
      "entropy": 5.3124712944030765,
      "epoch": 2.91931321953156,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004157611553529822,
      "loss": 4.946,
      "mean_token_accuracy": 0.18134967535734176,
      "num_tokens": 64438221.0,
      "step": 29415
    },
    {
      "entropy": 5.342079114913941,
      "epoch": 2.919809448193728,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041573351208508966,
      "loss": 5.0761,
      "mean_token_accuracy": 0.17538690865039824,
      "num_tokens": 64450139.0,
      "step": 29420
    },
    {
      "entropy": 5.378389120101929,
      "epoch": 2.9203056768558953,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041570586532730926,
      "loss": 4.9859,
      "mean_token_accuracy": 0.17863744795322417,
      "num_tokens": 64461523.0,
      "step": 29425
    },
    {
      "entropy": 5.338352298736572,
      "epoch": 2.9208019055180627,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004156782150803266,
      "loss": 5.061,
      "mean_token_accuracy": 0.17280968129634858,
      "num_tokens": 64474206.0,
      "step": 29430
    },
    {
      "entropy": 5.298125886917115,
      "epoch": 2.92129813418023,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041565056134482733,
      "loss": 5.0089,
      "mean_token_accuracy": 0.1776416465640068,
      "num_tokens": 64485585.0,
      "step": 29435
    },
    {
      "entropy": 5.4275120258331295,
      "epoch": 2.9217943628423977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004156229041214972,
      "loss": 5.0383,
      "mean_token_accuracy": 0.1791038081049919,
      "num_tokens": 64497237.0,
      "step": 29440
    },
    {
      "entropy": 5.3512836456298825,
      "epoch": 2.922290591504565,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004155952434110221,
      "loss": 4.9695,
      "mean_token_accuracy": 0.1803577572107315,
      "num_tokens": 64508235.0,
      "step": 29445
    },
    {
      "entropy": 5.390390253067016,
      "epoch": 2.922786820166733,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004155675792140879,
      "loss": 5.0098,
      "mean_token_accuracy": 0.1837974414229393,
      "num_tokens": 64520345.0,
      "step": 29450
    },
    {
      "entropy": 5.366340065002442,
      "epoch": 2.9232830488289006,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004155399115313806,
      "loss": 5.0141,
      "mean_token_accuracy": 0.17860111892223357,
      "num_tokens": 64531428.0,
      "step": 29455
    },
    {
      "entropy": 5.389728116989136,
      "epoch": 2.923779277491068,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041551224036358644,
      "loss": 5.0193,
      "mean_token_accuracy": 0.17727542370557786,
      "num_tokens": 64541489.0,
      "step": 29460
    },
    {
      "entropy": 5.254123544692993,
      "epoch": 2.9242755061532355,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041548456571139145,
      "loss": 4.8982,
      "mean_token_accuracy": 0.18916889727115632,
      "num_tokens": 64553026.0,
      "step": 29465
    },
    {
      "entropy": 5.355836868286133,
      "epoch": 2.924771734815403,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041545688757548184,
      "loss": 5.086,
      "mean_token_accuracy": 0.17783331871032715,
      "num_tokens": 64565988.0,
      "step": 29470
    },
    {
      "entropy": 5.357552909851075,
      "epoch": 2.9252679634775705,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041542920595654423,
      "loss": 4.9573,
      "mean_token_accuracy": 0.18164631724357605,
      "num_tokens": 64578098.0,
      "step": 29475
    },
    {
      "entropy": 5.285634613037109,
      "epoch": 2.925764192139738,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041540152085526476,
      "loss": 4.8783,
      "mean_token_accuracy": 0.1825031355023384,
      "num_tokens": 64588931.0,
      "step": 29480
    },
    {
      "entropy": 5.24184741973877,
      "epoch": 2.9262604208019054,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041537383227233023,
      "loss": 4.8797,
      "mean_token_accuracy": 0.1936873510479927,
      "num_tokens": 64599636.0,
      "step": 29485
    },
    {
      "entropy": 5.345392656326294,
      "epoch": 2.926756649464073,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004153461402084271,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18283215016126633,
      "num_tokens": 64609355.0,
      "step": 29490
    },
    {
      "entropy": 5.296345806121826,
      "epoch": 2.9272528781262404,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004153184446642421,
      "loss": 4.9429,
      "mean_token_accuracy": 0.1878829911351204,
      "num_tokens": 64619969.0,
      "step": 29495
    },
    {
      "entropy": 5.30253586769104,
      "epoch": 2.927749106788408,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041529074564046204,
      "loss": 4.931,
      "mean_token_accuracy": 0.19008105099201203,
      "num_tokens": 64629538.0,
      "step": 29500
    },
    {
      "entropy": 5.365242862701416,
      "epoch": 2.9282453354505757,
      "grad_norm": 1.0,
      "learning_rate": 0.0004152630431377738,
      "loss": 4.9856,
      "mean_token_accuracy": 0.18366603553295135,
      "num_tokens": 64640255.0,
      "step": 29505
    },
    {
      "entropy": 5.2740825653076175,
      "epoch": 2.9287415641127432,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004152353371568643,
      "loss": 4.8924,
      "mean_token_accuracy": 0.18824109733104705,
      "num_tokens": 64650051.0,
      "step": 29510
    },
    {
      "entropy": 5.281096410751343,
      "epoch": 2.9292377927749107,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041520762769842065,
      "loss": 4.957,
      "mean_token_accuracy": 0.1885521426796913,
      "num_tokens": 64660575.0,
      "step": 29515
    },
    {
      "entropy": 5.372226762771606,
      "epoch": 2.929734021437078,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041517991476312994,
      "loss": 5.0201,
      "mean_token_accuracy": 0.183464515209198,
      "num_tokens": 64671486.0,
      "step": 29520
    },
    {
      "entropy": 5.353983402252197,
      "epoch": 2.9302302500992456,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004151521983516794,
      "loss": 4.9058,
      "mean_token_accuracy": 0.18331848978996276,
      "num_tokens": 64682315.0,
      "step": 29525
    },
    {
      "entropy": 5.362731218338013,
      "epoch": 2.930726478761413,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004151244784647563,
      "loss": 5.0576,
      "mean_token_accuracy": 0.18200620859861374,
      "num_tokens": 64693940.0,
      "step": 29530
    },
    {
      "entropy": 5.339909315109253,
      "epoch": 2.931222707423581,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004150967551030482,
      "loss": 4.9706,
      "mean_token_accuracy": 0.18206655234098434,
      "num_tokens": 64703347.0,
      "step": 29535
    },
    {
      "entropy": 5.308848762512207,
      "epoch": 2.9317189360857485,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041506902826724234,
      "loss": 5.0172,
      "mean_token_accuracy": 0.1767240807414055,
      "num_tokens": 64714115.0,
      "step": 29540
    },
    {
      "entropy": 5.29938006401062,
      "epoch": 2.932215164747916,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004150412979580264,
      "loss": 4.9096,
      "mean_token_accuracy": 0.18565237820148467,
      "num_tokens": 64724888.0,
      "step": 29545
    },
    {
      "entropy": 5.4102271556854244,
      "epoch": 2.9327113934100835,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041501356417608816,
      "loss": 5.0274,
      "mean_token_accuracy": 0.17518818825483323,
      "num_tokens": 64735626.0,
      "step": 29550
    },
    {
      "entropy": 5.333011817932129,
      "epoch": 2.933207622072251,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041498582692211513,
      "loss": 4.9422,
      "mean_token_accuracy": 0.182943893969059,
      "num_tokens": 64747807.0,
      "step": 29555
    },
    {
      "entropy": 5.25140380859375,
      "epoch": 2.9337038507344184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041495808619679525,
      "loss": 4.9114,
      "mean_token_accuracy": 0.18851151019334794,
      "num_tokens": 64756845.0,
      "step": 29560
    },
    {
      "entropy": 5.291245698928833,
      "epoch": 2.934200079396586,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041493034200081643,
      "loss": 5.0228,
      "mean_token_accuracy": 0.17873993664979934,
      "num_tokens": 64767316.0,
      "step": 29565
    },
    {
      "entropy": 5.346328592300415,
      "epoch": 2.9346963080587534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004149025943348666,
      "loss": 4.9415,
      "mean_token_accuracy": 0.18808119148015975,
      "num_tokens": 64778329.0,
      "step": 29570
    },
    {
      "entropy": 5.320685625076294,
      "epoch": 2.935192536720921,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041487484319963395,
      "loss": 4.9884,
      "mean_token_accuracy": 0.18530306965112686,
      "num_tokens": 64789215.0,
      "step": 29575
    },
    {
      "entropy": 5.308544683456421,
      "epoch": 2.9356887653830883,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004148470885958066,
      "loss": 4.9501,
      "mean_token_accuracy": 0.1829306200146675,
      "num_tokens": 64801878.0,
      "step": 29580
    },
    {
      "entropy": 5.277546739578247,
      "epoch": 2.936184994045256,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004148193305240727,
      "loss": 4.8767,
      "mean_token_accuracy": 0.18137367367744445,
      "num_tokens": 64813442.0,
      "step": 29585
    },
    {
      "entropy": 5.312920570373535,
      "epoch": 2.9366812227074237,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041479156898512076,
      "loss": 4.9484,
      "mean_token_accuracy": 0.17792298644781113,
      "num_tokens": 64824095.0,
      "step": 29590
    },
    {
      "entropy": 5.345976114273071,
      "epoch": 2.937177451369591,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004147638039796391,
      "loss": 4.9725,
      "mean_token_accuracy": 0.18473187685012818,
      "num_tokens": 64833101.0,
      "step": 29595
    },
    {
      "entropy": 5.24795241355896,
      "epoch": 2.9376736800317587,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041473603550831636,
      "loss": 4.8981,
      "mean_token_accuracy": 0.18209708780050277,
      "num_tokens": 64844315.0,
      "step": 29600
    },
    {
      "entropy": 5.31929874420166,
      "epoch": 2.938169908693926,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041470826357184096,
      "loss": 4.9702,
      "mean_token_accuracy": 0.17970002442598343,
      "num_tokens": 64855090.0,
      "step": 29605
    },
    {
      "entropy": 5.2824718952178955,
      "epoch": 2.9386661373560936,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004146804881709017,
      "loss": 4.9349,
      "mean_token_accuracy": 0.17954566180706025,
      "num_tokens": 64865093.0,
      "step": 29610
    },
    {
      "entropy": 5.431281328201294,
      "epoch": 2.939162366018261,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004146527093061873,
      "loss": 5.055,
      "mean_token_accuracy": 0.17588279247283936,
      "num_tokens": 64876250.0,
      "step": 29615
    },
    {
      "entropy": 5.365763473510742,
      "epoch": 2.939658594680429,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004146249269783867,
      "loss": 4.9685,
      "mean_token_accuracy": 0.18396117240190507,
      "num_tokens": 64886339.0,
      "step": 29620
    },
    {
      "entropy": 5.348762083053589,
      "epoch": 2.9401548233425965,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004145971411881886,
      "loss": 5.0706,
      "mean_token_accuracy": 0.17358540296554564,
      "num_tokens": 64899473.0,
      "step": 29625
    },
    {
      "entropy": 5.377105569839477,
      "epoch": 2.940651052004764,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041456935193628236,
      "loss": 5.005,
      "mean_token_accuracy": 0.18035409599542618,
      "num_tokens": 64909919.0,
      "step": 29630
    },
    {
      "entropy": 5.360543298721313,
      "epoch": 2.9411472806669314,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004145415592233569,
      "loss": 4.9991,
      "mean_token_accuracy": 0.17764416933059693,
      "num_tokens": 64921069.0,
      "step": 29635
    },
    {
      "entropy": 5.2443712711334225,
      "epoch": 2.941643509329099,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041451376305010144,
      "loss": 4.9301,
      "mean_token_accuracy": 0.1899644747376442,
      "num_tokens": 64932173.0,
      "step": 29640
    },
    {
      "entropy": 5.318042898178101,
      "epoch": 2.9421397379912664,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004144859634172053,
      "loss": 4.8839,
      "mean_token_accuracy": 0.18925838470458983,
      "num_tokens": 64941650.0,
      "step": 29645
    },
    {
      "entropy": 5.273416328430176,
      "epoch": 2.942635966653434,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041445816032535784,
      "loss": 4.9264,
      "mean_token_accuracy": 0.1873728632926941,
      "num_tokens": 64952520.0,
      "step": 29650
    },
    {
      "entropy": 5.298548793792724,
      "epoch": 2.9431321953156013,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041443035377524844,
      "loss": 5.0911,
      "mean_token_accuracy": 0.17358495593070983,
      "num_tokens": 64965030.0,
      "step": 29655
    },
    {
      "entropy": 5.309768772125244,
      "epoch": 2.943628423977769,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004144025437675668,
      "loss": 4.9223,
      "mean_token_accuracy": 0.1848897859454155,
      "num_tokens": 64976260.0,
      "step": 29660
    },
    {
      "entropy": 5.4065842628479,
      "epoch": 2.9441246526399363,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004143747303030025,
      "loss": 5.066,
      "mean_token_accuracy": 0.18031062185764313,
      "num_tokens": 64988049.0,
      "step": 29665
    },
    {
      "entropy": 5.327679920196533,
      "epoch": 2.9446208813021038,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041434691338224506,
      "loss": 4.944,
      "mean_token_accuracy": 0.18356522619724275,
      "num_tokens": 64998506.0,
      "step": 29670
    },
    {
      "entropy": 5.25973596572876,
      "epoch": 2.9451171099642717,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004143190930059846,
      "loss": 4.9515,
      "mean_token_accuracy": 0.18412467688322068,
      "num_tokens": 65010533.0,
      "step": 29675
    },
    {
      "entropy": 5.37233076095581,
      "epoch": 2.945613338626439,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004142912691749107,
      "loss": 5.0174,
      "mean_token_accuracy": 0.1825263023376465,
      "num_tokens": 65021238.0,
      "step": 29680
    },
    {
      "entropy": 5.405715799331665,
      "epoch": 2.9461095672886066,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004142634418897135,
      "loss": 5.0584,
      "mean_token_accuracy": 0.17280271351337434,
      "num_tokens": 65033251.0,
      "step": 29685
    },
    {
      "entropy": 5.356669187545776,
      "epoch": 2.946605795950774,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004142356111510831,
      "loss": 4.9759,
      "mean_token_accuracy": 0.1765642762184143,
      "num_tokens": 65043845.0,
      "step": 29690
    },
    {
      "entropy": 5.292839527130127,
      "epoch": 2.9471020246129416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041420777695970945,
      "loss": 4.9671,
      "mean_token_accuracy": 0.18664987981319428,
      "num_tokens": 65055395.0,
      "step": 29695
    },
    {
      "entropy": 5.35147008895874,
      "epoch": 2.947598253275109,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000414179939316283,
      "loss": 5.0025,
      "mean_token_accuracy": 0.18018846064805985,
      "num_tokens": 65064931.0,
      "step": 29700
    },
    {
      "entropy": 5.366870498657226,
      "epoch": 2.948094481937277,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041415209822149397,
      "loss": 4.953,
      "mean_token_accuracy": 0.18154469579458238,
      "num_tokens": 65074678.0,
      "step": 29705
    },
    {
      "entropy": 5.363426208496094,
      "epoch": 2.9485907105994444,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004141242536760327,
      "loss": 4.9764,
      "mean_token_accuracy": 0.18792126178741456,
      "num_tokens": 65084562.0,
      "step": 29710
    },
    {
      "entropy": 5.3345130443572994,
      "epoch": 2.949086939261612,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004140964056805897,
      "loss": 4.9711,
      "mean_token_accuracy": 0.18069835007190704,
      "num_tokens": 65095924.0,
      "step": 29715
    },
    {
      "entropy": 5.354391622543335,
      "epoch": 2.9495831679237794,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004140685542358556,
      "loss": 5.0055,
      "mean_token_accuracy": 0.18347228318452835,
      "num_tokens": 65106802.0,
      "step": 29720
    },
    {
      "entropy": 5.298479795455933,
      "epoch": 2.950079396585947,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004140406993425211,
      "loss": 5.0377,
      "mean_token_accuracy": 0.17900027632713317,
      "num_tokens": 65118196.0,
      "step": 29725
    },
    {
      "entropy": 5.313843822479248,
      "epoch": 2.9505756252481143,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004140128410012767,
      "loss": 4.9526,
      "mean_token_accuracy": 0.18779850900173187,
      "num_tokens": 65129924.0,
      "step": 29730
    },
    {
      "entropy": 5.231953811645508,
      "epoch": 2.951071853910282,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041398497921281354,
      "loss": 4.8158,
      "mean_token_accuracy": 0.19160735309123994,
      "num_tokens": 65140994.0,
      "step": 29735
    },
    {
      "entropy": 5.3497209548950195,
      "epoch": 2.9515680825724493,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041395711397782236,
      "loss": 5.0586,
      "mean_token_accuracy": 0.17802572697401048,
      "num_tokens": 65153261.0,
      "step": 29740
    },
    {
      "entropy": 5.433164453506469,
      "epoch": 2.9520643112346168,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004139292452969941,
      "loss": 5.0918,
      "mean_token_accuracy": 0.17667055428028106,
      "num_tokens": 65164127.0,
      "step": 29745
    },
    {
      "entropy": 5.315621662139892,
      "epoch": 2.9525605398967842,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041390137317102,
      "loss": 4.9368,
      "mean_token_accuracy": 0.17721660137176515,
      "num_tokens": 65175603.0,
      "step": 29750
    },
    {
      "entropy": 5.34569787979126,
      "epoch": 2.9530567685589517,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004138734976005911,
      "loss": 4.912,
      "mean_token_accuracy": 0.17924889773130417,
      "num_tokens": 65186780.0,
      "step": 29755
    },
    {
      "entropy": 5.307768535614014,
      "epoch": 2.9535529972211196,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004138456185863986,
      "loss": 4.8629,
      "mean_token_accuracy": 0.18764034956693648,
      "num_tokens": 65198883.0,
      "step": 29760
    },
    {
      "entropy": 5.288889694213867,
      "epoch": 2.954049225883287,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041381773612913413,
      "loss": 4.9829,
      "mean_token_accuracy": 0.17947946786880492,
      "num_tokens": 65209088.0,
      "step": 29765
    },
    {
      "entropy": 5.2462958812713625,
      "epoch": 2.9545454545454546,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004137898502294889,
      "loss": 4.9607,
      "mean_token_accuracy": 0.18396978527307511,
      "num_tokens": 65220816.0,
      "step": 29770
    },
    {
      "entropy": 5.33585557937622,
      "epoch": 2.955041683207622,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004137619608881544,
      "loss": 4.9454,
      "mean_token_accuracy": 0.18131554871797562,
      "num_tokens": 65232549.0,
      "step": 29775
    },
    {
      "entropy": 5.3686902046203615,
      "epoch": 2.9555379118697895,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004137340681058223,
      "loss": 4.9941,
      "mean_token_accuracy": 0.1769046664237976,
      "num_tokens": 65243691.0,
      "step": 29780
    },
    {
      "entropy": 5.334276342391968,
      "epoch": 2.956034140531957,
      "grad_norm": 1.0,
      "learning_rate": 0.00041370617188318424,
      "loss": 4.9953,
      "mean_token_accuracy": 0.18572455793619155,
      "num_tokens": 65254651.0,
      "step": 29785
    },
    {
      "entropy": 5.35349669456482,
      "epoch": 2.956530369194125,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041367827222093195,
      "loss": 4.9447,
      "mean_token_accuracy": 0.18842816799879075,
      "num_tokens": 65264669.0,
      "step": 29790
    },
    {
      "entropy": 5.3421708106994625,
      "epoch": 2.9570265978562924,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004136503691197574,
      "loss": 4.9819,
      "mean_token_accuracy": 0.17907123565673827,
      "num_tokens": 65274663.0,
      "step": 29795
    },
    {
      "entropy": 5.3445079803466795,
      "epoch": 2.95752282651846,
      "grad_norm": 1.0,
      "learning_rate": 0.00041362246258035254,
      "loss": 5.0034,
      "mean_token_accuracy": 0.17827296108007432,
      "num_tokens": 65286383.0,
      "step": 29800
    },
    {
      "entropy": 5.287032270431519,
      "epoch": 2.9580190551806274,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004135945526034091,
      "loss": 4.8402,
      "mean_token_accuracy": 0.18465643972158433,
      "num_tokens": 65297156.0,
      "step": 29805
    },
    {
      "entropy": 5.349915790557861,
      "epoch": 2.958515283842795,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041356663918961956,
      "loss": 5.0013,
      "mean_token_accuracy": 0.17907775342464446,
      "num_tokens": 65306836.0,
      "step": 29810
    },
    {
      "entropy": 5.337243890762329,
      "epoch": 2.9590115125049623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041353872233967595,
      "loss": 5.0308,
      "mean_token_accuracy": 0.17742842584848403,
      "num_tokens": 65318355.0,
      "step": 29815
    },
    {
      "entropy": 5.359198474884034,
      "epoch": 2.95950774116713,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004135108020542705,
      "loss": 5.0085,
      "mean_token_accuracy": 0.178192375600338,
      "num_tokens": 65330039.0,
      "step": 29820
    },
    {
      "entropy": 5.284510612487793,
      "epoch": 2.9600039698292973,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041348287833409557,
      "loss": 4.9122,
      "mean_token_accuracy": 0.18914710283279418,
      "num_tokens": 65339771.0,
      "step": 29825
    },
    {
      "entropy": 5.316507339477539,
      "epoch": 2.9605001984914647,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004134549511798438,
      "loss": 4.9897,
      "mean_token_accuracy": 0.18730810433626174,
      "num_tokens": 65352108.0,
      "step": 29830
    },
    {
      "entropy": 5.325489950180054,
      "epoch": 2.960996427153632,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004134270205922075,
      "loss": 4.9236,
      "mean_token_accuracy": 0.18001531958580017,
      "num_tokens": 65364117.0,
      "step": 29835
    },
    {
      "entropy": 5.265122365951538,
      "epoch": 2.9614926558157997,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041339908657187933,
      "loss": 4.8954,
      "mean_token_accuracy": 0.1825769767165184,
      "num_tokens": 65374168.0,
      "step": 29840
    },
    {
      "entropy": 5.425363206863404,
      "epoch": 2.9619888844779676,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004133711491195522,
      "loss": 5.1127,
      "mean_token_accuracy": 0.1727147102355957,
      "num_tokens": 65385243.0,
      "step": 29845
    },
    {
      "entropy": 5.372970867156982,
      "epoch": 2.962485113140135,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041334320823591857,
      "loss": 4.9704,
      "mean_token_accuracy": 0.1753977745771408,
      "num_tokens": 65396674.0,
      "step": 29850
    },
    {
      "entropy": 5.339597463607788,
      "epoch": 2.9629813418023025,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041331526392167155,
      "loss": 5.0375,
      "mean_token_accuracy": 0.17244871854782104,
      "num_tokens": 65408121.0,
      "step": 29855
    },
    {
      "entropy": 5.371666097640992,
      "epoch": 2.96347757046447,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000413287316177504,
      "loss": 5.0163,
      "mean_token_accuracy": 0.1786966383457184,
      "num_tokens": 65420093.0,
      "step": 29860
    },
    {
      "entropy": 5.296400260925293,
      "epoch": 2.9639737991266375,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041325936500410905,
      "loss": 4.916,
      "mean_token_accuracy": 0.18476652652025222,
      "num_tokens": 65431338.0,
      "step": 29865
    },
    {
      "entropy": 5.3382830142974855,
      "epoch": 2.964470027788805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041323141040217976,
      "loss": 4.9553,
      "mean_token_accuracy": 0.18440219163894653,
      "num_tokens": 65442305.0,
      "step": 29870
    },
    {
      "entropy": 5.2721442699432375,
      "epoch": 2.9649662564509724,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041320345237240923,
      "loss": 4.9148,
      "mean_token_accuracy": 0.1862901121377945,
      "num_tokens": 65453028.0,
      "step": 29875
    },
    {
      "entropy": 5.274634313583374,
      "epoch": 2.9654624851131404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000413175490915491,
      "loss": 5.0179,
      "mean_token_accuracy": 0.18058642745018005,
      "num_tokens": 65463374.0,
      "step": 29880
    },
    {
      "entropy": 5.3180643081665036,
      "epoch": 2.965958713775308,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004131475260321183,
      "loss": 4.9303,
      "mean_token_accuracy": 0.1826159879565239,
      "num_tokens": 65474144.0,
      "step": 29885
    },
    {
      "entropy": 5.292810487747192,
      "epoch": 2.9664549424374753,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041311955772298464,
      "loss": 4.9223,
      "mean_token_accuracy": 0.19926407635211946,
      "num_tokens": 65485453.0,
      "step": 29890
    },
    {
      "entropy": 5.3180952072143555,
      "epoch": 2.966951171099643,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004130915859887837,
      "loss": 4.9333,
      "mean_token_accuracy": 0.18065074533224107,
      "num_tokens": 65496400.0,
      "step": 29895
    },
    {
      "entropy": 5.313723134994507,
      "epoch": 2.9674473997618103,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004130636108302088,
      "loss": 5.0097,
      "mean_token_accuracy": 0.1791549354791641,
      "num_tokens": 65508524.0,
      "step": 29900
    },
    {
      "entropy": 5.173094940185547,
      "epoch": 2.9679436284239777,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041303563224795393,
      "loss": 4.8655,
      "mean_token_accuracy": 0.19273101389408112,
      "num_tokens": 65519264.0,
      "step": 29905
    },
    {
      "entropy": 5.284279489517212,
      "epoch": 2.968439857086145,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004130076502427128,
      "loss": 4.8635,
      "mean_token_accuracy": 0.18816870003938674,
      "num_tokens": 65531049.0,
      "step": 29910
    },
    {
      "entropy": 5.354379653930664,
      "epoch": 2.9689360857483127,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041297966481517937,
      "loss": 5.0143,
      "mean_token_accuracy": 0.1812281459569931,
      "num_tokens": 65543126.0,
      "step": 29915
    },
    {
      "entropy": 5.3669816017150875,
      "epoch": 2.96943231441048,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004129516759660475,
      "loss": 5.0427,
      "mean_token_accuracy": 0.16666657328605652,
      "num_tokens": 65555050.0,
      "step": 29920
    },
    {
      "entropy": 5.337154054641724,
      "epoch": 2.9699285430726476,
      "grad_norm": 1.125,
      "learning_rate": 0.00041292368369601136,
      "loss": 5.0425,
      "mean_token_accuracy": 0.18812412470579148,
      "num_tokens": 65564156.0,
      "step": 29925
    },
    {
      "entropy": 5.349378395080566,
      "epoch": 2.9704247717348156,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000412895688005765,
      "loss": 5.0604,
      "mean_token_accuracy": 0.17734704315662383,
      "num_tokens": 65574639.0,
      "step": 29930
    },
    {
      "entropy": 5.346640348434448,
      "epoch": 2.970921000396983,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004128676888960028,
      "loss": 4.9144,
      "mean_token_accuracy": 0.18335703909397125,
      "num_tokens": 65585628.0,
      "step": 29935
    },
    {
      "entropy": 5.308150911331177,
      "epoch": 2.9714172290591505,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041283968636741903,
      "loss": 4.933,
      "mean_token_accuracy": 0.1851414382457733,
      "num_tokens": 65595544.0,
      "step": 29940
    },
    {
      "entropy": 5.355164670944214,
      "epoch": 2.971913457721318,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004128116804207081,
      "loss": 5.0282,
      "mean_token_accuracy": 0.18460913896560668,
      "num_tokens": 65607963.0,
      "step": 29945
    },
    {
      "entropy": 5.367991876602173,
      "epoch": 2.9724096863834855,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004127836710565642,
      "loss": 4.9337,
      "mean_token_accuracy": 0.1914518490433693,
      "num_tokens": 65619881.0,
      "step": 29950
    },
    {
      "entropy": 5.312719106674194,
      "epoch": 2.972905915045653,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041275565827568234,
      "loss": 5.0324,
      "mean_token_accuracy": 0.1792609930038452,
      "num_tokens": 65629967.0,
      "step": 29955
    },
    {
      "entropy": 5.315414524078369,
      "epoch": 2.9734021437078204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000412727642078757,
      "loss": 4.9519,
      "mean_token_accuracy": 0.18496492207050325,
      "num_tokens": 65640804.0,
      "step": 29960
    },
    {
      "entropy": 5.38997368812561,
      "epoch": 2.9738983723699883,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004126996224664829,
      "loss": 5.0511,
      "mean_token_accuracy": 0.17424743622541428,
      "num_tokens": 65652355.0,
      "step": 29965
    },
    {
      "entropy": 5.3565497398376465,
      "epoch": 2.974394601032156,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041267159943955483,
      "loss": 5.0062,
      "mean_token_accuracy": 0.17190022319555281,
      "num_tokens": 65664047.0,
      "step": 29970
    },
    {
      "entropy": 5.3383277416229244,
      "epoch": 2.9748908296943233,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004126435729986678,
      "loss": 4.9702,
      "mean_token_accuracy": 0.18165358901023865,
      "num_tokens": 65674942.0,
      "step": 29975
    },
    {
      "entropy": 5.3307061195373535,
      "epoch": 2.9753870583564908,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041261554314451674,
      "loss": 4.9162,
      "mean_token_accuracy": 0.18578920960426332,
      "num_tokens": 65685285.0,
      "step": 29980
    },
    {
      "entropy": 5.244826555252075,
      "epoch": 2.9758832870186582,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004125875098777967,
      "loss": 4.8871,
      "mean_token_accuracy": 0.19391443133354186,
      "num_tokens": 65695967.0,
      "step": 29985
    },
    {
      "entropy": 5.304629516601563,
      "epoch": 2.9763795156808257,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041255947319920297,
      "loss": 4.9758,
      "mean_token_accuracy": 0.18005457669496536,
      "num_tokens": 65707434.0,
      "step": 29990
    },
    {
      "entropy": 5.276933097839356,
      "epoch": 2.976875744342993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041253143310943064,
      "loss": 4.937,
      "mean_token_accuracy": 0.1901909217238426,
      "num_tokens": 65717836.0,
      "step": 29995
    },
    {
      "entropy": 5.345050430297851,
      "epoch": 2.9773719730051607,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041250338960917515,
      "loss": 4.9573,
      "mean_token_accuracy": 0.18878426998853684,
      "num_tokens": 65728527.0,
      "step": 30000
    },
    {
      "epoch": 2.9773719730051607,
      "eval_entropy": 5.167751105275226,
      "eval_loss": 5.1221208572387695,
      "eval_mean_token_accuracy": 0.18055338779033775,
      "eval_num_tokens": 65728527.0,
      "eval_runtime": 26.513,
      "eval_samples_per_second": 1269.904,
      "eval_steps_per_second": 158.752,
      "step": 30000
    },
    {
      "entropy": 5.265871953964234,
      "epoch": 2.977868201667328,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041247534269913197,
      "loss": 4.928,
      "mean_token_accuracy": 0.18543122857809066,
      "num_tokens": 65739724.0,
      "step": 30005
    },
    {
      "entropy": 5.332581996917725,
      "epoch": 2.9783644303294956,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004124472923799964,
      "loss": 5.0209,
      "mean_token_accuracy": 0.17655147463083268,
      "num_tokens": 65750057.0,
      "step": 30010
    },
    {
      "entropy": 5.3218498706817625,
      "epoch": 2.9788606589916635,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004124192386524642,
      "loss": 5.0069,
      "mean_token_accuracy": 0.18372008800506592,
      "num_tokens": 65760534.0,
      "step": 30015
    },
    {
      "entropy": 5.34131350517273,
      "epoch": 2.979356887653831,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000412391181517231,
      "loss": 5.0289,
      "mean_token_accuracy": 0.18162872046232223,
      "num_tokens": 65772524.0,
      "step": 30020
    },
    {
      "entropy": 5.350955772399902,
      "epoch": 2.9798531163159985,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041236312097499257,
      "loss": 4.9783,
      "mean_token_accuracy": 0.17406929433345794,
      "num_tokens": 65783003.0,
      "step": 30025
    },
    {
      "entropy": 5.2759112358093265,
      "epoch": 2.980349344978166,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004123350570264447,
      "loss": 4.9652,
      "mean_token_accuracy": 0.17730950564146042,
      "num_tokens": 65793219.0,
      "step": 30030
    },
    {
      "entropy": 5.373185348510742,
      "epoch": 2.9808455736403334,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041230698967228334,
      "loss": 5.0677,
      "mean_token_accuracy": 0.1837572067975998,
      "num_tokens": 65804105.0,
      "step": 30035
    },
    {
      "entropy": 5.498500823974609,
      "epoch": 2.981341802302501,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004122789189132045,
      "loss": 5.0632,
      "mean_token_accuracy": 0.18126633763313293,
      "num_tokens": 65814211.0,
      "step": 30040
    },
    {
      "entropy": 5.392000293731689,
      "epoch": 2.9818380309646684,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004122508447499043,
      "loss": 5.0727,
      "mean_token_accuracy": 0.17113547027111053,
      "num_tokens": 65825304.0,
      "step": 30045
    },
    {
      "entropy": 5.256764125823975,
      "epoch": 2.9823342596268363,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041222276718307896,
      "loss": 4.891,
      "mean_token_accuracy": 0.18840173482894898,
      "num_tokens": 65835933.0,
      "step": 30050
    },
    {
      "entropy": 5.308320426940918,
      "epoch": 2.9828304882890038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004121946862134246,
      "loss": 5.0101,
      "mean_token_accuracy": 0.18038589358329774,
      "num_tokens": 65846479.0,
      "step": 30055
    },
    {
      "entropy": 5.282874393463135,
      "epoch": 2.9833267169511712,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041216660184163767,
      "loss": 4.8927,
      "mean_token_accuracy": 0.1870732918381691,
      "num_tokens": 65857033.0,
      "step": 30060
    },
    {
      "entropy": 5.37871994972229,
      "epoch": 2.9838229456133387,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041213851406841456,
      "loss": 5.0636,
      "mean_token_accuracy": 0.17830247879028321,
      "num_tokens": 65867983.0,
      "step": 30065
    },
    {
      "entropy": 5.390155458450318,
      "epoch": 2.984319174275506,
      "grad_norm": 0.875,
      "learning_rate": 0.0004121104228944518,
      "loss": 5.0109,
      "mean_token_accuracy": 0.17368921637535095,
      "num_tokens": 65878535.0,
      "step": 30070
    },
    {
      "entropy": 5.380351257324219,
      "epoch": 2.9848154029376737,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004120823283204461,
      "loss": 4.9704,
      "mean_token_accuracy": 0.18688935786485672,
      "num_tokens": 65890484.0,
      "step": 30075
    },
    {
      "entropy": 5.254448890686035,
      "epoch": 2.985311631599841,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000412054230347094,
      "loss": 4.9132,
      "mean_token_accuracy": 0.1892281174659729,
      "num_tokens": 65901220.0,
      "step": 30080
    },
    {
      "entropy": 5.364172840118409,
      "epoch": 2.9858078602620086,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004120261289750923,
      "loss": 4.9783,
      "mean_token_accuracy": 0.1777174159884453,
      "num_tokens": 65911875.0,
      "step": 30085
    },
    {
      "entropy": 5.277479791641236,
      "epoch": 2.986304088924176,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041199802420513785,
      "loss": 4.9548,
      "mean_token_accuracy": 0.18398118019104004,
      "num_tokens": 65921915.0,
      "step": 30090
    },
    {
      "entropy": 5.275164556503296,
      "epoch": 2.9868003175863436,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004119699160379276,
      "loss": 4.9145,
      "mean_token_accuracy": 0.18196410238742827,
      "num_tokens": 65932061.0,
      "step": 30095
    },
    {
      "entropy": 5.299293661117554,
      "epoch": 2.987296546248511,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041194180447415855,
      "loss": 4.9274,
      "mean_token_accuracy": 0.18856136202812196,
      "num_tokens": 65942818.0,
      "step": 30100
    },
    {
      "entropy": 5.313733148574829,
      "epoch": 2.987792774910679,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004119136895145279,
      "loss": 4.9478,
      "mean_token_accuracy": 0.18271782547235488,
      "num_tokens": 65953715.0,
      "step": 30105
    },
    {
      "entropy": 5.371411561965942,
      "epoch": 2.9882890035728464,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004118855711597327,
      "loss": 5.0168,
      "mean_token_accuracy": 0.1787567064166069,
      "num_tokens": 65964235.0,
      "step": 30110
    },
    {
      "entropy": 5.292625665664673,
      "epoch": 2.988785232235014,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041185744941047033,
      "loss": 4.9956,
      "mean_token_accuracy": 0.17721310257911682,
      "num_tokens": 65976457.0,
      "step": 30115
    },
    {
      "entropy": 5.32046480178833,
      "epoch": 2.9892814608971814,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004118293242674382,
      "loss": 4.9638,
      "mean_token_accuracy": 0.18141742795705795,
      "num_tokens": 65986810.0,
      "step": 30120
    },
    {
      "entropy": 5.362882614135742,
      "epoch": 2.989777689559349,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004118011957313336,
      "loss": 5.0624,
      "mean_token_accuracy": 0.17303382009267806,
      "num_tokens": 65998019.0,
      "step": 30125
    },
    {
      "entropy": 5.377602624893188,
      "epoch": 2.9902739182215163,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041177306380285405,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18162559270858764,
      "num_tokens": 66009404.0,
      "step": 30130
    },
    {
      "entropy": 5.273078107833863,
      "epoch": 2.9907701468836843,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004117449284826974,
      "loss": 4.8968,
      "mean_token_accuracy": 0.18744518905878066,
      "num_tokens": 66020610.0,
      "step": 30135
    },
    {
      "entropy": 5.3369756698608395,
      "epoch": 2.9912663755458517,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041171678977156104,
      "loss": 5.0114,
      "mean_token_accuracy": 0.17639297991991043,
      "num_tokens": 66032576.0,
      "step": 30140
    },
    {
      "entropy": 5.334190607070923,
      "epoch": 2.991762604208019,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041168864767014295,
      "loss": 4.9756,
      "mean_token_accuracy": 0.18723649382591248,
      "num_tokens": 66043378.0,
      "step": 30145
    },
    {
      "entropy": 5.30084114074707,
      "epoch": 2.9922588328701867,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000411660502179141,
      "loss": 4.9135,
      "mean_token_accuracy": 0.18194007128477097,
      "num_tokens": 66053960.0,
      "step": 30150
    },
    {
      "entropy": 5.306016874313355,
      "epoch": 2.992755061532354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004116323532992529,
      "loss": 4.9835,
      "mean_token_accuracy": 0.17876940965652466,
      "num_tokens": 66066069.0,
      "step": 30155
    },
    {
      "entropy": 5.2935319423675535,
      "epoch": 2.9932512901945216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041160420103117704,
      "loss": 4.9316,
      "mean_token_accuracy": 0.18616168349981307,
      "num_tokens": 66076789.0,
      "step": 30160
    },
    {
      "entropy": 5.290328598022461,
      "epoch": 2.993747518856689,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041157604537561123,
      "loss": 4.9062,
      "mean_token_accuracy": 0.1829723224043846,
      "num_tokens": 66088267.0,
      "step": 30165
    },
    {
      "entropy": 5.249536180496216,
      "epoch": 2.9942437475188566,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041154788633325387,
      "loss": 4.888,
      "mean_token_accuracy": 0.1871301308274269,
      "num_tokens": 66098789.0,
      "step": 30170
    },
    {
      "entropy": 5.284776306152343,
      "epoch": 2.994739976181024,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004115197239048031,
      "loss": 4.9601,
      "mean_token_accuracy": 0.17907484322786332,
      "num_tokens": 66110881.0,
      "step": 30175
    },
    {
      "entropy": 5.361056852340698,
      "epoch": 2.9952362048431915,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041149155809095734,
      "loss": 4.9537,
      "mean_token_accuracy": 0.1791049063205719,
      "num_tokens": 66121728.0,
      "step": 30180
    },
    {
      "entropy": 5.227986478805542,
      "epoch": 2.995732433505359,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041146338889241506,
      "loss": 4.8593,
      "mean_token_accuracy": 0.18790058344602584,
      "num_tokens": 66131386.0,
      "step": 30185
    },
    {
      "entropy": 5.338031339645386,
      "epoch": 2.996228662167527,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00041143521630987483,
      "loss": 5.0486,
      "mean_token_accuracy": 0.1788963258266449,
      "num_tokens": 66142905.0,
      "step": 30190
    },
    {
      "entropy": 5.394472265243531,
      "epoch": 2.9967248908296944,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041140704034403527,
      "loss": 5.0936,
      "mean_token_accuracy": 0.17383337765932083,
      "num_tokens": 66152708.0,
      "step": 30195
    },
    {
      "entropy": 5.4560401916503904,
      "epoch": 2.997221119491862,
      "grad_norm": 1.078125,
      "learning_rate": 0.000411378860995595,
      "loss": 5.1373,
      "mean_token_accuracy": 0.1783331662416458,
      "num_tokens": 66164351.0,
      "step": 30200
    },
    {
      "entropy": 5.374107837677002,
      "epoch": 2.9977173481540293,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004113506782652528,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18049195110797883,
      "num_tokens": 66176641.0,
      "step": 30205
    },
    {
      "entropy": 5.445544719696045,
      "epoch": 2.998213576816197,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004113224921537076,
      "loss": 5.0684,
      "mean_token_accuracy": 0.17354937493801117,
      "num_tokens": 66187822.0,
      "step": 30210
    },
    {
      "entropy": 5.330350065231324,
      "epoch": 2.9987098054783643,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041129430266165836,
      "loss": 4.9884,
      "mean_token_accuracy": 0.1776341661810875,
      "num_tokens": 66199636.0,
      "step": 30215
    },
    {
      "entropy": 5.371321344375611,
      "epoch": 2.999206034140532,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004112661097898041,
      "loss": 4.976,
      "mean_token_accuracy": 0.17818184345960617,
      "num_tokens": 66210649.0,
      "step": 30220
    },
    {
      "entropy": 5.332385110855102,
      "epoch": 2.9997022628026997,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041123791353884397,
      "loss": 4.9464,
      "mean_token_accuracy": 0.18333798050880432,
      "num_tokens": 66221806.0,
      "step": 30225
    },
    {
      "entropy": 5.270693731307984,
      "epoch": 3.000198491464867,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004112097139094771,
      "loss": 4.9181,
      "mean_token_accuracy": 0.1823803514242172,
      "num_tokens": 66232016.0,
      "step": 30230
    },
    {
      "entropy": 5.286133050918579,
      "epoch": 3.0006947201270346,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041118151090240284,
      "loss": 4.8708,
      "mean_token_accuracy": 0.18741313964128495,
      "num_tokens": 66242852.0,
      "step": 30235
    },
    {
      "entropy": 5.367388820648193,
      "epoch": 3.001190948789202,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041115330451832065,
      "loss": 4.9117,
      "mean_token_accuracy": 0.18387403935194016,
      "num_tokens": 66253254.0,
      "step": 30240
    },
    {
      "entropy": 5.342922067642212,
      "epoch": 3.0016871774513696,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041112509475792983,
      "loss": 4.8995,
      "mean_token_accuracy": 0.18251138031482697,
      "num_tokens": 66262721.0,
      "step": 30245
    },
    {
      "entropy": 5.265235519409179,
      "epoch": 3.002183406113537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041109688162193,
      "loss": 4.8619,
      "mean_token_accuracy": 0.18433444797992707,
      "num_tokens": 66273718.0,
      "step": 30250
    },
    {
      "entropy": 5.317084693908692,
      "epoch": 3.0026796347757045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004110686651110207,
      "loss": 4.9353,
      "mean_token_accuracy": 0.18033940047025682,
      "num_tokens": 66285527.0,
      "step": 30255
    },
    {
      "entropy": 5.343463611602783,
      "epoch": 3.003175863437872,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004110404452259017,
      "loss": 4.8922,
      "mean_token_accuracy": 0.18687526434659957,
      "num_tokens": 66295564.0,
      "step": 30260
    },
    {
      "entropy": 5.261044263839722,
      "epoch": 3.0036720921000395,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004110122219672729,
      "loss": 4.8547,
      "mean_token_accuracy": 0.19496615678071977,
      "num_tokens": 66306721.0,
      "step": 30265
    },
    {
      "entropy": 5.2801741600036625,
      "epoch": 3.0041683207622074,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000410983995335834,
      "loss": 4.8692,
      "mean_token_accuracy": 0.1948615476489067,
      "num_tokens": 66316990.0,
      "step": 30270
    },
    {
      "entropy": 5.295068407058716,
      "epoch": 3.004664549424375,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000410955765332285,
      "loss": 4.9031,
      "mean_token_accuracy": 0.19090454429388046,
      "num_tokens": 66327814.0,
      "step": 30275
    },
    {
      "entropy": 5.265821218490601,
      "epoch": 3.0051607780865424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041092753195732607,
      "loss": 4.8496,
      "mean_token_accuracy": 0.18605807721614837,
      "num_tokens": 66338373.0,
      "step": 30280
    },
    {
      "entropy": 5.304681491851807,
      "epoch": 3.00565700674871,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041089929521165725,
      "loss": 4.9265,
      "mean_token_accuracy": 0.18985942751169205,
      "num_tokens": 66349311.0,
      "step": 30285
    },
    {
      "entropy": 5.368531560897827,
      "epoch": 3.0061532354108773,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004108710550959787,
      "loss": 4.915,
      "mean_token_accuracy": 0.1896743580698967,
      "num_tokens": 66360771.0,
      "step": 30290
    },
    {
      "entropy": 5.336238813400269,
      "epoch": 3.006649464073045,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041084281161099073,
      "loss": 4.9074,
      "mean_token_accuracy": 0.1865750938653946,
      "num_tokens": 66371348.0,
      "step": 30295
    },
    {
      "entropy": 5.356964826583862,
      "epoch": 3.0071456927352123,
      "grad_norm": 1.0,
      "learning_rate": 0.0004108145647573938,
      "loss": 4.9587,
      "mean_token_accuracy": 0.18578348010778428,
      "num_tokens": 66382489.0,
      "step": 30300
    },
    {
      "entropy": 5.301280927658081,
      "epoch": 3.0076419213973797,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041078631453588834,
      "loss": 4.8238,
      "mean_token_accuracy": 0.19429711252450943,
      "num_tokens": 66392825.0,
      "step": 30305
    },
    {
      "entropy": 5.30101752281189,
      "epoch": 3.0081381500595477,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004107580609471748,
      "loss": 4.9049,
      "mean_token_accuracy": 0.1845463678240776,
      "num_tokens": 66403646.0,
      "step": 30310
    },
    {
      "entropy": 5.246071529388428,
      "epoch": 3.008634378721715,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041072980399195394,
      "loss": 4.8135,
      "mean_token_accuracy": 0.1976514309644699,
      "num_tokens": 66414199.0,
      "step": 30315
    },
    {
      "entropy": 5.26142725944519,
      "epoch": 3.0091306073838826,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004107015436709264,
      "loss": 4.8242,
      "mean_token_accuracy": 0.18893151432275773,
      "num_tokens": 66426331.0,
      "step": 30320
    },
    {
      "entropy": 5.259121656417847,
      "epoch": 3.00962683604605,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041067327998479296,
      "loss": 4.8266,
      "mean_token_accuracy": 0.1869245156645775,
      "num_tokens": 66436854.0,
      "step": 30325
    },
    {
      "entropy": 5.281003856658936,
      "epoch": 3.0101230647082176,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041064501293425453,
      "loss": 4.9338,
      "mean_token_accuracy": 0.18995184004306792,
      "num_tokens": 66447757.0,
      "step": 30330
    },
    {
      "entropy": 5.301751375198364,
      "epoch": 3.010619293370385,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004106167425200121,
      "loss": 4.8842,
      "mean_token_accuracy": 0.18637774884700775,
      "num_tokens": 66458395.0,
      "step": 30335
    },
    {
      "entropy": 5.256955242156982,
      "epoch": 3.0111155220325525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004105884687427667,
      "loss": 4.7697,
      "mean_token_accuracy": 0.19525453746318816,
      "num_tokens": 66467728.0,
      "step": 30340
    },
    {
      "entropy": 5.227864980697632,
      "epoch": 3.01161175069472,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004105601916032194,
      "loss": 4.8765,
      "mean_token_accuracy": 0.19574571698904036,
      "num_tokens": 66478113.0,
      "step": 30345
    },
    {
      "entropy": 5.284300470352173,
      "epoch": 3.0121079793568875,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004105319111020714,
      "loss": 4.8569,
      "mean_token_accuracy": 0.18784415423870088,
      "num_tokens": 66489055.0,
      "step": 30350
    },
    {
      "entropy": 5.290496492385865,
      "epoch": 3.0126042080190554,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041050362724002413,
      "loss": 4.861,
      "mean_token_accuracy": 0.19001045376062392,
      "num_tokens": 66499516.0,
      "step": 30355
    },
    {
      "entropy": 5.2940051555633545,
      "epoch": 3.013100436681223,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041047534001777884,
      "loss": 4.936,
      "mean_token_accuracy": 0.1869167149066925,
      "num_tokens": 66510832.0,
      "step": 30360
    },
    {
      "entropy": 5.36661467552185,
      "epoch": 3.0135966653433903,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004104470494360371,
      "loss": 4.9705,
      "mean_token_accuracy": 0.18099557608366013,
      "num_tokens": 66522457.0,
      "step": 30365
    },
    {
      "entropy": 5.336384057998657,
      "epoch": 3.014092894005558,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041041875549550043,
      "loss": 4.9369,
      "mean_token_accuracy": 0.18616467714309692,
      "num_tokens": 66532527.0,
      "step": 30370
    },
    {
      "entropy": 5.342232322692871,
      "epoch": 3.0145891226677253,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004103904581968703,
      "loss": 4.9075,
      "mean_token_accuracy": 0.18597796112298964,
      "num_tokens": 66542437.0,
      "step": 30375
    },
    {
      "entropy": 5.258966541290283,
      "epoch": 3.0150853513298927,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041036215754084864,
      "loss": 4.8566,
      "mean_token_accuracy": 0.19025434255599977,
      "num_tokens": 66552927.0,
      "step": 30380
    },
    {
      "entropy": 5.313636350631714,
      "epoch": 3.0155815799920602,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041033385352813714,
      "loss": 4.983,
      "mean_token_accuracy": 0.18103162944316864,
      "num_tokens": 66563664.0,
      "step": 30385
    },
    {
      "entropy": 5.323551607131958,
      "epoch": 3.0160778086542277,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004103055461594377,
      "loss": 4.886,
      "mean_token_accuracy": 0.1900268852710724,
      "num_tokens": 66573968.0,
      "step": 30390
    },
    {
      "entropy": 5.327582216262817,
      "epoch": 3.0165740373163956,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004102772354354522,
      "loss": 4.9719,
      "mean_token_accuracy": 0.1811283841729164,
      "num_tokens": 66585048.0,
      "step": 30395
    },
    {
      "entropy": 5.314001131057739,
      "epoch": 3.017070265978563,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041024892135688287,
      "loss": 4.9492,
      "mean_token_accuracy": 0.18292193859815598,
      "num_tokens": 66594813.0,
      "step": 30400
    },
    {
      "entropy": 5.367702198028565,
      "epoch": 3.0175664946407306,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004102206039244316,
      "loss": 4.9104,
      "mean_token_accuracy": 0.18650465607643127,
      "num_tokens": 66607096.0,
      "step": 30405
    },
    {
      "entropy": 5.3379401683807375,
      "epoch": 3.018062723302898,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004101922831388008,
      "loss": 4.911,
      "mean_token_accuracy": 0.17782786190509797,
      "num_tokens": 66617001.0,
      "step": 30410
    },
    {
      "entropy": 5.325883626937866,
      "epoch": 3.0185589519650655,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041016395900069267,
      "loss": 4.8972,
      "mean_token_accuracy": 0.18533137142658235,
      "num_tokens": 66627036.0,
      "step": 30415
    },
    {
      "entropy": 5.2928221225738525,
      "epoch": 3.019055180627233,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004101356315108097,
      "loss": 4.9144,
      "mean_token_accuracy": 0.18151905834674836,
      "num_tokens": 66637526.0,
      "step": 30420
    },
    {
      "entropy": 5.297168970108032,
      "epoch": 3.0195514092894005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041010730066985414,
      "loss": 4.8525,
      "mean_token_accuracy": 0.19439444839954376,
      "num_tokens": 66650046.0,
      "step": 30425
    },
    {
      "entropy": 5.247808742523193,
      "epoch": 3.020047637951568,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004100789664785287,
      "loss": 4.8008,
      "mean_token_accuracy": 0.1905344933271408,
      "num_tokens": 66660922.0,
      "step": 30430
    },
    {
      "entropy": 5.32895188331604,
      "epoch": 3.0205438666137354,
      "grad_norm": 1.125,
      "learning_rate": 0.0004100506289375359,
      "loss": 4.9633,
      "mean_token_accuracy": 0.18288267105817796,
      "num_tokens": 66671861.0,
      "step": 30435
    },
    {
      "entropy": 5.314322185516358,
      "epoch": 3.0210400952759033,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004100222880475786,
      "loss": 4.9032,
      "mean_token_accuracy": 0.18777859210968018,
      "num_tokens": 66682834.0,
      "step": 30440
    },
    {
      "entropy": 5.285633373260498,
      "epoch": 3.021536323938071,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040999394380935943,
      "loss": 4.9102,
      "mean_token_accuracy": 0.188670052587986,
      "num_tokens": 66693692.0,
      "step": 30445
    },
    {
      "entropy": 5.283768892288208,
      "epoch": 3.0220325526002383,
      "grad_norm": 1.0,
      "learning_rate": 0.00040996559622358145,
      "loss": 4.8998,
      "mean_token_accuracy": 0.18967295438051224,
      "num_tokens": 66705035.0,
      "step": 30450
    },
    {
      "entropy": 5.229093408584594,
      "epoch": 3.0225287812624058,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004099372452909474,
      "loss": 4.8139,
      "mean_token_accuracy": 0.1976830393075943,
      "num_tokens": 66714713.0,
      "step": 30455
    },
    {
      "entropy": 5.257751750946045,
      "epoch": 3.0230250099245732,
      "grad_norm": 1.0,
      "learning_rate": 0.0004099088910121605,
      "loss": 4.9476,
      "mean_token_accuracy": 0.181852488219738,
      "num_tokens": 66726454.0,
      "step": 30460
    },
    {
      "entropy": 5.285331106185913,
      "epoch": 3.0235212385867407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040988053338792367,
      "loss": 4.8832,
      "mean_token_accuracy": 0.1882677525281906,
      "num_tokens": 66736103.0,
      "step": 30465
    },
    {
      "entropy": 5.292697334289551,
      "epoch": 3.024017467248908,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004098521724189404,
      "loss": 4.9412,
      "mean_token_accuracy": 0.17969736009836196,
      "num_tokens": 66746554.0,
      "step": 30470
    },
    {
      "entropy": 5.346834182739258,
      "epoch": 3.0245136959110757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040982380810591386,
      "loss": 4.8678,
      "mean_token_accuracy": 0.19035170525312423,
      "num_tokens": 66756568.0,
      "step": 30475
    },
    {
      "entropy": 5.288014221191406,
      "epoch": 3.025009924573243,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004097954404495473,
      "loss": 4.8751,
      "mean_token_accuracy": 0.18766970485448836,
      "num_tokens": 66769036.0,
      "step": 30480
    },
    {
      "entropy": 5.254004001617432,
      "epoch": 3.025506153235411,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040976706945054434,
      "loss": 4.7937,
      "mean_token_accuracy": 0.20542415082454682,
      "num_tokens": 66779868.0,
      "step": 30485
    },
    {
      "entropy": 5.3002612590789795,
      "epoch": 3.0260023818975785,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040973869510960837,
      "loss": 4.8975,
      "mean_token_accuracy": 0.18165594041347505,
      "num_tokens": 66791903.0,
      "step": 30490
    },
    {
      "entropy": 5.313199758529663,
      "epoch": 3.026498610559746,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040971031742744314,
      "loss": 4.9355,
      "mean_token_accuracy": 0.1816021516919136,
      "num_tokens": 66802118.0,
      "step": 30495
    },
    {
      "entropy": 5.353033590316772,
      "epoch": 3.0269948392219135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040968193640475236,
      "loss": 4.9381,
      "mean_token_accuracy": 0.18719416856765747,
      "num_tokens": 66813653.0,
      "step": 30500
    },
    {
      "entropy": 5.419765186309815,
      "epoch": 3.027491067884081,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004096535520422397,
      "loss": 4.988,
      "mean_token_accuracy": 0.1760167211294174,
      "num_tokens": 66824752.0,
      "step": 30505
    },
    {
      "entropy": 5.3168439865112305,
      "epoch": 3.0279872965462484,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040962516434060915,
      "loss": 4.8562,
      "mean_token_accuracy": 0.19456040412187575,
      "num_tokens": 66835483.0,
      "step": 30510
    },
    {
      "entropy": 5.323139619827271,
      "epoch": 3.028483525208416,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004095967733005646,
      "loss": 4.9346,
      "mean_token_accuracy": 0.18285020887851716,
      "num_tokens": 66846890.0,
      "step": 30515
    },
    {
      "entropy": 5.213891315460205,
      "epoch": 3.0289797538705834,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040956837892281014,
      "loss": 4.8176,
      "mean_token_accuracy": 0.1923963189125061,
      "num_tokens": 66857326.0,
      "step": 30520
    },
    {
      "entropy": 5.328210544586182,
      "epoch": 3.0294759825327513,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040953998120804976,
      "loss": 4.8933,
      "mean_token_accuracy": 0.18658428639173508,
      "num_tokens": 66868266.0,
      "step": 30525
    },
    {
      "entropy": 5.356304025650024,
      "epoch": 3.0299722111949188,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004095115801569878,
      "loss": 4.9325,
      "mean_token_accuracy": 0.18099202662706376,
      "num_tokens": 66879121.0,
      "step": 30530
    },
    {
      "entropy": 5.295130872726441,
      "epoch": 3.0304684398570862,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004094831757703285,
      "loss": 4.8706,
      "mean_token_accuracy": 0.18671713322401046,
      "num_tokens": 66889804.0,
      "step": 30535
    },
    {
      "entropy": 5.274500179290771,
      "epoch": 3.0309646685192537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040945476804877623,
      "loss": 4.8941,
      "mean_token_accuracy": 0.1892985001206398,
      "num_tokens": 66900421.0,
      "step": 30540
    },
    {
      "entropy": 5.341641521453857,
      "epoch": 3.031460897181421,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004094263569930355,
      "loss": 4.9687,
      "mean_token_accuracy": 0.1791709139943123,
      "num_tokens": 66910754.0,
      "step": 30545
    },
    {
      "entropy": 5.427459001541138,
      "epoch": 3.0319571258435887,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040939794260381074,
      "loss": 4.9521,
      "mean_token_accuracy": 0.183514966070652,
      "num_tokens": 66921107.0,
      "step": 30550
    },
    {
      "entropy": 5.2726442337036135,
      "epoch": 3.032453354505756,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004093695248818065,
      "loss": 4.8682,
      "mean_token_accuracy": 0.18951306045055388,
      "num_tokens": 66932851.0,
      "step": 30555
    },
    {
      "entropy": 5.323325061798096,
      "epoch": 3.0329495831679236,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004093411038277277,
      "loss": 4.9239,
      "mean_token_accuracy": 0.17990732938051224,
      "num_tokens": 66943403.0,
      "step": 30560
    },
    {
      "entropy": 5.391712093353272,
      "epoch": 3.033445811830091,
      "grad_norm": 1.03125,
      "learning_rate": 0.000409312679442279,
      "loss": 5.0381,
      "mean_token_accuracy": 0.18226863145828248,
      "num_tokens": 66955021.0,
      "step": 30565
    },
    {
      "entropy": 5.353145360946655,
      "epoch": 3.033942040492259,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040928425172616524,
      "loss": 4.9215,
      "mean_token_accuracy": 0.18425218015909195,
      "num_tokens": 66966147.0,
      "step": 30570
    },
    {
      "entropy": 5.313010549545288,
      "epoch": 3.0344382691544265,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040925582068009145,
      "loss": 4.8458,
      "mean_token_accuracy": 0.186372072994709,
      "num_tokens": 66977946.0,
      "step": 30575
    },
    {
      "entropy": 5.343583297729492,
      "epoch": 3.034934497816594,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004092273863047625,
      "loss": 4.9664,
      "mean_token_accuracy": 0.18595708906650543,
      "num_tokens": 66988014.0,
      "step": 30580
    },
    {
      "entropy": 5.352772808074951,
      "epoch": 3.0354307264787614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040919894860088366,
      "loss": 4.9189,
      "mean_token_accuracy": 0.18408778607845305,
      "num_tokens": 66999142.0,
      "step": 30585
    },
    {
      "entropy": 5.259269142150879,
      "epoch": 3.035926955140929,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004091705075691601,
      "loss": 4.9605,
      "mean_token_accuracy": 0.17683627605438232,
      "num_tokens": 67010849.0,
      "step": 30590
    },
    {
      "entropy": 5.289477062225342,
      "epoch": 3.0364231838030964,
      "grad_norm": 1.0625,
      "learning_rate": 0.000409142063210297,
      "loss": 4.8583,
      "mean_token_accuracy": 0.19421707838773727,
      "num_tokens": 67021681.0,
      "step": 30595
    },
    {
      "entropy": 5.338026666641236,
      "epoch": 3.036919412465264,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040911361552499974,
      "loss": 4.9497,
      "mean_token_accuracy": 0.17591291964054107,
      "num_tokens": 67034080.0,
      "step": 30600
    },
    {
      "entropy": 5.320521879196167,
      "epoch": 3.0374156411274313,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040908516451397395,
      "loss": 4.8342,
      "mean_token_accuracy": 0.18653670996427535,
      "num_tokens": 67045528.0,
      "step": 30605
    },
    {
      "entropy": 5.302772283554077,
      "epoch": 3.0379118697895993,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004090567101779249,
      "loss": 4.8649,
      "mean_token_accuracy": 0.18709441125392914,
      "num_tokens": 67057611.0,
      "step": 30610
    },
    {
      "entropy": 5.261604070663452,
      "epoch": 3.0384080984517667,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004090282525175583,
      "loss": 4.8891,
      "mean_token_accuracy": 0.18792577534914018,
      "num_tokens": 67067843.0,
      "step": 30615
    },
    {
      "entropy": 5.348330736160278,
      "epoch": 3.038904327113934,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004089997915335799,
      "loss": 4.9831,
      "mean_token_accuracy": 0.18639727383852006,
      "num_tokens": 67079125.0,
      "step": 30620
    },
    {
      "entropy": 5.298739719390869,
      "epoch": 3.0394005557761017,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004089713272266954,
      "loss": 4.9028,
      "mean_token_accuracy": 0.18508587926626205,
      "num_tokens": 67089750.0,
      "step": 30625
    },
    {
      "entropy": 5.335669755935669,
      "epoch": 3.039896784438269,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004089428595976106,
      "loss": 4.955,
      "mean_token_accuracy": 0.18565613627433777,
      "num_tokens": 67101738.0,
      "step": 30630
    },
    {
      "entropy": 5.292442178726196,
      "epoch": 3.0403930131004366,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004089143886470316,
      "loss": 4.854,
      "mean_token_accuracy": 0.19091516882181167,
      "num_tokens": 67111941.0,
      "step": 30635
    },
    {
      "entropy": 5.3361071109771725,
      "epoch": 3.040889241762604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004088859143756643,
      "loss": 5.0033,
      "mean_token_accuracy": 0.17981231808662415,
      "num_tokens": 67124755.0,
      "step": 30640
    },
    {
      "entropy": 5.39765772819519,
      "epoch": 3.0413854704247716,
      "grad_norm": 1.125,
      "learning_rate": 0.0004088574367842148,
      "loss": 5.0816,
      "mean_token_accuracy": 0.18164192140102386,
      "num_tokens": 67136597.0,
      "step": 30645
    },
    {
      "entropy": 5.368074655532837,
      "epoch": 3.041881699086939,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004088289558733893,
      "loss": 4.9515,
      "mean_token_accuracy": 0.19030878841876983,
      "num_tokens": 67146790.0,
      "step": 30650
    },
    {
      "entropy": 5.347726106643677,
      "epoch": 3.042377927749107,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004088004716438941,
      "loss": 4.926,
      "mean_token_accuracy": 0.19198968261480331,
      "num_tokens": 67158054.0,
      "step": 30655
    },
    {
      "entropy": 5.3011644840240475,
      "epoch": 3.0428741564112745,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004087719840964355,
      "loss": 4.842,
      "mean_token_accuracy": 0.19172983020544052,
      "num_tokens": 67168244.0,
      "step": 30660
    },
    {
      "entropy": 5.283528423309326,
      "epoch": 3.043370385073442,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040874349323171994,
      "loss": 4.8704,
      "mean_token_accuracy": 0.1906850144267082,
      "num_tokens": 67179165.0,
      "step": 30665
    },
    {
      "entropy": 5.25514235496521,
      "epoch": 3.0438666137356094,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004087149990504539,
      "loss": 4.8719,
      "mean_token_accuracy": 0.190183562040329,
      "num_tokens": 67189653.0,
      "step": 30670
    },
    {
      "entropy": 5.29747428894043,
      "epoch": 3.044362842397777,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040868650155334406,
      "loss": 4.9475,
      "mean_token_accuracy": 0.1796001374721527,
      "num_tokens": 67203933.0,
      "step": 30675
    },
    {
      "entropy": 5.24266242980957,
      "epoch": 3.0448590710599444,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040865800074109713,
      "loss": 4.8587,
      "mean_token_accuracy": 0.18902930468320847,
      "num_tokens": 67215590.0,
      "step": 30680
    },
    {
      "entropy": 5.300674057006836,
      "epoch": 3.045355299722112,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004086294966144196,
      "loss": 4.8537,
      "mean_token_accuracy": 0.18947855979204178,
      "num_tokens": 67225416.0,
      "step": 30685
    },
    {
      "entropy": 5.2821897029876705,
      "epoch": 3.0458515283842793,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004086009891740187,
      "loss": 4.8165,
      "mean_token_accuracy": 0.1889408975839615,
      "num_tokens": 67234492.0,
      "step": 30690
    },
    {
      "entropy": 5.257295417785644,
      "epoch": 3.046347757046447,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040857247842060105,
      "loss": 4.9,
      "mean_token_accuracy": 0.18166453689336776,
      "num_tokens": 67245630.0,
      "step": 30695
    },
    {
      "entropy": 5.343411159515381,
      "epoch": 3.0468439857086147,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004085439643548737,
      "loss": 4.9897,
      "mean_token_accuracy": 0.1834544286131859,
      "num_tokens": 67257014.0,
      "step": 30700
    },
    {
      "entropy": 5.2039693832397464,
      "epoch": 3.047340214370782,
      "grad_norm": 1.0,
      "learning_rate": 0.0004085154469775439,
      "loss": 4.7817,
      "mean_token_accuracy": 0.19578063040971755,
      "num_tokens": 67267165.0,
      "step": 30705
    },
    {
      "entropy": 5.305708980560302,
      "epoch": 3.0478364430329496,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004084869262893187,
      "loss": 4.8925,
      "mean_token_accuracy": 0.18981412351131438,
      "num_tokens": 67277077.0,
      "step": 30710
    },
    {
      "entropy": 5.2969330787658695,
      "epoch": 3.048332671695117,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004084584022909053,
      "loss": 4.9169,
      "mean_token_accuracy": 0.1865010753273964,
      "num_tokens": 67287941.0,
      "step": 30715
    },
    {
      "entropy": 5.180174016952515,
      "epoch": 3.0488289003572846,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040842987498301115,
      "loss": 4.7968,
      "mean_token_accuracy": 0.18959407806396483,
      "num_tokens": 67299764.0,
      "step": 30720
    },
    {
      "entropy": 5.2974567890167235,
      "epoch": 3.049325129019452,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004084013443663437,
      "loss": 4.8696,
      "mean_token_accuracy": 0.18633117377758027,
      "num_tokens": 67310527.0,
      "step": 30725
    },
    {
      "entropy": 5.331798076629639,
      "epoch": 3.0498213576816195,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004083728104416102,
      "loss": 4.8991,
      "mean_token_accuracy": 0.18736923038959502,
      "num_tokens": 67321369.0,
      "step": 30730
    },
    {
      "entropy": 5.258451652526856,
      "epoch": 3.050317586343787,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004083442732095185,
      "loss": 4.8365,
      "mean_token_accuracy": 0.19927524626255036,
      "num_tokens": 67332215.0,
      "step": 30735
    },
    {
      "entropy": 5.32018551826477,
      "epoch": 3.050813815005955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004083157326707762,
      "loss": 5.0207,
      "mean_token_accuracy": 0.17866481393575667,
      "num_tokens": 67344407.0,
      "step": 30740
    },
    {
      "entropy": 5.324433088302612,
      "epoch": 3.0513100436681224,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040828718882609085,
      "loss": 4.8737,
      "mean_token_accuracy": 0.18492981791496277,
      "num_tokens": 67355513.0,
      "step": 30745
    },
    {
      "entropy": 5.281755065917968,
      "epoch": 3.05180627233029,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040825864167617047,
      "loss": 4.9151,
      "mean_token_accuracy": 0.18315714448690415,
      "num_tokens": 67366515.0,
      "step": 30750
    },
    {
      "entropy": 5.3294209957122805,
      "epoch": 3.0523025009924574,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000408230091221723,
      "loss": 4.9108,
      "mean_token_accuracy": 0.1871731087565422,
      "num_tokens": 67376568.0,
      "step": 30755
    },
    {
      "entropy": 5.3504602909088135,
      "epoch": 3.052798729654625,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004082015374634563,
      "loss": 4.9701,
      "mean_token_accuracy": 0.17987498193979262,
      "num_tokens": 67388639.0,
      "step": 30760
    },
    {
      "entropy": 5.327446985244751,
      "epoch": 3.0532949583167923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004081729804020785,
      "loss": 4.9071,
      "mean_token_accuracy": 0.17831323444843292,
      "num_tokens": 67400330.0,
      "step": 30765
    },
    {
      "entropy": 5.281221389770508,
      "epoch": 3.05379118697896,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004081444200382978,
      "loss": 4.8652,
      "mean_token_accuracy": 0.18893893361091613,
      "num_tokens": 67411067.0,
      "step": 30770
    },
    {
      "entropy": 5.3120321273803714,
      "epoch": 3.0542874156411273,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040811585637282234,
      "loss": 4.8497,
      "mean_token_accuracy": 0.18820465505123138,
      "num_tokens": 67421142.0,
      "step": 30775
    },
    {
      "entropy": 5.253071355819702,
      "epoch": 3.0547836443032947,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004080872894063605,
      "loss": 4.8503,
      "mean_token_accuracy": 0.19042923897504807,
      "num_tokens": 67433609.0,
      "step": 30780
    },
    {
      "entropy": 5.405457067489624,
      "epoch": 3.0552798729654627,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040805871913962056,
      "loss": 5.0535,
      "mean_token_accuracy": 0.16958091408014297,
      "num_tokens": 67445281.0,
      "step": 30785
    },
    {
      "entropy": 5.261765718460083,
      "epoch": 3.05577610162763,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004080301455733112,
      "loss": 4.8717,
      "mean_token_accuracy": 0.18146812915802002,
      "num_tokens": 67455831.0,
      "step": 30790
    },
    {
      "entropy": 5.311526870727539,
      "epoch": 3.0562723302897976,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00040800156870814087,
      "loss": 4.9734,
      "mean_token_accuracy": 0.18589683026075363,
      "num_tokens": 67467195.0,
      "step": 30795
    },
    {
      "entropy": 5.256455326080323,
      "epoch": 3.056768558951965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040797298854481826,
      "loss": 4.8502,
      "mean_token_accuracy": 0.18842951506376265,
      "num_tokens": 67478332.0,
      "step": 30800
    },
    {
      "entropy": 5.281278085708618,
      "epoch": 3.0572647876141326,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040794440508405195,
      "loss": 4.8835,
      "mean_token_accuracy": 0.18999885767698288,
      "num_tokens": 67489600.0,
      "step": 30805
    },
    {
      "entropy": 5.36547269821167,
      "epoch": 3.0577610162763,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000407915818326551,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18032452911138536,
      "num_tokens": 67498837.0,
      "step": 30810
    },
    {
      "entropy": 5.407568740844726,
      "epoch": 3.0582572449384675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000407887228273024,
      "loss": 4.9489,
      "mean_token_accuracy": 0.18314982801675797,
      "num_tokens": 67509118.0,
      "step": 30815
    },
    {
      "entropy": 5.271299839019775,
      "epoch": 3.058753473600635,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004078586349241802,
      "loss": 4.8492,
      "mean_token_accuracy": 0.1864874079823494,
      "num_tokens": 67520577.0,
      "step": 30820
    },
    {
      "entropy": 5.302979373931885,
      "epoch": 3.059249702262803,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040783003828072855,
      "loss": 4.8493,
      "mean_token_accuracy": 0.18635737895965576,
      "num_tokens": 67532020.0,
      "step": 30825
    },
    {
      "entropy": 5.271734189987183,
      "epoch": 3.0597459309249704,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004078014383433782,
      "loss": 4.8727,
      "mean_token_accuracy": 0.18828414082527162,
      "num_tokens": 67540845.0,
      "step": 30830
    },
    {
      "entropy": 5.252586078643799,
      "epoch": 3.060242159587138,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004077728351128383,
      "loss": 4.8945,
      "mean_token_accuracy": 0.18870023936033248,
      "num_tokens": 67552698.0,
      "step": 30835
    },
    {
      "entropy": 5.2014542579650875,
      "epoch": 3.0607383882493053,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004077442285898181,
      "loss": 4.8656,
      "mean_token_accuracy": 0.18511125147342683,
      "num_tokens": 67563772.0,
      "step": 30840
    },
    {
      "entropy": 5.349909019470215,
      "epoch": 3.061234616911473,
      "grad_norm": 0.890625,
      "learning_rate": 0.00040771561877502727,
      "loss": 4.9076,
      "mean_token_accuracy": 0.18483437597751617,
      "num_tokens": 67574955.0,
      "step": 30845
    },
    {
      "entropy": 5.317493915557861,
      "epoch": 3.0617308455736403,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004076870056691749,
      "loss": 4.8591,
      "mean_token_accuracy": 0.1907067432999611,
      "num_tokens": 67586086.0,
      "step": 30850
    },
    {
      "entropy": 5.313659763336181,
      "epoch": 3.0622270742358078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040765838927297077,
      "loss": 4.9311,
      "mean_token_accuracy": 0.1907110959291458,
      "num_tokens": 67597036.0,
      "step": 30855
    },
    {
      "entropy": 5.345471906661987,
      "epoch": 3.0627233028979752,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004076297695871244,
      "loss": 4.8146,
      "mean_token_accuracy": 0.20624638348817825,
      "num_tokens": 67606499.0,
      "step": 30860
    },
    {
      "entropy": 5.3490289688110355,
      "epoch": 3.0632195315601427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040760114661234553,
      "loss": 4.9499,
      "mean_token_accuracy": 0.18539584577083587,
      "num_tokens": 67617820.0,
      "step": 30865
    },
    {
      "entropy": 5.291796875,
      "epoch": 3.0637157602223106,
      "grad_norm": 1.015625,
      "learning_rate": 0.000407572520349344,
      "loss": 4.8191,
      "mean_token_accuracy": 0.1927574247121811,
      "num_tokens": 67628122.0,
      "step": 30870
    },
    {
      "entropy": 5.243136358261109,
      "epoch": 3.064211988884478,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004075438907988296,
      "loss": 4.9115,
      "mean_token_accuracy": 0.18780407011508943,
      "num_tokens": 67639127.0,
      "step": 30875
    },
    {
      "entropy": 5.281428432464599,
      "epoch": 3.0647082175466456,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004075152579615124,
      "loss": 4.8093,
      "mean_token_accuracy": 0.197114759683609,
      "num_tokens": 67650122.0,
      "step": 30880
    },
    {
      "entropy": 5.320765018463135,
      "epoch": 3.065204446208813,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004074866218381023,
      "loss": 4.9306,
      "mean_token_accuracy": 0.17893226593732833,
      "num_tokens": 67660427.0,
      "step": 30885
    },
    {
      "entropy": 5.2808177947998045,
      "epoch": 3.0657006748709805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040745798242930947,
      "loss": 5.0002,
      "mean_token_accuracy": 0.184769469499588,
      "num_tokens": 67671151.0,
      "step": 30890
    },
    {
      "entropy": 5.36487135887146,
      "epoch": 3.066196903533148,
      "grad_norm": 0.96875,
      "learning_rate": 0.000407429339735844,
      "loss": 4.9259,
      "mean_token_accuracy": 0.17924682199954986,
      "num_tokens": 67681338.0,
      "step": 30895
    },
    {
      "entropy": 5.2613850116729735,
      "epoch": 3.0666931321953155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040740069375841637,
      "loss": 4.8386,
      "mean_token_accuracy": 0.18892231583595276,
      "num_tokens": 67691884.0,
      "step": 30900
    },
    {
      "entropy": 5.250843620300293,
      "epoch": 3.067189360857483,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004073720444977368,
      "loss": 4.9056,
      "mean_token_accuracy": 0.18700968772172927,
      "num_tokens": 67703011.0,
      "step": 30905
    },
    {
      "entropy": 5.413768720626831,
      "epoch": 3.067685589519651,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004073433919545157,
      "loss": 5.0693,
      "mean_token_accuracy": 0.17891449183225633,
      "num_tokens": 67714042.0,
      "step": 30910
    },
    {
      "entropy": 5.341200590133667,
      "epoch": 3.0681818181818183,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004073147361294638,
      "loss": 4.9365,
      "mean_token_accuracy": 0.1857965812087059,
      "num_tokens": 67725449.0,
      "step": 30915
    },
    {
      "entropy": 5.366840839385986,
      "epoch": 3.068678046843986,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004072860770232915,
      "loss": 4.9932,
      "mean_token_accuracy": 0.18376203775405883,
      "num_tokens": 67737334.0,
      "step": 30920
    },
    {
      "entropy": 5.288955783843994,
      "epoch": 3.0691742755061533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004072574146367095,
      "loss": 4.8879,
      "mean_token_accuracy": 0.1873987451195717,
      "num_tokens": 67747777.0,
      "step": 30925
    },
    {
      "entropy": 5.291474008560181,
      "epoch": 3.0696705041683208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040722874897042863,
      "loss": 4.8565,
      "mean_token_accuracy": 0.19431440234184266,
      "num_tokens": 67758584.0,
      "step": 30930
    },
    {
      "entropy": 5.262974119186401,
      "epoch": 3.0701667328304882,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004072000800251597,
      "loss": 4.8757,
      "mean_token_accuracy": 0.19069294780492782,
      "num_tokens": 67769168.0,
      "step": 30935
    },
    {
      "entropy": 5.283299112319947,
      "epoch": 3.0706629614926557,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040717140780161367,
      "loss": 4.8891,
      "mean_token_accuracy": 0.19419799596071244,
      "num_tokens": 67779680.0,
      "step": 30940
    },
    {
      "entropy": 5.290833950042725,
      "epoch": 3.071159190154823,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040714273230050153,
      "loss": 4.8965,
      "mean_token_accuracy": 0.17966223657131195,
      "num_tokens": 67789678.0,
      "step": 30945
    },
    {
      "entropy": 5.315385866165161,
      "epoch": 3.0716554188169907,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004071140535225343,
      "loss": 4.9482,
      "mean_token_accuracy": 0.19023759216070174,
      "num_tokens": 67800991.0,
      "step": 30950
    },
    {
      "entropy": 5.400873756408691,
      "epoch": 3.0721516474791586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004070853714684233,
      "loss": 5.0232,
      "mean_token_accuracy": 0.17621572464704513,
      "num_tokens": 67811319.0,
      "step": 30955
    },
    {
      "entropy": 5.402422618865967,
      "epoch": 3.072647876141326,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004070566861388797,
      "loss": 5.0068,
      "mean_token_accuracy": 0.18513171076774598,
      "num_tokens": 67823704.0,
      "step": 30960
    },
    {
      "entropy": 5.34969687461853,
      "epoch": 3.0731441048034935,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040702799753461477,
      "loss": 4.978,
      "mean_token_accuracy": 0.17857130914926528,
      "num_tokens": 67835460.0,
      "step": 30965
    },
    {
      "entropy": 5.275319910049438,
      "epoch": 3.073640333465661,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004069993056563401,
      "loss": 4.8636,
      "mean_token_accuracy": 0.19124802052974701,
      "num_tokens": 67845464.0,
      "step": 30970
    },
    {
      "entropy": 5.433015203475952,
      "epoch": 3.0741365621278285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000406970610504767,
      "loss": 4.9838,
      "mean_token_accuracy": 0.17452260702848435,
      "num_tokens": 67856718.0,
      "step": 30975
    },
    {
      "entropy": 5.298393869400025,
      "epoch": 3.074632790789996,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004069419120806072,
      "loss": 4.8892,
      "mean_token_accuracy": 0.18670626133680343,
      "num_tokens": 67867066.0,
      "step": 30980
    },
    {
      "entropy": 5.289642953872681,
      "epoch": 3.0751290194521634,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040691321038457224,
      "loss": 4.9343,
      "mean_token_accuracy": 0.184998582303524,
      "num_tokens": 67878213.0,
      "step": 30985
    },
    {
      "entropy": 5.336636066436768,
      "epoch": 3.075625248114331,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004068845054173739,
      "loss": 4.937,
      "mean_token_accuracy": 0.18188368529081345,
      "num_tokens": 67888503.0,
      "step": 30990
    },
    {
      "entropy": 5.376238918304443,
      "epoch": 3.076121476776499,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004068557971797241,
      "loss": 4.9193,
      "mean_token_accuracy": 0.17876862436532975,
      "num_tokens": 67899035.0,
      "step": 30995
    },
    {
      "entropy": 5.228286790847778,
      "epoch": 3.0766177054386663,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040682708567233464,
      "loss": 4.8786,
      "mean_token_accuracy": 0.19142864346504213,
      "num_tokens": 67909137.0,
      "step": 31000
    },
    {
      "entropy": 5.323409461975098,
      "epoch": 3.077113934100834,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040679837089591743,
      "loss": 4.9276,
      "mean_token_accuracy": 0.18037762939929963,
      "num_tokens": 67919026.0,
      "step": 31005
    },
    {
      "entropy": 5.4073278427124025,
      "epoch": 3.0776101627630013,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040676965285118465,
      "loss": 5.0047,
      "mean_token_accuracy": 0.18411382883787156,
      "num_tokens": 67930333.0,
      "step": 31010
    },
    {
      "entropy": 5.308143472671508,
      "epoch": 3.0781063914251687,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004067409315388484,
      "loss": 4.8385,
      "mean_token_accuracy": 0.19369235038757324,
      "num_tokens": 67940448.0,
      "step": 31015
    },
    {
      "entropy": 5.285804796218872,
      "epoch": 3.078602620087336,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040671220695962104,
      "loss": 4.8829,
      "mean_token_accuracy": 0.18887413293123245,
      "num_tokens": 67951398.0,
      "step": 31020
    },
    {
      "entropy": 5.321771812438965,
      "epoch": 3.0790988487495037,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004066834791142146,
      "loss": 4.9289,
      "mean_token_accuracy": 0.18370055109262468,
      "num_tokens": 67962741.0,
      "step": 31025
    },
    {
      "entropy": 5.3339310646057125,
      "epoch": 3.079595077411671,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004066547480033417,
      "loss": 4.9084,
      "mean_token_accuracy": 0.18845605701208115,
      "num_tokens": 67975015.0,
      "step": 31030
    },
    {
      "entropy": 5.364108276367188,
      "epoch": 3.0800913060738386,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004066260136277148,
      "loss": 4.9776,
      "mean_token_accuracy": 0.1770806059241295,
      "num_tokens": 67985918.0,
      "step": 31035
    },
    {
      "entropy": 5.2725893497467045,
      "epoch": 3.0805875347360065,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040659727598804635,
      "loss": 4.9282,
      "mean_token_accuracy": 0.1856217175722122,
      "num_tokens": 67997419.0,
      "step": 31040
    },
    {
      "entropy": 5.284546995162964,
      "epoch": 3.081083763398174,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040656853508504897,
      "loss": 4.8859,
      "mean_token_accuracy": 0.1909284308552742,
      "num_tokens": 68008739.0,
      "step": 31045
    },
    {
      "entropy": 5.319670486450195,
      "epoch": 3.0815799920603415,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040653979091943555,
      "loss": 4.9271,
      "mean_token_accuracy": 0.18315291851758958,
      "num_tokens": 68020707.0,
      "step": 31050
    },
    {
      "entropy": 5.392688989639282,
      "epoch": 3.082076220722509,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040651104349191864,
      "loss": 4.8503,
      "mean_token_accuracy": 0.19311174005270004,
      "num_tokens": 68030825.0,
      "step": 31055
    },
    {
      "entropy": 5.261142349243164,
      "epoch": 3.0825724493846764,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040648229280321126,
      "loss": 4.911,
      "mean_token_accuracy": 0.18748724907636644,
      "num_tokens": 68041385.0,
      "step": 31060
    },
    {
      "entropy": 5.328071784973145,
      "epoch": 3.083068678046844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004064535388540264,
      "loss": 5.0006,
      "mean_token_accuracy": 0.18405609726905822,
      "num_tokens": 68053123.0,
      "step": 31065
    },
    {
      "entropy": 5.29747371673584,
      "epoch": 3.0835649067090114,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004064247816450769,
      "loss": 4.8831,
      "mean_token_accuracy": 0.18434059619903564,
      "num_tokens": 68064529.0,
      "step": 31070
    },
    {
      "entropy": 5.4389561176300045,
      "epoch": 3.084061135371179,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004063960211770761,
      "loss": 4.9846,
      "mean_token_accuracy": 0.18496274948120117,
      "num_tokens": 68074091.0,
      "step": 31075
    },
    {
      "entropy": 5.319818353652954,
      "epoch": 3.084557364033347,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040636725745073714,
      "loss": 4.9038,
      "mean_token_accuracy": 0.18909524977207184,
      "num_tokens": 68085185.0,
      "step": 31080
    },
    {
      "entropy": 5.246987009048462,
      "epoch": 3.0850535926955143,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004063384904667732,
      "loss": 4.8086,
      "mean_token_accuracy": 0.1928607240319252,
      "num_tokens": 68095977.0,
      "step": 31085
    },
    {
      "entropy": 5.265205717086792,
      "epoch": 3.0855498213576817,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040630972022589774,
      "loss": 4.9305,
      "mean_token_accuracy": 0.18369110673666,
      "num_tokens": 68107469.0,
      "step": 31090
    },
    {
      "entropy": 5.423060035705566,
      "epoch": 3.086046050019849,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040628094672882415,
      "loss": 5.078,
      "mean_token_accuracy": 0.1682389035820961,
      "num_tokens": 68119153.0,
      "step": 31095
    },
    {
      "entropy": 5.31736192703247,
      "epoch": 3.0865422786820167,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000406252169976266,
      "loss": 4.871,
      "mean_token_accuracy": 0.18533989936113357,
      "num_tokens": 68130336.0,
      "step": 31100
    },
    {
      "entropy": 5.369941711425781,
      "epoch": 3.087038507344184,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004062233899689368,
      "loss": 4.8945,
      "mean_token_accuracy": 0.1844942107796669,
      "num_tokens": 68141231.0,
      "step": 31105
    },
    {
      "entropy": 5.274157428741455,
      "epoch": 3.0875347360063516,
      "grad_norm": 1.125,
      "learning_rate": 0.0004061946067075503,
      "loss": 4.8464,
      "mean_token_accuracy": 0.19641185849905013,
      "num_tokens": 68151891.0,
      "step": 31110
    },
    {
      "entropy": 5.275098896026611,
      "epoch": 3.088030964668519,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004061658201928203,
      "loss": 4.8791,
      "mean_token_accuracy": 0.18977589905261993,
      "num_tokens": 68162537.0,
      "step": 31115
    },
    {
      "entropy": 5.3796590805053714,
      "epoch": 3.0885271933306866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004061370304254606,
      "loss": 5.0737,
      "mean_token_accuracy": 0.1782889947295189,
      "num_tokens": 68173393.0,
      "step": 31120
    },
    {
      "entropy": 5.331187438964844,
      "epoch": 3.0890234219928545,
      "grad_norm": 1.03125,
      "learning_rate": 0.000406108237406185,
      "loss": 4.9419,
      "mean_token_accuracy": 0.18552072197198868,
      "num_tokens": 68184937.0,
      "step": 31125
    },
    {
      "entropy": 5.325215864181518,
      "epoch": 3.089519650655022,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040607944113570777,
      "loss": 4.9585,
      "mean_token_accuracy": 0.18821211755275727,
      "num_tokens": 68196962.0,
      "step": 31130
    },
    {
      "entropy": 5.3650373935699465,
      "epoch": 3.0900158793171895,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004060506416147428,
      "loss": 4.9391,
      "mean_token_accuracy": 0.18435475677251817,
      "num_tokens": 68208603.0,
      "step": 31135
    },
    {
      "entropy": 5.30541033744812,
      "epoch": 3.090512107979357,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004060218388440043,
      "loss": 4.9216,
      "mean_token_accuracy": 0.19023598432540895,
      "num_tokens": 68219424.0,
      "step": 31140
    },
    {
      "entropy": 5.288882160186768,
      "epoch": 3.0910083366415244,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004059930328242065,
      "loss": 4.8659,
      "mean_token_accuracy": 0.1917388394474983,
      "num_tokens": 68229758.0,
      "step": 31145
    },
    {
      "entropy": 5.215186977386475,
      "epoch": 3.091504565303692,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040596422355606383,
      "loss": 4.848,
      "mean_token_accuracy": 0.19155573546886445,
      "num_tokens": 68239146.0,
      "step": 31150
    },
    {
      "entropy": 5.277754640579223,
      "epoch": 3.0920007939658594,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004059354110402905,
      "loss": 4.9328,
      "mean_token_accuracy": 0.17966222167015075,
      "num_tokens": 68249339.0,
      "step": 31155
    },
    {
      "entropy": 5.3320389747619625,
      "epoch": 3.092497022628027,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040590659527760123,
      "loss": 4.9026,
      "mean_token_accuracy": 0.18747304528951644,
      "num_tokens": 68260136.0,
      "step": 31160
    },
    {
      "entropy": 5.298227310180664,
      "epoch": 3.0929932512901943,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040587777626871046,
      "loss": 4.911,
      "mean_token_accuracy": 0.1840634748339653,
      "num_tokens": 68270033.0,
      "step": 31165
    },
    {
      "entropy": 5.320271015167236,
      "epoch": 3.0934894799523622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004058489540143328,
      "loss": 4.9675,
      "mean_token_accuracy": 0.18014969825744628,
      "num_tokens": 68283344.0,
      "step": 31170
    },
    {
      "entropy": 5.321994924545288,
      "epoch": 3.0939857086145297,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004058201285151831,
      "loss": 4.8925,
      "mean_token_accuracy": 0.18640560209751128,
      "num_tokens": 68294604.0,
      "step": 31175
    },
    {
      "entropy": 5.312696504592895,
      "epoch": 3.094481937276697,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004057912997719761,
      "loss": 4.8621,
      "mean_token_accuracy": 0.19480335116386413,
      "num_tokens": 68304436.0,
      "step": 31180
    },
    {
      "entropy": 5.236658239364624,
      "epoch": 3.0949781659388647,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004057624677854266,
      "loss": 4.8413,
      "mean_token_accuracy": 0.1917276367545128,
      "num_tokens": 68315593.0,
      "step": 31185
    },
    {
      "entropy": 5.299475193023682,
      "epoch": 3.095474394601032,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040573363255624976,
      "loss": 4.9409,
      "mean_token_accuracy": 0.1939719185233116,
      "num_tokens": 68327540.0,
      "step": 31190
    },
    {
      "entropy": 5.406162071228027,
      "epoch": 3.0959706232631996,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004057047940851605,
      "loss": 5.0153,
      "mean_token_accuracy": 0.18066019266843797,
      "num_tokens": 68339784.0,
      "step": 31195
    },
    {
      "entropy": 5.30894250869751,
      "epoch": 3.096466851925367,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004056759523728739,
      "loss": 4.9149,
      "mean_token_accuracy": 0.18158739805221558,
      "num_tokens": 68351182.0,
      "step": 31200
    },
    {
      "entropy": 5.302634811401367,
      "epoch": 3.0969630805875346,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040564710742010545,
      "loss": 4.9607,
      "mean_token_accuracy": 0.18224334567785264,
      "num_tokens": 68361726.0,
      "step": 31205
    },
    {
      "entropy": 5.3128979206085205,
      "epoch": 3.0974593092497025,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004056182592275702,
      "loss": 4.9304,
      "mean_token_accuracy": 0.18476719558238983,
      "num_tokens": 68373167.0,
      "step": 31210
    },
    {
      "entropy": 5.343462038040161,
      "epoch": 3.09795553791187,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004055894077959835,
      "loss": 4.9704,
      "mean_token_accuracy": 0.1861750826239586,
      "num_tokens": 68384646.0,
      "step": 31215
    },
    {
      "entropy": 5.3098842144012455,
      "epoch": 3.0984517665740374,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004055605531260609,
      "loss": 4.9249,
      "mean_token_accuracy": 0.18401724994182586,
      "num_tokens": 68395953.0,
      "step": 31220
    },
    {
      "entropy": 5.348503065109253,
      "epoch": 3.098947995236205,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040553169521851794,
      "loss": 4.9654,
      "mean_token_accuracy": 0.17950978875160217,
      "num_tokens": 68407167.0,
      "step": 31225
    },
    {
      "entropy": 5.314312314987182,
      "epoch": 3.0994442238983724,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040550283407407025,
      "loss": 4.9893,
      "mean_token_accuracy": 0.1843811884522438,
      "num_tokens": 68419530.0,
      "step": 31230
    },
    {
      "entropy": 5.323176431655884,
      "epoch": 3.09994045256054,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004054739696934334,
      "loss": 4.925,
      "mean_token_accuracy": 0.18117129504680635,
      "num_tokens": 68430970.0,
      "step": 31235
    },
    {
      "entropy": 5.317126989364624,
      "epoch": 3.1004366812227073,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040544510207732337,
      "loss": 4.9074,
      "mean_token_accuracy": 0.18501434326171876,
      "num_tokens": 68441257.0,
      "step": 31240
    },
    {
      "entropy": 5.251178646087647,
      "epoch": 3.100932909884875,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040541623122645576,
      "loss": 4.8971,
      "mean_token_accuracy": 0.1882983237504959,
      "num_tokens": 68452493.0,
      "step": 31245
    },
    {
      "entropy": 5.272414207458496,
      "epoch": 3.1014291385470423,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040538735714154675,
      "loss": 4.8418,
      "mean_token_accuracy": 0.20019327402114867,
      "num_tokens": 68464023.0,
      "step": 31250
    },
    {
      "entropy": 5.337690544128418,
      "epoch": 3.10192536720921,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004053584798233122,
      "loss": 4.9248,
      "mean_token_accuracy": 0.19343167245388032,
      "num_tokens": 68474519.0,
      "step": 31255
    },
    {
      "entropy": 5.343744516372681,
      "epoch": 3.1024215958713777,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040532959927246814,
      "loss": 4.9137,
      "mean_token_accuracy": 0.19006831645965577,
      "num_tokens": 68486239.0,
      "step": 31260
    },
    {
      "entropy": 5.334685134887695,
      "epoch": 3.102917824533545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000405300715489731,
      "loss": 4.967,
      "mean_token_accuracy": 0.1811900332570076,
      "num_tokens": 68495921.0,
      "step": 31265
    },
    {
      "entropy": 5.380409145355225,
      "epoch": 3.1034140531957126,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004052718284758169,
      "loss": 4.8627,
      "mean_token_accuracy": 0.19212860763072967,
      "num_tokens": 68505773.0,
      "step": 31270
    },
    {
      "entropy": 5.385918474197387,
      "epoch": 3.10391028185788,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000405242938231442,
      "loss": 4.9834,
      "mean_token_accuracy": 0.1828041523694992,
      "num_tokens": 68516563.0,
      "step": 31275
    },
    {
      "entropy": 5.314352941513062,
      "epoch": 3.1044065105200476,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000405214044757323,
      "loss": 4.9369,
      "mean_token_accuracy": 0.19126426428556442,
      "num_tokens": 68527952.0,
      "step": 31280
    },
    {
      "entropy": 5.341227626800537,
      "epoch": 3.104902739182215,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004051851480541763,
      "loss": 4.9118,
      "mean_token_accuracy": 0.1815643012523651,
      "num_tokens": 68538450.0,
      "step": 31285
    },
    {
      "entropy": 5.268524074554444,
      "epoch": 3.1053989678443825,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004051562481227184,
      "loss": 4.8722,
      "mean_token_accuracy": 0.18379394263029097,
      "num_tokens": 68547941.0,
      "step": 31290
    },
    {
      "entropy": 5.249780654907227,
      "epoch": 3.10589519650655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000405127344963666,
      "loss": 4.8629,
      "mean_token_accuracy": 0.19278552532196044,
      "num_tokens": 68558525.0,
      "step": 31295
    },
    {
      "entropy": 5.252750110626221,
      "epoch": 3.106391425168718,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004050984385777359,
      "loss": 4.857,
      "mean_token_accuracy": 0.1887189045548439,
      "num_tokens": 68569487.0,
      "step": 31300
    },
    {
      "entropy": 5.30226092338562,
      "epoch": 3.1068876538308854,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004050695289656448,
      "loss": 4.8962,
      "mean_token_accuracy": 0.18835106790065764,
      "num_tokens": 68580069.0,
      "step": 31305
    },
    {
      "entropy": 5.320374059677124,
      "epoch": 3.107383882493053,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040504061612810964,
      "loss": 4.8738,
      "mean_token_accuracy": 0.18854302316904067,
      "num_tokens": 68591511.0,
      "step": 31310
    },
    {
      "entropy": 5.207733201980591,
      "epoch": 3.1078801111552203,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004050117000658475,
      "loss": 4.9194,
      "mean_token_accuracy": 0.18775196522474288,
      "num_tokens": 68602464.0,
      "step": 31315
    },
    {
      "entropy": 5.332287263870239,
      "epoch": 3.108376339817388,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040498278077957534,
      "loss": 4.8708,
      "mean_token_accuracy": 0.18583334535360335,
      "num_tokens": 68613492.0,
      "step": 31320
    },
    {
      "entropy": 5.2472679138183596,
      "epoch": 3.1088725684795553,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004049538582700102,
      "loss": 4.8349,
      "mean_token_accuracy": 0.19183728843927383,
      "num_tokens": 68625259.0,
      "step": 31325
    },
    {
      "entropy": 5.323595428466797,
      "epoch": 3.1093687971417228,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004049249325378695,
      "loss": 4.9778,
      "mean_token_accuracy": 0.182860766351223,
      "num_tokens": 68635455.0,
      "step": 31330
    },
    {
      "entropy": 5.329261302947998,
      "epoch": 3.1098650258038902,
      "grad_norm": 0.875,
      "learning_rate": 0.00040489600358387036,
      "loss": 4.9818,
      "mean_token_accuracy": 0.18514414578676225,
      "num_tokens": 68648102.0,
      "step": 31335
    },
    {
      "entropy": 5.336288738250732,
      "epoch": 3.110361254466058,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004048670714087303,
      "loss": 4.9423,
      "mean_token_accuracy": 0.188930943608284,
      "num_tokens": 68660854.0,
      "step": 31340
    },
    {
      "entropy": 5.1920631408691404,
      "epoch": 3.1108574831282256,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004048381360131667,
      "loss": 4.7677,
      "mean_token_accuracy": 0.20397522002458573,
      "num_tokens": 68671839.0,
      "step": 31345
    },
    {
      "entropy": 5.359168148040771,
      "epoch": 3.111353711790393,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004048091973978971,
      "loss": 4.9227,
      "mean_token_accuracy": 0.19097465723752977,
      "num_tokens": 68682535.0,
      "step": 31350
    },
    {
      "entropy": 5.2828648567199705,
      "epoch": 3.1118499404525606,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004047802555636391,
      "loss": 4.8935,
      "mean_token_accuracy": 0.18944518715143205,
      "num_tokens": 68694283.0,
      "step": 31355
    },
    {
      "entropy": 5.321843671798706,
      "epoch": 3.112346169114728,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004047513105111104,
      "loss": 4.9195,
      "mean_token_accuracy": 0.182905675470829,
      "num_tokens": 68704268.0,
      "step": 31360
    },
    {
      "entropy": 5.283941459655762,
      "epoch": 3.1128423977768955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040472236224102883,
      "loss": 4.826,
      "mean_token_accuracy": 0.18626630008220674,
      "num_tokens": 68714762.0,
      "step": 31365
    },
    {
      "entropy": 5.26569676399231,
      "epoch": 3.113338626439063,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040469341075411217,
      "loss": 4.8648,
      "mean_token_accuracy": 0.18009740114212036,
      "num_tokens": 68725302.0,
      "step": 31370
    },
    {
      "entropy": 5.317063570022583,
      "epoch": 3.1138348551012305,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004046644560510783,
      "loss": 4.9742,
      "mean_token_accuracy": 0.17642245590686798,
      "num_tokens": 68736893.0,
      "step": 31375
    },
    {
      "entropy": 5.27666163444519,
      "epoch": 3.114331083763398,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040463549813264557,
      "loss": 4.8625,
      "mean_token_accuracy": 0.1901179850101471,
      "num_tokens": 68748202.0,
      "step": 31380
    },
    {
      "entropy": 5.375661563873291,
      "epoch": 3.114827312425566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040460653699953166,
      "loss": 4.8931,
      "mean_token_accuracy": 0.19433989077806474,
      "num_tokens": 68759033.0,
      "step": 31385
    },
    {
      "entropy": 5.292493486404419,
      "epoch": 3.1153235410877333,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040457757265245485,
      "loss": 4.8699,
      "mean_token_accuracy": 0.19111834913492204,
      "num_tokens": 68770879.0,
      "step": 31390
    },
    {
      "entropy": 5.254086208343506,
      "epoch": 3.115819769749901,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004045486050921336,
      "loss": 4.8418,
      "mean_token_accuracy": 0.1964523121714592,
      "num_tokens": 68782087.0,
      "step": 31395
    },
    {
      "entropy": 5.299632883071899,
      "epoch": 3.1163159984120683,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000404519634319286,
      "loss": 4.8893,
      "mean_token_accuracy": 0.1857980340719223,
      "num_tokens": 68792355.0,
      "step": 31400
    },
    {
      "entropy": 5.289393949508667,
      "epoch": 3.1168122270742358,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040449066033463053,
      "loss": 4.9135,
      "mean_token_accuracy": 0.1838744044303894,
      "num_tokens": 68804255.0,
      "step": 31405
    },
    {
      "entropy": 5.272180891036987,
      "epoch": 3.1173084557364032,
      "grad_norm": 1.0,
      "learning_rate": 0.0004044616831388858,
      "loss": 4.8675,
      "mean_token_accuracy": 0.18819535076618193,
      "num_tokens": 68816102.0,
      "step": 31410
    },
    {
      "entropy": 5.296196222305298,
      "epoch": 3.1178046843985707,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004044327027327703,
      "loss": 4.8955,
      "mean_token_accuracy": 0.1898640662431717,
      "num_tokens": 68827303.0,
      "step": 31415
    },
    {
      "entropy": 5.298006153106689,
      "epoch": 3.118300913060738,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004044037191170026,
      "loss": 4.8822,
      "mean_token_accuracy": 0.19024822860956192,
      "num_tokens": 68838418.0,
      "step": 31420
    },
    {
      "entropy": 5.3678505420684814,
      "epoch": 3.118797141722906,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040437473229230155,
      "loss": 4.9385,
      "mean_token_accuracy": 0.18362030535936355,
      "num_tokens": 68848431.0,
      "step": 31425
    },
    {
      "entropy": 5.267582130432129,
      "epoch": 3.1192933703850736,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004043457422593859,
      "loss": 4.8347,
      "mean_token_accuracy": 0.18409923315048218,
      "num_tokens": 68858983.0,
      "step": 31430
    },
    {
      "entropy": 5.2895763397216795,
      "epoch": 3.119789599047241,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004043167490189746,
      "loss": 4.9086,
      "mean_token_accuracy": 0.18996853977441788,
      "num_tokens": 68870170.0,
      "step": 31435
    },
    {
      "entropy": 5.299309110641479,
      "epoch": 3.1202858277094085,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040428775257178655,
      "loss": 4.9385,
      "mean_token_accuracy": 0.19314604103565217,
      "num_tokens": 68880242.0,
      "step": 31440
    },
    {
      "entropy": 5.358444118499756,
      "epoch": 3.120782056371576,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040425875291854083,
      "loss": 5.0014,
      "mean_token_accuracy": 0.1790206700563431,
      "num_tokens": 68892150.0,
      "step": 31445
    },
    {
      "entropy": 5.399437522888183,
      "epoch": 3.1212782850337435,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040422975005995656,
      "loss": 4.9809,
      "mean_token_accuracy": 0.18015486001968384,
      "num_tokens": 68903164.0,
      "step": 31450
    },
    {
      "entropy": 5.279314804077148,
      "epoch": 3.121774513695911,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004042007439967529,
      "loss": 4.8639,
      "mean_token_accuracy": 0.1883976399898529,
      "num_tokens": 68914514.0,
      "step": 31455
    },
    {
      "entropy": 5.2970850467681885,
      "epoch": 3.1222707423580784,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004041717347296493,
      "loss": 4.8951,
      "mean_token_accuracy": 0.18990073204040528,
      "num_tokens": 68925366.0,
      "step": 31460
    },
    {
      "entropy": 5.279950523376465,
      "epoch": 3.122766971020246,
      "grad_norm": 1.0,
      "learning_rate": 0.0004041427222593649,
      "loss": 4.845,
      "mean_token_accuracy": 0.1878553092479706,
      "num_tokens": 68936019.0,
      "step": 31465
    },
    {
      "entropy": 5.186664056777954,
      "epoch": 3.123263199682414,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040411370658661935,
      "loss": 4.8218,
      "mean_token_accuracy": 0.19330547749996185,
      "num_tokens": 68947405.0,
      "step": 31470
    },
    {
      "entropy": 5.2955827713012695,
      "epoch": 3.1237594283445813,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040408468771213206,
      "loss": 4.9139,
      "mean_token_accuracy": 0.1856538474559784,
      "num_tokens": 68958822.0,
      "step": 31475
    },
    {
      "entropy": 5.317957830429077,
      "epoch": 3.124255657006749,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040405566563662265,
      "loss": 4.9434,
      "mean_token_accuracy": 0.18680141121149063,
      "num_tokens": 68969992.0,
      "step": 31480
    },
    {
      "entropy": 5.290104818344116,
      "epoch": 3.1247518856689163,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040402664036081086,
      "loss": 4.8817,
      "mean_token_accuracy": 0.1818320631980896,
      "num_tokens": 68982081.0,
      "step": 31485
    },
    {
      "entropy": 5.356375122070313,
      "epoch": 3.1252481143310837,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004039976118854164,
      "loss": 4.92,
      "mean_token_accuracy": 0.18407747149467468,
      "num_tokens": 68993430.0,
      "step": 31490
    },
    {
      "entropy": 5.390209674835205,
      "epoch": 3.125744342993251,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004039685802111591,
      "loss": 4.9867,
      "mean_token_accuracy": 0.18603895604610443,
      "num_tokens": 69004939.0,
      "step": 31495
    },
    {
      "entropy": 5.2925787448883055,
      "epoch": 3.1262405716554187,
      "grad_norm": 1.0,
      "learning_rate": 0.00040393954533875904,
      "loss": 4.9141,
      "mean_token_accuracy": 0.18988742381334306,
      "num_tokens": 69016293.0,
      "step": 31500
    },
    {
      "entropy": 5.3498681545257565,
      "epoch": 3.126736800317586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040391050726893595,
      "loss": 5.0199,
      "mean_token_accuracy": 0.1784585878252983,
      "num_tokens": 69029087.0,
      "step": 31505
    },
    {
      "entropy": 5.327571773529053,
      "epoch": 3.127233028979754,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004038814660024101,
      "loss": 4.9022,
      "mean_token_accuracy": 0.18308320492506028,
      "num_tokens": 69040937.0,
      "step": 31510
    },
    {
      "entropy": 5.362440490722657,
      "epoch": 3.1277292576419216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004038524215399016,
      "loss": 4.9903,
      "mean_token_accuracy": 0.18333331644535064,
      "num_tokens": 69052212.0,
      "step": 31515
    },
    {
      "entropy": 5.332947397232056,
      "epoch": 3.128225486304089,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004038233738821308,
      "loss": 4.8853,
      "mean_token_accuracy": 0.19323972165584563,
      "num_tokens": 69061951.0,
      "step": 31520
    },
    {
      "entropy": 5.326661586761475,
      "epoch": 3.1287217149662565,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004037943230298178,
      "loss": 4.8403,
      "mean_token_accuracy": 0.19664566367864608,
      "num_tokens": 69072156.0,
      "step": 31525
    },
    {
      "entropy": 5.286450481414795,
      "epoch": 3.129217943628424,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040376526898368314,
      "loss": 4.8966,
      "mean_token_accuracy": 0.18480471074581145,
      "num_tokens": 69084032.0,
      "step": 31530
    },
    {
      "entropy": 5.312008190155029,
      "epoch": 3.1297141722905915,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040373621174444736,
      "loss": 4.9522,
      "mean_token_accuracy": 0.1804022952914238,
      "num_tokens": 69096436.0,
      "step": 31535
    },
    {
      "entropy": 5.2932336807250975,
      "epoch": 3.130210400952759,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040370715131283076,
      "loss": 4.8666,
      "mean_token_accuracy": 0.18938516080379486,
      "num_tokens": 69107090.0,
      "step": 31540
    },
    {
      "entropy": 5.301195049285889,
      "epoch": 3.1307066296149264,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004036780876895543,
      "loss": 4.834,
      "mean_token_accuracy": 0.19179490953683853,
      "num_tokens": 69117644.0,
      "step": 31545
    },
    {
      "entropy": 5.2501692295074465,
      "epoch": 3.131202858277094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004036490208753385,
      "loss": 4.9097,
      "mean_token_accuracy": 0.1843639776110649,
      "num_tokens": 69129639.0,
      "step": 31550
    },
    {
      "entropy": 5.264796590805053,
      "epoch": 3.131699086939262,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040361995087090407,
      "loss": 4.8669,
      "mean_token_accuracy": 0.18711763620376587,
      "num_tokens": 69139429.0,
      "step": 31555
    },
    {
      "entropy": 5.186162996292114,
      "epoch": 3.1321953156014293,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040359087767697223,
      "loss": 4.7995,
      "mean_token_accuracy": 0.1972327634692192,
      "num_tokens": 69150777.0,
      "step": 31560
    },
    {
      "entropy": 5.31248369216919,
      "epoch": 3.1326915442635968,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040356180129426356,
      "loss": 4.8776,
      "mean_token_accuracy": 0.1829502537846565,
      "num_tokens": 69161970.0,
      "step": 31565
    },
    {
      "entropy": 5.228343772888183,
      "epoch": 3.1331877729257642,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004035327217234993,
      "loss": 4.8702,
      "mean_token_accuracy": 0.18989089131355286,
      "num_tokens": 69172990.0,
      "step": 31570
    },
    {
      "entropy": 5.273339414596558,
      "epoch": 3.1336840015879317,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040350363896540054,
      "loss": 4.9061,
      "mean_token_accuracy": 0.1856796532869339,
      "num_tokens": 69184044.0,
      "step": 31575
    },
    {
      "entropy": 5.230098009109497,
      "epoch": 3.134180230250099,
      "grad_norm": 1.0,
      "learning_rate": 0.00040347455302068835,
      "loss": 4.8725,
      "mean_token_accuracy": 0.17592404186725616,
      "num_tokens": 69194841.0,
      "step": 31580
    },
    {
      "entropy": 5.305024576187134,
      "epoch": 3.1346764589122667,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040344546389008406,
      "loss": 4.9179,
      "mean_token_accuracy": 0.185160094499588,
      "num_tokens": 69205465.0,
      "step": 31585
    },
    {
      "entropy": 5.322886800765991,
      "epoch": 3.135172687574434,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040341637157430916,
      "loss": 4.9501,
      "mean_token_accuracy": 0.1831107646226883,
      "num_tokens": 69215831.0,
      "step": 31590
    },
    {
      "entropy": 5.30539755821228,
      "epoch": 3.135668916236602,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004033872760740849,
      "loss": 4.8897,
      "mean_token_accuracy": 0.19352262765169143,
      "num_tokens": 69225156.0,
      "step": 31595
    },
    {
      "entropy": 5.255848455429077,
      "epoch": 3.1361651448987695,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004033581773901328,
      "loss": 4.9312,
      "mean_token_accuracy": 0.1829553186893463,
      "num_tokens": 69237590.0,
      "step": 31600
    },
    {
      "entropy": 5.331939268112182,
      "epoch": 3.136661373560937,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040332907552317453,
      "loss": 4.8872,
      "mean_token_accuracy": 0.19307206124067305,
      "num_tokens": 69246904.0,
      "step": 31605
    },
    {
      "entropy": 5.455360746383667,
      "epoch": 3.1371576022231045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004032999704739317,
      "loss": 5.0063,
      "mean_token_accuracy": 0.17620910108089446,
      "num_tokens": 69258289.0,
      "step": 31610
    },
    {
      "entropy": 5.287320470809936,
      "epoch": 3.137653830885272,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004032708622431261,
      "loss": 4.9015,
      "mean_token_accuracy": 0.18390351235866548,
      "num_tokens": 69268232.0,
      "step": 31615
    },
    {
      "entropy": 5.292891645431519,
      "epoch": 3.1381500595474394,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004032417508314795,
      "loss": 4.8804,
      "mean_token_accuracy": 0.18013941496610641,
      "num_tokens": 69278596.0,
      "step": 31620
    },
    {
      "entropy": 5.309746837615966,
      "epoch": 3.138646288209607,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040321263623971376,
      "loss": 4.9841,
      "mean_token_accuracy": 0.18259136229753495,
      "num_tokens": 69290675.0,
      "step": 31625
    },
    {
      "entropy": 5.358918714523315,
      "epoch": 3.1391425168717744,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000403183518468551,
      "loss": 4.9864,
      "mean_token_accuracy": 0.18748716562986373,
      "num_tokens": 69303600.0,
      "step": 31630
    },
    {
      "entropy": 5.396495389938354,
      "epoch": 3.139638745533942,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004031543975187132,
      "loss": 5.0077,
      "mean_token_accuracy": 0.1808657929301262,
      "num_tokens": 69314672.0,
      "step": 31635
    },
    {
      "entropy": 5.267249870300293,
      "epoch": 3.1401349741961098,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040312527339092237,
      "loss": 4.8591,
      "mean_token_accuracy": 0.19518404752016066,
      "num_tokens": 69325574.0,
      "step": 31640
    },
    {
      "entropy": 5.241053771972656,
      "epoch": 3.1406312028582772,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040309614608590095,
      "loss": 4.8531,
      "mean_token_accuracy": 0.18419217616319655,
      "num_tokens": 69336430.0,
      "step": 31645
    },
    {
      "entropy": 5.239997625350952,
      "epoch": 3.1411274315204447,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004030670156043711,
      "loss": 4.77,
      "mean_token_accuracy": 0.20274640321731568,
      "num_tokens": 69347007.0,
      "step": 31650
    },
    {
      "entropy": 5.345808601379394,
      "epoch": 3.141623660182612,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040303788194705523,
      "loss": 4.9207,
      "mean_token_accuracy": 0.18516316562891005,
      "num_tokens": 69357796.0,
      "step": 31655
    },
    {
      "entropy": 5.308848857879639,
      "epoch": 3.1421198888447797,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040300874511467574,
      "loss": 4.9303,
      "mean_token_accuracy": 0.1879841148853302,
      "num_tokens": 69370098.0,
      "step": 31660
    },
    {
      "entropy": 5.279611206054687,
      "epoch": 3.142616117506947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040297960510795526,
      "loss": 4.9138,
      "mean_token_accuracy": 0.1749950811266899,
      "num_tokens": 69380991.0,
      "step": 31665
    },
    {
      "entropy": 5.332782363891601,
      "epoch": 3.1431123461691146,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004029504619276163,
      "loss": 4.8968,
      "mean_token_accuracy": 0.18251143246889115,
      "num_tokens": 69391748.0,
      "step": 31670
    },
    {
      "entropy": 5.3742029666900635,
      "epoch": 3.143608574831282,
      "grad_norm": 0.890625,
      "learning_rate": 0.00040292131557438166,
      "loss": 4.9426,
      "mean_token_accuracy": 0.17778950333595275,
      "num_tokens": 69404964.0,
      "step": 31675
    },
    {
      "entropy": 5.237097644805909,
      "epoch": 3.14410480349345,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000402892166048974,
      "loss": 4.885,
      "mean_token_accuracy": 0.19233371317386627,
      "num_tokens": 69416045.0,
      "step": 31680
    },
    {
      "entropy": 5.252412271499634,
      "epoch": 3.1446010321556175,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004028630133521162,
      "loss": 4.8665,
      "mean_token_accuracy": 0.18257110118865966,
      "num_tokens": 69427489.0,
      "step": 31685
    },
    {
      "entropy": 5.287076377868653,
      "epoch": 3.145097260817785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004028338574845312,
      "loss": 4.8693,
      "mean_token_accuracy": 0.18724656999111175,
      "num_tokens": 69437929.0,
      "step": 31690
    },
    {
      "entropy": 5.3351860523223875,
      "epoch": 3.1455934894799524,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040280469844694203,
      "loss": 4.922,
      "mean_token_accuracy": 0.1895683690905571,
      "num_tokens": 69449449.0,
      "step": 31695
    },
    {
      "entropy": 5.268132591247559,
      "epoch": 3.14608971814212,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004027755362400716,
      "loss": 4.9331,
      "mean_token_accuracy": 0.18419099748134612,
      "num_tokens": 69459451.0,
      "step": 31700
    },
    {
      "entropy": 5.365521049499511,
      "epoch": 3.1465859468042874,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004027463708646434,
      "loss": 4.9494,
      "mean_token_accuracy": 0.18192971795797347,
      "num_tokens": 69470615.0,
      "step": 31705
    },
    {
      "entropy": 5.351056337356567,
      "epoch": 3.147082175466455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004027172023213803,
      "loss": 4.8954,
      "mean_token_accuracy": 0.18886639028787613,
      "num_tokens": 69481123.0,
      "step": 31710
    },
    {
      "entropy": 5.333863735198975,
      "epoch": 3.1475784041286223,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004026880306110058,
      "loss": 4.9346,
      "mean_token_accuracy": 0.18404506295919418,
      "num_tokens": 69491590.0,
      "step": 31715
    },
    {
      "entropy": 5.236373043060302,
      "epoch": 3.14807463279079,
      "grad_norm": 1.0,
      "learning_rate": 0.00040265885573424336,
      "loss": 4.8913,
      "mean_token_accuracy": 0.18725781589746476,
      "num_tokens": 69501883.0,
      "step": 31720
    },
    {
      "entropy": 5.269284915924072,
      "epoch": 3.1485708614529577,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040262967769181634,
      "loss": 4.8653,
      "mean_token_accuracy": 0.1854100152850151,
      "num_tokens": 69514556.0,
      "step": 31725
    },
    {
      "entropy": 5.403923749923706,
      "epoch": 3.149067090115125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040260049648444836,
      "loss": 4.9916,
      "mean_token_accuracy": 0.18379950523376465,
      "num_tokens": 69525483.0,
      "step": 31730
    },
    {
      "entropy": 5.355910921096802,
      "epoch": 3.1495633187772927,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040257131211286307,
      "loss": 4.9323,
      "mean_token_accuracy": 0.18630514591932296,
      "num_tokens": 69535681.0,
      "step": 31735
    },
    {
      "entropy": 5.3409584045410154,
      "epoch": 3.15005954743946,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000402542124577784,
      "loss": 5.0655,
      "mean_token_accuracy": 0.167495259642601,
      "num_tokens": 69547825.0,
      "step": 31740
    },
    {
      "entropy": 5.259937047958374,
      "epoch": 3.1505557761016276,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040251293387993513,
      "loss": 4.8978,
      "mean_token_accuracy": 0.1859361231327057,
      "num_tokens": 69557922.0,
      "step": 31745
    },
    {
      "entropy": 5.319004058837891,
      "epoch": 3.151052004763795,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004024837400200403,
      "loss": 4.9571,
      "mean_token_accuracy": 0.18541821837425232,
      "num_tokens": 69568165.0,
      "step": 31750
    },
    {
      "entropy": 5.245988941192627,
      "epoch": 3.1515482334259626,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040245454299882343,
      "loss": 4.7497,
      "mean_token_accuracy": 0.19644167572259902,
      "num_tokens": 69578828.0,
      "step": 31755
    },
    {
      "entropy": 5.3537742614746096,
      "epoch": 3.15204446208813,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040242534281700843,
      "loss": 4.9577,
      "mean_token_accuracy": 0.17756276577711105,
      "num_tokens": 69590527.0,
      "step": 31760
    },
    {
      "entropy": 5.334926319122315,
      "epoch": 3.152540690750298,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040239613947531965,
      "loss": 4.919,
      "mean_token_accuracy": 0.18715827614068986,
      "num_tokens": 69602032.0,
      "step": 31765
    },
    {
      "entropy": 5.260985422134399,
      "epoch": 3.1530369194124654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040236693297448107,
      "loss": 4.9209,
      "mean_token_accuracy": 0.18140140920877457,
      "num_tokens": 69613490.0,
      "step": 31770
    },
    {
      "entropy": 5.377650403976441,
      "epoch": 3.153533148074633,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004023377233152169,
      "loss": 5.0035,
      "mean_token_accuracy": 0.17712698876857758,
      "num_tokens": 69624623.0,
      "step": 31775
    },
    {
      "entropy": 5.365924501419068,
      "epoch": 3.1540293767368004,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040230851049825173,
      "loss": 4.9724,
      "mean_token_accuracy": 0.18062271773815156,
      "num_tokens": 69634956.0,
      "step": 31780
    },
    {
      "entropy": 5.319891309738159,
      "epoch": 3.154525605398968,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040227929452430974,
      "loss": 4.8824,
      "mean_token_accuracy": 0.1900096207857132,
      "num_tokens": 69645370.0,
      "step": 31785
    },
    {
      "entropy": 5.366716623306274,
      "epoch": 3.1550218340611353,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00040225007539411553,
      "loss": 4.9722,
      "mean_token_accuracy": 0.17325596809387206,
      "num_tokens": 69656328.0,
      "step": 31790
    },
    {
      "entropy": 5.2842926502227785,
      "epoch": 3.155518062723303,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004022208531083936,
      "loss": 4.9021,
      "mean_token_accuracy": 0.18627944886684417,
      "num_tokens": 69668584.0,
      "step": 31795
    },
    {
      "entropy": 5.279483413696289,
      "epoch": 3.1560142913854703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004021916276678688,
      "loss": 4.9307,
      "mean_token_accuracy": 0.18037495911121368,
      "num_tokens": 69677977.0,
      "step": 31800
    },
    {
      "entropy": 5.295825719833374,
      "epoch": 3.1565105200476378,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040216239907326555,
      "loss": 4.9126,
      "mean_token_accuracy": 0.18898969292640685,
      "num_tokens": 69688979.0,
      "step": 31805
    },
    {
      "entropy": 5.2933837413787845,
      "epoch": 3.1570067487098052,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004021331673253089,
      "loss": 4.9587,
      "mean_token_accuracy": 0.190383842587471,
      "num_tokens": 69698567.0,
      "step": 31810
    },
    {
      "entropy": 5.330805778503418,
      "epoch": 3.157502977371973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004021039324247236,
      "loss": 4.8654,
      "mean_token_accuracy": 0.18453508764505386,
      "num_tokens": 69708623.0,
      "step": 31815
    },
    {
      "entropy": 5.268275403976441,
      "epoch": 3.1579992060341406,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040207469437223466,
      "loss": 4.9302,
      "mean_token_accuracy": 0.18515763133764268,
      "num_tokens": 69719613.0,
      "step": 31820
    },
    {
      "entropy": 5.301593971252442,
      "epoch": 3.158495434696308,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040204545316856714,
      "loss": 4.9214,
      "mean_token_accuracy": 0.18621871918439864,
      "num_tokens": 69731407.0,
      "step": 31825
    },
    {
      "entropy": 5.269184637069702,
      "epoch": 3.1589916633584756,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004020162088144461,
      "loss": 4.8581,
      "mean_token_accuracy": 0.1856224164366722,
      "num_tokens": 69742363.0,
      "step": 31830
    },
    {
      "entropy": 5.267366027832031,
      "epoch": 3.159487892020643,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040198696131059684,
      "loss": 4.8601,
      "mean_token_accuracy": 0.19207378029823302,
      "num_tokens": 69754878.0,
      "step": 31835
    },
    {
      "entropy": 5.284064865112304,
      "epoch": 3.1599841206828105,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040195771065774455,
      "loss": 4.8439,
      "mean_token_accuracy": 0.19277852922677993,
      "num_tokens": 69764986.0,
      "step": 31840
    },
    {
      "entropy": 5.2761815071105955,
      "epoch": 3.160480349344978,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004019284568566146,
      "loss": 4.888,
      "mean_token_accuracy": 0.1865689367055893,
      "num_tokens": 69775762.0,
      "step": 31845
    },
    {
      "entropy": 5.309274387359619,
      "epoch": 3.1609765780071455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040189919990793243,
      "loss": 4.9512,
      "mean_token_accuracy": 0.1850581392645836,
      "num_tokens": 69786411.0,
      "step": 31850
    },
    {
      "entropy": 5.365603876113892,
      "epoch": 3.1614728066693134,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040186993981242353,
      "loss": 4.9026,
      "mean_token_accuracy": 0.185126368701458,
      "num_tokens": 69797140.0,
      "step": 31855
    },
    {
      "entropy": 5.260383558273316,
      "epoch": 3.161969035331481,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004018406765708135,
      "loss": 4.7933,
      "mean_token_accuracy": 0.19425053000450135,
      "num_tokens": 69807452.0,
      "step": 31860
    },
    {
      "entropy": 5.236143255233765,
      "epoch": 3.1624652639936484,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040181141018382806,
      "loss": 4.9023,
      "mean_token_accuracy": 0.18749292939901352,
      "num_tokens": 69819650.0,
      "step": 31865
    },
    {
      "entropy": 5.277122735977173,
      "epoch": 3.162961492655816,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040178214065219285,
      "loss": 4.8815,
      "mean_token_accuracy": 0.18540531843900682,
      "num_tokens": 69830411.0,
      "step": 31870
    },
    {
      "entropy": 5.280612373352051,
      "epoch": 3.1634577213179833,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004017528679766337,
      "loss": 4.8722,
      "mean_token_accuracy": 0.19434474557638168,
      "num_tokens": 69841795.0,
      "step": 31875
    },
    {
      "entropy": 5.252249908447266,
      "epoch": 3.163953949980151,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004017235921578767,
      "loss": 4.883,
      "mean_token_accuracy": 0.182961542904377,
      "num_tokens": 69851633.0,
      "step": 31880
    },
    {
      "entropy": 5.269833278656006,
      "epoch": 3.1644501786423183,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040169431319664756,
      "loss": 4.8033,
      "mean_token_accuracy": 0.1997184306383133,
      "num_tokens": 69862398.0,
      "step": 31885
    },
    {
      "entropy": 5.271602630615234,
      "epoch": 3.1649464073044857,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004016650310936725,
      "loss": 4.8856,
      "mean_token_accuracy": 0.18724875301122665,
      "num_tokens": 69874397.0,
      "step": 31890
    },
    {
      "entropy": 5.266568231582641,
      "epoch": 3.165442635966653,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004016357458496776,
      "loss": 4.849,
      "mean_token_accuracy": 0.1959835797548294,
      "num_tokens": 69883950.0,
      "step": 31895
    },
    {
      "entropy": 5.27046275138855,
      "epoch": 3.165938864628821,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000401606457465389,
      "loss": 4.8212,
      "mean_token_accuracy": 0.19115711003541946,
      "num_tokens": 69895586.0,
      "step": 31900
    },
    {
      "entropy": 5.282785224914551,
      "epoch": 3.1664350932909886,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004015771659415332,
      "loss": 4.8716,
      "mean_token_accuracy": 0.19440239369869233,
      "num_tokens": 69906093.0,
      "step": 31905
    },
    {
      "entropy": 5.2363416194915775,
      "epoch": 3.166931321953156,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004015478712788364,
      "loss": 4.834,
      "mean_token_accuracy": 0.18752485811710357,
      "num_tokens": 69916381.0,
      "step": 31910
    },
    {
      "entropy": 5.245379257202148,
      "epoch": 3.1674275506153236,
      "grad_norm": 1.125,
      "learning_rate": 0.00040151857347802506,
      "loss": 4.8653,
      "mean_token_accuracy": 0.18954726159572602,
      "num_tokens": 69926059.0,
      "step": 31915
    },
    {
      "entropy": 5.2781288623809814,
      "epoch": 3.167923779277491,
      "grad_norm": 1.0,
      "learning_rate": 0.00040148927253982577,
      "loss": 4.9115,
      "mean_token_accuracy": 0.1827086701989174,
      "num_tokens": 69937677.0,
      "step": 31920
    },
    {
      "entropy": 5.255718421936035,
      "epoch": 3.1684200079396585,
      "grad_norm": 1.0,
      "learning_rate": 0.00040145996846496497,
      "loss": 4.8687,
      "mean_token_accuracy": 0.18736159056425095,
      "num_tokens": 69949399.0,
      "step": 31925
    },
    {
      "entropy": 5.301953744888306,
      "epoch": 3.168916236601826,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004014306612541695,
      "loss": 4.9326,
      "mean_token_accuracy": 0.1946787178516388,
      "num_tokens": 69960595.0,
      "step": 31930
    },
    {
      "entropy": 5.320375633239746,
      "epoch": 3.1694124652639935,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004014013509081661,
      "loss": 4.8896,
      "mean_token_accuracy": 0.18861862570047377,
      "num_tokens": 69971018.0,
      "step": 31935
    },
    {
      "entropy": 5.305331659317017,
      "epoch": 3.1699086939261614,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040137203742768155,
      "loss": 4.8821,
      "mean_token_accuracy": 0.18675601929426194,
      "num_tokens": 69982507.0,
      "step": 31940
    },
    {
      "entropy": 5.288206243515015,
      "epoch": 3.170404922588329,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004013427208134427,
      "loss": 4.8691,
      "mean_token_accuracy": 0.19823046028614044,
      "num_tokens": 69994190.0,
      "step": 31945
    },
    {
      "entropy": 5.30462007522583,
      "epoch": 3.1709011512504963,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040131340106617673,
      "loss": 5.0572,
      "mean_token_accuracy": 0.17996717393398284,
      "num_tokens": 70005522.0,
      "step": 31950
    },
    {
      "entropy": 5.418384981155396,
      "epoch": 3.171397379912664,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004012840781866106,
      "loss": 4.9703,
      "mean_token_accuracy": 0.19026967734098435,
      "num_tokens": 70016706.0,
      "step": 31955
    },
    {
      "entropy": 5.328632831573486,
      "epoch": 3.1718936085748313,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040125475217547135,
      "loss": 4.8601,
      "mean_token_accuracy": 0.18698727935552598,
      "num_tokens": 70028577.0,
      "step": 31960
    },
    {
      "entropy": 5.26553111076355,
      "epoch": 3.1723898372369987,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040122542303348634,
      "loss": 4.867,
      "mean_token_accuracy": 0.18887021541595458,
      "num_tokens": 70040258.0,
      "step": 31965
    },
    {
      "entropy": 5.277250576019287,
      "epoch": 3.172886065899166,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004011960907613828,
      "loss": 4.9911,
      "mean_token_accuracy": 0.18084984719753266,
      "num_tokens": 70052105.0,
      "step": 31970
    },
    {
      "entropy": 5.302644872665406,
      "epoch": 3.1733822945613337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040116675535988825,
      "loss": 4.8591,
      "mean_token_accuracy": 0.18826743811368943,
      "num_tokens": 70062272.0,
      "step": 31975
    },
    {
      "entropy": 5.317049217224121,
      "epoch": 3.173878523223501,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040113741682972994,
      "loss": 4.8958,
      "mean_token_accuracy": 0.18951108902692795,
      "num_tokens": 70073614.0,
      "step": 31980
    },
    {
      "entropy": 5.226245784759522,
      "epoch": 3.174374751885669,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004011080751716356,
      "loss": 4.8952,
      "mean_token_accuracy": 0.18302251547574996,
      "num_tokens": 70085191.0,
      "step": 31985
    },
    {
      "entropy": 5.323028612136841,
      "epoch": 3.1748709805478366,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040107873038633255,
      "loss": 4.8949,
      "mean_token_accuracy": 0.18393363654613495,
      "num_tokens": 70097610.0,
      "step": 31990
    },
    {
      "entropy": 5.255635547637939,
      "epoch": 3.175367209210004,
      "grad_norm": 1.0,
      "learning_rate": 0.0004010493824745487,
      "loss": 4.8302,
      "mean_token_accuracy": 0.19064592719078063,
      "num_tokens": 70109040.0,
      "step": 31995
    },
    {
      "entropy": 5.271688652038574,
      "epoch": 3.1758634378721715,
      "grad_norm": 1.0,
      "learning_rate": 0.00040102003143701194,
      "loss": 4.9153,
      "mean_token_accuracy": 0.1814916178584099,
      "num_tokens": 70120472.0,
      "step": 32000
    },
    {
      "entropy": 5.239098978042603,
      "epoch": 3.176359666534339,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040099067727444984,
      "loss": 4.7974,
      "mean_token_accuracy": 0.19133377820253372,
      "num_tokens": 70130481.0,
      "step": 32005
    },
    {
      "entropy": 5.252885246276856,
      "epoch": 3.1768558951965065,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004009613199875904,
      "loss": 4.8259,
      "mean_token_accuracy": 0.19237719029188155,
      "num_tokens": 70140783.0,
      "step": 32010
    },
    {
      "entropy": 5.288695383071899,
      "epoch": 3.177352123858674,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004009319595771617,
      "loss": 4.935,
      "mean_token_accuracy": 0.18788602501153945,
      "num_tokens": 70151624.0,
      "step": 32015
    },
    {
      "entropy": 5.343128633499146,
      "epoch": 3.1778483525208414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004009025960438918,
      "loss": 4.9148,
      "mean_token_accuracy": 0.1863637238740921,
      "num_tokens": 70162600.0,
      "step": 32020
    },
    {
      "entropy": 5.32385025024414,
      "epoch": 3.1783445811830093,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040087322938850867,
      "loss": 4.929,
      "mean_token_accuracy": 0.1879892259836197,
      "num_tokens": 70175563.0,
      "step": 32025
    },
    {
      "entropy": 5.291247129440308,
      "epoch": 3.178840809845177,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004008438596117408,
      "loss": 4.8541,
      "mean_token_accuracy": 0.18302326500415803,
      "num_tokens": 70186358.0,
      "step": 32030
    },
    {
      "entropy": 5.333696508407593,
      "epoch": 3.1793370385073443,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040081448671431635,
      "loss": 4.8966,
      "mean_token_accuracy": 0.18276099413633345,
      "num_tokens": 70196872.0,
      "step": 32035
    },
    {
      "entropy": 5.25417423248291,
      "epoch": 3.1798332671695118,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040078511069696377,
      "loss": 4.882,
      "mean_token_accuracy": 0.1808314546942711,
      "num_tokens": 70208136.0,
      "step": 32040
    },
    {
      "entropy": 5.3414403915405275,
      "epoch": 3.1803294958316792,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040075573156041146,
      "loss": 5.1355,
      "mean_token_accuracy": 0.17055115401744841,
      "num_tokens": 70219716.0,
      "step": 32045
    },
    {
      "entropy": 5.326167249679566,
      "epoch": 3.1808257244938467,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040072634930538796,
      "loss": 4.9342,
      "mean_token_accuracy": 0.18206951767206192,
      "num_tokens": 70230502.0,
      "step": 32050
    },
    {
      "entropy": 5.3879876136779785,
      "epoch": 3.181321953156014,
      "grad_norm": 1.0,
      "learning_rate": 0.00040069696393262196,
      "loss": 4.9397,
      "mean_token_accuracy": 0.1836100250482559,
      "num_tokens": 70241820.0,
      "step": 32055
    },
    {
      "entropy": 5.326156282424927,
      "epoch": 3.1818181818181817,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004006675754428421,
      "loss": 4.8648,
      "mean_token_accuracy": 0.18773905634880067,
      "num_tokens": 70252243.0,
      "step": 32060
    },
    {
      "entropy": 5.244920921325684,
      "epoch": 3.182314410480349,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040063818383677716,
      "loss": 4.7851,
      "mean_token_accuracy": 0.1910450652241707,
      "num_tokens": 70261835.0,
      "step": 32065
    },
    {
      "entropy": 5.322255182266235,
      "epoch": 3.182810639142517,
      "grad_norm": 1.03125,
      "learning_rate": 0.000400608789115156,
      "loss": 4.9619,
      "mean_token_accuracy": 0.18473488241434097,
      "num_tokens": 70273210.0,
      "step": 32070
    },
    {
      "entropy": 5.301050281524658,
      "epoch": 3.1833068678046845,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040057939127870754,
      "loss": 4.9244,
      "mean_token_accuracy": 0.17972665131092072,
      "num_tokens": 70283904.0,
      "step": 32075
    },
    {
      "entropy": 5.352426195144654,
      "epoch": 3.183803096466852,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004005499903281607,
      "loss": 4.9123,
      "mean_token_accuracy": 0.1904539719223976,
      "num_tokens": 70295199.0,
      "step": 32080
    },
    {
      "entropy": 5.2093555450439455,
      "epoch": 3.1842993251290195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040052058626424464,
      "loss": 4.7976,
      "mean_token_accuracy": 0.19386321306228638,
      "num_tokens": 70305418.0,
      "step": 32085
    },
    {
      "entropy": 5.2848457336425785,
      "epoch": 3.184795553791187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004004911790876886,
      "loss": 4.9785,
      "mean_token_accuracy": 0.18507099747657776,
      "num_tokens": 70316131.0,
      "step": 32090
    },
    {
      "entropy": 5.298650503158569,
      "epoch": 3.1852917824533544,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040046176879922156,
      "loss": 4.8289,
      "mean_token_accuracy": 0.18625362664461137,
      "num_tokens": 70326804.0,
      "step": 32095
    },
    {
      "entropy": 5.435314607620239,
      "epoch": 3.185788011115522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040043235539957315,
      "loss": 5.041,
      "mean_token_accuracy": 0.17425031661987306,
      "num_tokens": 70338125.0,
      "step": 32100
    },
    {
      "entropy": 5.319746398925782,
      "epoch": 3.1862842397776894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004004029388894726,
      "loss": 4.8776,
      "mean_token_accuracy": 0.18766166418790817,
      "num_tokens": 70348614.0,
      "step": 32105
    },
    {
      "entropy": 5.261679220199585,
      "epoch": 3.1867804684398573,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004003735192696494,
      "loss": 4.9494,
      "mean_token_accuracy": 0.18025965839624405,
      "num_tokens": 70358735.0,
      "step": 32110
    },
    {
      "entropy": 5.222465705871582,
      "epoch": 3.1872766971020248,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000400344096540833,
      "loss": 4.7979,
      "mean_token_accuracy": 0.19682619124650955,
      "num_tokens": 70369448.0,
      "step": 32115
    },
    {
      "entropy": 5.270518016815186,
      "epoch": 3.1877729257641922,
      "grad_norm": 1.0,
      "learning_rate": 0.00040031467070375315,
      "loss": 4.8682,
      "mean_token_accuracy": 0.1890062853693962,
      "num_tokens": 70379462.0,
      "step": 32120
    },
    {
      "entropy": 5.285385417938232,
      "epoch": 3.1882691544263597,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004002852417591395,
      "loss": 4.906,
      "mean_token_accuracy": 0.1813950464129448,
      "num_tokens": 70389283.0,
      "step": 32125
    },
    {
      "entropy": 5.298059463500977,
      "epoch": 3.188765383088527,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004002558097077218,
      "loss": 4.9429,
      "mean_token_accuracy": 0.1892665892839432,
      "num_tokens": 70400058.0,
      "step": 32130
    },
    {
      "entropy": 5.238427686691284,
      "epoch": 3.1892616117506947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004002263745502299,
      "loss": 4.806,
      "mean_token_accuracy": 0.19291511476039885,
      "num_tokens": 70411614.0,
      "step": 32135
    },
    {
      "entropy": 5.265917253494263,
      "epoch": 3.189757840412862,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040019693628739377,
      "loss": 4.8603,
      "mean_token_accuracy": 0.18478748053312302,
      "num_tokens": 70422404.0,
      "step": 32140
    },
    {
      "entropy": 5.352966690063477,
      "epoch": 3.1902540690750296,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004001674949199434,
      "loss": 4.9459,
      "mean_token_accuracy": 0.1882609486579895,
      "num_tokens": 70432219.0,
      "step": 32145
    },
    {
      "entropy": 5.293938016891479,
      "epoch": 3.190750297737197,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004001380504486089,
      "loss": 4.9127,
      "mean_token_accuracy": 0.18344465345144273,
      "num_tokens": 70443129.0,
      "step": 32150
    },
    {
      "entropy": 5.276419353485108,
      "epoch": 3.191246526399365,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004001086028741204,
      "loss": 4.9274,
      "mean_token_accuracy": 0.19008594751358032,
      "num_tokens": 70453225.0,
      "step": 32155
    },
    {
      "entropy": 5.245329284667969,
      "epoch": 3.1917427550615325,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004000791521972081,
      "loss": 4.9119,
      "mean_token_accuracy": 0.1857820063829422,
      "num_tokens": 70464909.0,
      "step": 32160
    },
    {
      "entropy": 5.319140815734864,
      "epoch": 3.1922389837237,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040004969841860233,
      "loss": 4.9035,
      "mean_token_accuracy": 0.18811959773302078,
      "num_tokens": 70475211.0,
      "step": 32165
    },
    {
      "entropy": 5.4211790561676025,
      "epoch": 3.1927352123858674,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040002024153903347,
      "loss": 4.9216,
      "mean_token_accuracy": 0.18732904642820358,
      "num_tokens": 70485603.0,
      "step": 32170
    },
    {
      "entropy": 5.282697486877441,
      "epoch": 3.193231441048035,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003999907815592321,
      "loss": 4.8264,
      "mean_token_accuracy": 0.19308691769838332,
      "num_tokens": 70497163.0,
      "step": 32175
    },
    {
      "entropy": 5.242024660110474,
      "epoch": 3.1937276697102024,
      "grad_norm": 1.0,
      "learning_rate": 0.00039996131847992863,
      "loss": 4.8636,
      "mean_token_accuracy": 0.1939951777458191,
      "num_tokens": 70509323.0,
      "step": 32180
    },
    {
      "entropy": 5.318041896820068,
      "epoch": 3.19422389837237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003999318523018537,
      "loss": 4.9997,
      "mean_token_accuracy": 0.17877112179994584,
      "num_tokens": 70520767.0,
      "step": 32185
    },
    {
      "entropy": 5.271632385253906,
      "epoch": 3.1947201270345373,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003999023830257382,
      "loss": 4.8447,
      "mean_token_accuracy": 0.1862033873796463,
      "num_tokens": 70531671.0,
      "step": 32190
    },
    {
      "entropy": 5.2948112964630125,
      "epoch": 3.1952163556967053,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003998729106523126,
      "loss": 4.8866,
      "mean_token_accuracy": 0.1861721843481064,
      "num_tokens": 70542309.0,
      "step": 32195
    },
    {
      "entropy": 5.356155157089233,
      "epoch": 3.1957125843588727,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003998434351823078,
      "loss": 4.969,
      "mean_token_accuracy": 0.18209062665700912,
      "num_tokens": 70553003.0,
      "step": 32200
    },
    {
      "entropy": 5.294030714035034,
      "epoch": 3.19620881302104,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000399813956616455,
      "loss": 4.896,
      "mean_token_accuracy": 0.18621761798858644,
      "num_tokens": 70564161.0,
      "step": 32205
    },
    {
      "entropy": 5.319119787216186,
      "epoch": 3.1967050416832077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039978447495548497,
      "loss": 4.8717,
      "mean_token_accuracy": 0.18912599980831146,
      "num_tokens": 70575891.0,
      "step": 32210
    },
    {
      "entropy": 5.360777044296265,
      "epoch": 3.197201270345375,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039975499020012886,
      "loss": 5.029,
      "mean_token_accuracy": 0.18415554314851762,
      "num_tokens": 70586691.0,
      "step": 32215
    },
    {
      "entropy": 5.30278754234314,
      "epoch": 3.1976974990075426,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039972550235111776,
      "loss": 4.8242,
      "mean_token_accuracy": 0.19209970235824586,
      "num_tokens": 70597423.0,
      "step": 32220
    },
    {
      "entropy": 5.302385902404785,
      "epoch": 3.19819372766971,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000399696011409183,
      "loss": 4.9151,
      "mean_token_accuracy": 0.1868718758225441,
      "num_tokens": 70609828.0,
      "step": 32225
    },
    {
      "entropy": 5.210149669647217,
      "epoch": 3.1986899563318776,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039966651737505585,
      "loss": 4.8224,
      "mean_token_accuracy": 0.1879908874630928,
      "num_tokens": 70620480.0,
      "step": 32230
    },
    {
      "entropy": 5.286655473709106,
      "epoch": 3.199186184994045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003996370202494677,
      "loss": 4.8889,
      "mean_token_accuracy": 0.19381917119026185,
      "num_tokens": 70630045.0,
      "step": 32235
    },
    {
      "entropy": 5.259391355514526,
      "epoch": 3.199682413656213,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039960752003315,
      "loss": 4.7963,
      "mean_token_accuracy": 0.19661949425935746,
      "num_tokens": 70640142.0,
      "step": 32240
    },
    {
      "entropy": 5.287434244155884,
      "epoch": 3.2001786423183805,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003995780167268343,
      "loss": 4.8859,
      "mean_token_accuracy": 0.18667418509721756,
      "num_tokens": 70652062.0,
      "step": 32245
    },
    {
      "entropy": 5.292778062820434,
      "epoch": 3.200674870980548,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003995485103312523,
      "loss": 4.9159,
      "mean_token_accuracy": 0.18896034210920334,
      "num_tokens": 70666187.0,
      "step": 32250
    },
    {
      "entropy": 5.349783992767334,
      "epoch": 3.2011710996427154,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039951900084713555,
      "loss": 4.9638,
      "mean_token_accuracy": 0.1850681021809578,
      "num_tokens": 70677432.0,
      "step": 32255
    },
    {
      "entropy": 5.29609899520874,
      "epoch": 3.201667328304883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003994894882752159,
      "loss": 4.8432,
      "mean_token_accuracy": 0.19101462066173552,
      "num_tokens": 70686963.0,
      "step": 32260
    },
    {
      "entropy": 5.392936420440674,
      "epoch": 3.2021635569670504,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039945997261622526,
      "loss": 5.056,
      "mean_token_accuracy": 0.17277976274490356,
      "num_tokens": 70698345.0,
      "step": 32265
    },
    {
      "entropy": 5.297130966186524,
      "epoch": 3.202659785629218,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003994304538708954,
      "loss": 4.8921,
      "mean_token_accuracy": 0.1899528369307518,
      "num_tokens": 70709695.0,
      "step": 32270
    },
    {
      "entropy": 5.255371236801148,
      "epoch": 3.2031560142913853,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003994009320399584,
      "loss": 4.816,
      "mean_token_accuracy": 0.19080153852701187,
      "num_tokens": 70720843.0,
      "step": 32275
    },
    {
      "entropy": 5.236575031280518,
      "epoch": 3.203652242953553,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039937140712414644,
      "loss": 4.9418,
      "mean_token_accuracy": 0.1883989840745926,
      "num_tokens": 70731886.0,
      "step": 32280
    },
    {
      "entropy": 5.291489839553833,
      "epoch": 3.2041484716157207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003993418791241915,
      "loss": 4.9062,
      "mean_token_accuracy": 0.1923538014292717,
      "num_tokens": 70741838.0,
      "step": 32285
    },
    {
      "entropy": 5.330525588989258,
      "epoch": 3.204644700277888,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003993123480408259,
      "loss": 4.8293,
      "mean_token_accuracy": 0.19658887833356858,
      "num_tokens": 70751454.0,
      "step": 32290
    },
    {
      "entropy": 5.339622831344604,
      "epoch": 3.2051409289400556,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039928281387478207,
      "loss": 4.9189,
      "mean_token_accuracy": 0.18177811354398726,
      "num_tokens": 70762294.0,
      "step": 32295
    },
    {
      "entropy": 5.249457502365113,
      "epoch": 3.205637157602223,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003992532766267921,
      "loss": 4.8751,
      "mean_token_accuracy": 0.19194059520959855,
      "num_tokens": 70773186.0,
      "step": 32300
    },
    {
      "entropy": 5.311899900436401,
      "epoch": 3.2061333862643906,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003992237362975887,
      "loss": 4.9121,
      "mean_token_accuracy": 0.1816537633538246,
      "num_tokens": 70783984.0,
      "step": 32305
    },
    {
      "entropy": 5.294994735717774,
      "epoch": 3.206629614926558,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039919419288790423,
      "loss": 4.8818,
      "mean_token_accuracy": 0.18718979954719545,
      "num_tokens": 70795779.0,
      "step": 32310
    },
    {
      "entropy": 5.316264581680298,
      "epoch": 3.2071258435887255,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039916464639847145,
      "loss": 4.8771,
      "mean_token_accuracy": 0.18468999713659287,
      "num_tokens": 70807247.0,
      "step": 32315
    },
    {
      "entropy": 5.323252010345459,
      "epoch": 3.207622072250893,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003991350968300229,
      "loss": 4.9083,
      "mean_token_accuracy": 0.1844802737236023,
      "num_tokens": 70818507.0,
      "step": 32320
    },
    {
      "entropy": 5.285073375701904,
      "epoch": 3.208118300913061,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003991055441832916,
      "loss": 4.9218,
      "mean_token_accuracy": 0.1823875516653061,
      "num_tokens": 70830055.0,
      "step": 32325
    },
    {
      "entropy": 5.353236532211303,
      "epoch": 3.2086145295752284,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039907598845901016,
      "loss": 5.0049,
      "mean_token_accuracy": 0.1835538238286972,
      "num_tokens": 70841052.0,
      "step": 32330
    },
    {
      "entropy": 5.37023286819458,
      "epoch": 3.209110758237396,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003990464296579114,
      "loss": 4.9396,
      "mean_token_accuracy": 0.190088252723217,
      "num_tokens": 70852148.0,
      "step": 32335
    },
    {
      "entropy": 5.275377893447876,
      "epoch": 3.2096069868995634,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003990168677807287,
      "loss": 4.935,
      "mean_token_accuracy": 0.18183023780584334,
      "num_tokens": 70863998.0,
      "step": 32340
    },
    {
      "entropy": 5.306494903564453,
      "epoch": 3.210103215561731,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003989873028281948,
      "loss": 4.9947,
      "mean_token_accuracy": 0.18065525740385055,
      "num_tokens": 70874013.0,
      "step": 32345
    },
    {
      "entropy": 5.291953992843628,
      "epoch": 3.2105994442238983,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000398957734801043,
      "loss": 4.8154,
      "mean_token_accuracy": 0.19465869814157485,
      "num_tokens": 70885600.0,
      "step": 32350
    },
    {
      "entropy": 5.215686082839966,
      "epoch": 3.211095672886066,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039892816370000657,
      "loss": 4.7782,
      "mean_token_accuracy": 0.20246593207120894,
      "num_tokens": 70896447.0,
      "step": 32355
    },
    {
      "entropy": 5.343898057937622,
      "epoch": 3.2115919015482333,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003988985895258186,
      "loss": 4.9684,
      "mean_token_accuracy": 0.184929521381855,
      "num_tokens": 70907558.0,
      "step": 32360
    },
    {
      "entropy": 5.305607080459595,
      "epoch": 3.212088130210401,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039886901227921263,
      "loss": 4.9149,
      "mean_token_accuracy": 0.18392559438943862,
      "num_tokens": 70919687.0,
      "step": 32365
    },
    {
      "entropy": 5.282931709289551,
      "epoch": 3.2125843588725687,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003988394319609221,
      "loss": 4.8588,
      "mean_token_accuracy": 0.19216416031122208,
      "num_tokens": 70930417.0,
      "step": 32370
    },
    {
      "entropy": 5.266162395477295,
      "epoch": 3.213080587534736,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003988098485716804,
      "loss": 4.8658,
      "mean_token_accuracy": 0.1890067368745804,
      "num_tokens": 70941721.0,
      "step": 32375
    },
    {
      "entropy": 5.255045557022095,
      "epoch": 3.2135768161969036,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039878026211222136,
      "loss": 4.8479,
      "mean_token_accuracy": 0.19330516904592515,
      "num_tokens": 70952616.0,
      "step": 32380
    },
    {
      "entropy": 5.228937673568725,
      "epoch": 3.214073044859071,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039875067258327853,
      "loss": 4.8495,
      "mean_token_accuracy": 0.1878674879670143,
      "num_tokens": 70963236.0,
      "step": 32385
    },
    {
      "entropy": 5.322563171386719,
      "epoch": 3.2145692735212386,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003987210799855856,
      "loss": 4.9112,
      "mean_token_accuracy": 0.18558530658483505,
      "num_tokens": 70973376.0,
      "step": 32390
    },
    {
      "entropy": 5.265290975570679,
      "epoch": 3.215065502183406,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039869148431987653,
      "loss": 4.9036,
      "mean_token_accuracy": 0.18549921810626985,
      "num_tokens": 70984937.0,
      "step": 32395
    },
    {
      "entropy": 5.377542400360108,
      "epoch": 3.2155617308455735,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003986618855868852,
      "loss": 4.9764,
      "mean_token_accuracy": 0.1872401863336563,
      "num_tokens": 70997023.0,
      "step": 32400
    },
    {
      "entropy": 5.302580785751343,
      "epoch": 3.216057959507741,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003986322837873455,
      "loss": 4.8365,
      "mean_token_accuracy": 0.19109082221984863,
      "num_tokens": 71008304.0,
      "step": 32405
    },
    {
      "entropy": 5.249147987365722,
      "epoch": 3.216554188169909,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039860267892199167,
      "loss": 4.9033,
      "mean_token_accuracy": 0.18948803395032882,
      "num_tokens": 71020160.0,
      "step": 32410
    },
    {
      "entropy": 5.325675582885742,
      "epoch": 3.2170504168320764,
      "grad_norm": 1.125,
      "learning_rate": 0.00039857307099155773,
      "loss": 4.939,
      "mean_token_accuracy": 0.19227443188428878,
      "num_tokens": 71031701.0,
      "step": 32415
    },
    {
      "entropy": 5.330224847793579,
      "epoch": 3.217546645494244,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039854345999677777,
      "loss": 4.9711,
      "mean_token_accuracy": 0.1864713490009308,
      "num_tokens": 71041955.0,
      "step": 32420
    },
    {
      "entropy": 5.336421775817871,
      "epoch": 3.2180428741564113,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039851384593838637,
      "loss": 4.9756,
      "mean_token_accuracy": 0.18422434180974961,
      "num_tokens": 71052884.0,
      "step": 32425
    },
    {
      "entropy": 5.349724340438843,
      "epoch": 3.218539102818579,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039848422881711765,
      "loss": 4.8784,
      "mean_token_accuracy": 0.19071923643350602,
      "num_tokens": 71065091.0,
      "step": 32430
    },
    {
      "entropy": 5.328426361083984,
      "epoch": 3.2190353314807463,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003984546086337061,
      "loss": 4.9589,
      "mean_token_accuracy": 0.1861597403883934,
      "num_tokens": 71075823.0,
      "step": 32435
    },
    {
      "entropy": 5.330241012573242,
      "epoch": 3.2195315601429138,
      "grad_norm": 1.0,
      "learning_rate": 0.00039842498538888646,
      "loss": 4.9058,
      "mean_token_accuracy": 0.19076012670993805,
      "num_tokens": 71087408.0,
      "step": 32440
    },
    {
      "entropy": 5.337849712371826,
      "epoch": 3.2200277888050812,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039839535908339297,
      "loss": 4.8871,
      "mean_token_accuracy": 0.18959749788045882,
      "num_tokens": 71097474.0,
      "step": 32445
    },
    {
      "entropy": 5.260224437713623,
      "epoch": 3.2205240174672487,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039836572971796045,
      "loss": 4.8448,
      "mean_token_accuracy": 0.1951851263642311,
      "num_tokens": 71107064.0,
      "step": 32450
    },
    {
      "entropy": 5.246933031082153,
      "epoch": 3.2210202461294166,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003983360972933239,
      "loss": 4.9023,
      "mean_token_accuracy": 0.19033001810312272,
      "num_tokens": 71118233.0,
      "step": 32455
    },
    {
      "entropy": 5.252711629867553,
      "epoch": 3.221516474791584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003983064618102177,
      "loss": 4.8163,
      "mean_token_accuracy": 0.19805368781089783,
      "num_tokens": 71129922.0,
      "step": 32460
    },
    {
      "entropy": 5.277643346786499,
      "epoch": 3.2220127034537516,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003982768232693769,
      "loss": 4.9171,
      "mean_token_accuracy": 0.1868602976202965,
      "num_tokens": 71139073.0,
      "step": 32465
    },
    {
      "entropy": 5.268294811248779,
      "epoch": 3.222508932115919,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039824718167153666,
      "loss": 4.8208,
      "mean_token_accuracy": 0.1867639273405075,
      "num_tokens": 71152240.0,
      "step": 32470
    },
    {
      "entropy": 5.29614028930664,
      "epoch": 3.2230051607780865,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003982175370174319,
      "loss": 4.8966,
      "mean_token_accuracy": 0.18479514718055726,
      "num_tokens": 71163001.0,
      "step": 32475
    },
    {
      "entropy": 5.290539932250977,
      "epoch": 3.223501389440254,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003981878893077976,
      "loss": 4.8683,
      "mean_token_accuracy": 0.1902368351817131,
      "num_tokens": 71172811.0,
      "step": 32480
    },
    {
      "entropy": 5.230042600631714,
      "epoch": 3.2239976181024215,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039815823854336916,
      "loss": 4.7948,
      "mean_token_accuracy": 0.20089155286550522,
      "num_tokens": 71182276.0,
      "step": 32485
    },
    {
      "entropy": 5.216751623153686,
      "epoch": 3.224493846764589,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003981285847248818,
      "loss": 4.7875,
      "mean_token_accuracy": 0.19634587019681932,
      "num_tokens": 71193154.0,
      "step": 32490
    },
    {
      "entropy": 5.307789850234985,
      "epoch": 3.2249900754267564,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003980989278530709,
      "loss": 4.9476,
      "mean_token_accuracy": 0.1814580336213112,
      "num_tokens": 71204394.0,
      "step": 32495
    },
    {
      "entropy": 5.241375684738159,
      "epoch": 3.2254863040889243,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003980692679286718,
      "loss": 4.7869,
      "mean_token_accuracy": 0.19192990511655808,
      "num_tokens": 71215311.0,
      "step": 32500
    },
    {
      "entropy": 5.277870464324951,
      "epoch": 3.225982532751092,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039803960495241996,
      "loss": 4.9027,
      "mean_token_accuracy": 0.19418093264102937,
      "num_tokens": 71225800.0,
      "step": 32505
    },
    {
      "entropy": 5.314246749877929,
      "epoch": 3.2264787614132593,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003980099389250511,
      "loss": 4.8719,
      "mean_token_accuracy": 0.18313958644866943,
      "num_tokens": 71237846.0,
      "step": 32510
    },
    {
      "entropy": 5.283829975128174,
      "epoch": 3.2269749900754268,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039798026984730084,
      "loss": 4.892,
      "mean_token_accuracy": 0.18974294662475585,
      "num_tokens": 71248221.0,
      "step": 32515
    },
    {
      "entropy": 5.275758552551269,
      "epoch": 3.2274712187375942,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003979505977199049,
      "loss": 4.9084,
      "mean_token_accuracy": 0.1852722570300102,
      "num_tokens": 71259950.0,
      "step": 32520
    },
    {
      "entropy": 5.3022924900054935,
      "epoch": 3.2279674473997617,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039792092254359903,
      "loss": 4.9383,
      "mean_token_accuracy": 0.18661603778600694,
      "num_tokens": 71270541.0,
      "step": 32525
    },
    {
      "entropy": 5.30677490234375,
      "epoch": 3.228463676061929,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039789124431911923,
      "loss": 4.9095,
      "mean_token_accuracy": 0.18427392095327377,
      "num_tokens": 71280545.0,
      "step": 32530
    },
    {
      "entropy": 5.371171855926514,
      "epoch": 3.2289599047240967,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039786156304720126,
      "loss": 4.9762,
      "mean_token_accuracy": 0.18603410869836806,
      "num_tokens": 71292368.0,
      "step": 32535
    },
    {
      "entropy": 5.34957184791565,
      "epoch": 3.2294561333862646,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003978318787285814,
      "loss": 4.8751,
      "mean_token_accuracy": 0.18716879040002823,
      "num_tokens": 71302570.0,
      "step": 32540
    },
    {
      "entropy": 5.27445421218872,
      "epoch": 3.229952362048432,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039780219136399556,
      "loss": 4.9103,
      "mean_token_accuracy": 0.1919171392917633,
      "num_tokens": 71312629.0,
      "step": 32545
    },
    {
      "entropy": 5.3260573387146,
      "epoch": 3.2304485907105995,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039777250095418,
      "loss": 4.9668,
      "mean_token_accuracy": 0.18379939645528792,
      "num_tokens": 71324313.0,
      "step": 32550
    },
    {
      "entropy": 5.351593017578125,
      "epoch": 3.230944819372767,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003977428074998709,
      "loss": 4.9839,
      "mean_token_accuracy": 0.1848885625600815,
      "num_tokens": 71336136.0,
      "step": 32555
    },
    {
      "entropy": 5.249500703811646,
      "epoch": 3.2314410480349345,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039771311100180476,
      "loss": 4.8344,
      "mean_token_accuracy": 0.19508963376283645,
      "num_tokens": 71347267.0,
      "step": 32560
    },
    {
      "entropy": 5.24223256111145,
      "epoch": 3.231937276697102,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039768341146071784,
      "loss": 4.8592,
      "mean_token_accuracy": 0.1989669010043144,
      "num_tokens": 71357268.0,
      "step": 32565
    },
    {
      "entropy": 5.399132204055786,
      "epoch": 3.2324335053592694,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003976537088773467,
      "loss": 4.9854,
      "mean_token_accuracy": 0.18006051629781722,
      "num_tokens": 71368365.0,
      "step": 32570
    },
    {
      "entropy": 5.390395784378052,
      "epoch": 3.232929734021437,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039762400325242785,
      "loss": 4.9845,
      "mean_token_accuracy": 0.18089616000652314,
      "num_tokens": 71379623.0,
      "step": 32575
    },
    {
      "entropy": 5.2479291439056395,
      "epoch": 3.2334259626836044,
      "grad_norm": 1.125,
      "learning_rate": 0.00039759429458669805,
      "loss": 4.8865,
      "mean_token_accuracy": 0.19214936941862107,
      "num_tokens": 71388522.0,
      "step": 32580
    },
    {
      "entropy": 5.237315320968628,
      "epoch": 3.2339221913457723,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039756458288089377,
      "loss": 4.8623,
      "mean_token_accuracy": 0.19166333377361297,
      "num_tokens": 71399127.0,
      "step": 32585
    },
    {
      "entropy": 5.429157066345215,
      "epoch": 3.2344184200079398,
      "grad_norm": 1.09375,
      "learning_rate": 0.000397534868135752,
      "loss": 4.9963,
      "mean_token_accuracy": 0.18218237906694412,
      "num_tokens": 71410780.0,
      "step": 32590
    },
    {
      "entropy": 5.374507284164428,
      "epoch": 3.2349146486701073,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039750515035200947,
      "loss": 5.019,
      "mean_token_accuracy": 0.17325063943862914,
      "num_tokens": 71422887.0,
      "step": 32595
    },
    {
      "entropy": 5.2980124950408936,
      "epoch": 3.2354108773322747,
      "grad_norm": 1.0,
      "learning_rate": 0.00039747542953040327,
      "loss": 4.9202,
      "mean_token_accuracy": 0.19216493666172027,
      "num_tokens": 71432525.0,
      "step": 32600
    },
    {
      "entropy": 5.273099374771118,
      "epoch": 3.235907105994442,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003974457056716703,
      "loss": 4.9313,
      "mean_token_accuracy": 0.18890963941812516,
      "num_tokens": 71444802.0,
      "step": 32605
    },
    {
      "entropy": 5.229762840270996,
      "epoch": 3.2364033346566097,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003974159787765477,
      "loss": 4.8284,
      "mean_token_accuracy": 0.1904064193367958,
      "num_tokens": 71455978.0,
      "step": 32610
    },
    {
      "entropy": 5.221776008605957,
      "epoch": 3.236899563318777,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003973862488457726,
      "loss": 4.7848,
      "mean_token_accuracy": 0.18940970450639724,
      "num_tokens": 71466981.0,
      "step": 32615
    },
    {
      "entropy": 5.317994117736816,
      "epoch": 3.2373957919809446,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003973565158800822,
      "loss": 4.9228,
      "mean_token_accuracy": 0.18424446731805802,
      "num_tokens": 71478856.0,
      "step": 32620
    },
    {
      "entropy": 5.35066499710083,
      "epoch": 3.2378920206431125,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039732677988021394,
      "loss": 4.9559,
      "mean_token_accuracy": 0.18485590666532517,
      "num_tokens": 71489449.0,
      "step": 32625
    },
    {
      "entropy": 5.291810417175293,
      "epoch": 3.23838824930528,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003972970408469051,
      "loss": 4.9364,
      "mean_token_accuracy": 0.1821162983775139,
      "num_tokens": 71499687.0,
      "step": 32630
    },
    {
      "entropy": 5.372183418273925,
      "epoch": 3.2388844779674475,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003972672987808932,
      "loss": 4.8968,
      "mean_token_accuracy": 0.18742562234401702,
      "num_tokens": 71510087.0,
      "step": 32635
    },
    {
      "entropy": 5.311279821395874,
      "epoch": 3.239380706629615,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039723755368291564,
      "loss": 4.873,
      "mean_token_accuracy": 0.18713998645544053,
      "num_tokens": 71520504.0,
      "step": 32640
    },
    {
      "entropy": 5.331241178512573,
      "epoch": 3.2398769352917824,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003972078055537102,
      "loss": 4.9488,
      "mean_token_accuracy": 0.17950180023908616,
      "num_tokens": 71532342.0,
      "step": 32645
    },
    {
      "entropy": 5.361578893661499,
      "epoch": 3.24037316395395,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003971780543940146,
      "loss": 4.9159,
      "mean_token_accuracy": 0.18106812238693237,
      "num_tokens": 71544523.0,
      "step": 32650
    },
    {
      "entropy": 5.293410110473633,
      "epoch": 3.2408693926161174,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003971483002045665,
      "loss": 4.9064,
      "mean_token_accuracy": 0.18776200711727142,
      "num_tokens": 71555162.0,
      "step": 32655
    },
    {
      "entropy": 5.265429592132568,
      "epoch": 3.241365621278285,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003971185429861037,
      "loss": 4.8905,
      "mean_token_accuracy": 0.19286435842514038,
      "num_tokens": 71566407.0,
      "step": 32660
    },
    {
      "entropy": 5.334396409988403,
      "epoch": 3.2418618499404523,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039708878273936424,
      "loss": 4.9177,
      "mean_token_accuracy": 0.18828495889902114,
      "num_tokens": 71577328.0,
      "step": 32665
    },
    {
      "entropy": 5.391731643676758,
      "epoch": 3.2423580786026203,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003970590194650859,
      "loss": 5.0382,
      "mean_token_accuracy": 0.1783861756324768,
      "num_tokens": 71587716.0,
      "step": 32670
    },
    {
      "entropy": 5.279687070846558,
      "epoch": 3.2428543072647877,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003970292531640071,
      "loss": 4.8664,
      "mean_token_accuracy": 0.1907387927174568,
      "num_tokens": 71599575.0,
      "step": 32675
    },
    {
      "entropy": 5.29763765335083,
      "epoch": 3.243350535926955,
      "grad_norm": 0.859375,
      "learning_rate": 0.00039699948383686563,
      "loss": 4.9417,
      "mean_token_accuracy": 0.1867110699415207,
      "num_tokens": 71610925.0,
      "step": 32680
    },
    {
      "entropy": 5.288529253005981,
      "epoch": 3.2438467645891227,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003969697114843998,
      "loss": 4.8794,
      "mean_token_accuracy": 0.1924777388572693,
      "num_tokens": 71620250.0,
      "step": 32685
    },
    {
      "entropy": 5.338484764099121,
      "epoch": 3.24434299325129,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039693993610734807,
      "loss": 4.9997,
      "mean_token_accuracy": 0.1817205458879471,
      "num_tokens": 71632633.0,
      "step": 32690
    },
    {
      "entropy": 5.342201900482178,
      "epoch": 3.2448392219134576,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039691015770644875,
      "loss": 4.9824,
      "mean_token_accuracy": 0.18010225296020507,
      "num_tokens": 71644031.0,
      "step": 32695
    },
    {
      "entropy": 5.349267053604126,
      "epoch": 3.245335450575625,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039688037628244,
      "loss": 4.9417,
      "mean_token_accuracy": 0.19769650101661682,
      "num_tokens": 71654785.0,
      "step": 32700
    },
    {
      "entropy": 5.347609519958496,
      "epoch": 3.2458316792377926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039685059183606065,
      "loss": 4.9036,
      "mean_token_accuracy": 0.19249048233032226,
      "num_tokens": 71667014.0,
      "step": 32705
    },
    {
      "entropy": 5.314712619781494,
      "epoch": 3.2463279078999605,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039682080436804926,
      "loss": 4.9242,
      "mean_token_accuracy": 0.19081036299467086,
      "num_tokens": 71677157.0,
      "step": 32710
    },
    {
      "entropy": 5.349314117431641,
      "epoch": 3.246824136562128,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003967910138791442,
      "loss": 4.9636,
      "mean_token_accuracy": 0.18291127681732178,
      "num_tokens": 71688542.0,
      "step": 32715
    },
    {
      "entropy": 5.286740922927857,
      "epoch": 3.2473203652242955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039676122037008473,
      "loss": 4.8976,
      "mean_token_accuracy": 0.18938620537519454,
      "num_tokens": 71699364.0,
      "step": 32720
    },
    {
      "entropy": 5.325748443603516,
      "epoch": 3.247816593886463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003967314238416092,
      "loss": 4.9286,
      "mean_token_accuracy": 0.1822100907564163,
      "num_tokens": 71710248.0,
      "step": 32725
    },
    {
      "entropy": 5.313327884674072,
      "epoch": 3.2483128225486304,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003967016242944567,
      "loss": 4.9309,
      "mean_token_accuracy": 0.18233461380004884,
      "num_tokens": 71720616.0,
      "step": 32730
    },
    {
      "entropy": 5.4134032249450685,
      "epoch": 3.248809051210798,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003966718217293661,
      "loss": 4.9276,
      "mean_token_accuracy": 0.18610207587480546,
      "num_tokens": 71731536.0,
      "step": 32735
    },
    {
      "entropy": 5.316247606277466,
      "epoch": 3.2493052798729654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039664201614707657,
      "loss": 4.9395,
      "mean_token_accuracy": 0.19066007286310196,
      "num_tokens": 71742459.0,
      "step": 32740
    },
    {
      "entropy": 5.232972002029419,
      "epoch": 3.249801508535133,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003966122075483271,
      "loss": 4.912,
      "mean_token_accuracy": 0.18325185328722,
      "num_tokens": 71753611.0,
      "step": 32745
    },
    {
      "entropy": 5.3031334400177,
      "epoch": 3.2502977371973003,
      "grad_norm": 1.0,
      "learning_rate": 0.00039658239593385697,
      "loss": 4.8688,
      "mean_token_accuracy": 0.19244902580976486,
      "num_tokens": 71764924.0,
      "step": 32750
    },
    {
      "entropy": 5.295020818710327,
      "epoch": 3.2507939658594682,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039655258130440543,
      "loss": 4.9661,
      "mean_token_accuracy": 0.1715869665145874,
      "num_tokens": 71776078.0,
      "step": 32755
    },
    {
      "entropy": 5.327985382080078,
      "epoch": 3.2512901945216357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003965227636607117,
      "loss": 4.8879,
      "mean_token_accuracy": 0.18446722626686096,
      "num_tokens": 71785988.0,
      "step": 32760
    },
    {
      "entropy": 5.265541315078735,
      "epoch": 3.251786423183803,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039649294300351547,
      "loss": 4.8968,
      "mean_token_accuracy": 0.1877792403101921,
      "num_tokens": 71796382.0,
      "step": 32765
    },
    {
      "entropy": 5.275164651870727,
      "epoch": 3.2522826518459707,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003964631193335559,
      "loss": 4.8662,
      "mean_token_accuracy": 0.1810305744409561,
      "num_tokens": 71806533.0,
      "step": 32770
    },
    {
      "entropy": 5.271871042251587,
      "epoch": 3.252778880508138,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039643329265157275,
      "loss": 4.9094,
      "mean_token_accuracy": 0.19145602881908416,
      "num_tokens": 71816893.0,
      "step": 32775
    },
    {
      "entropy": 5.348955345153809,
      "epoch": 3.2532751091703056,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039640346295830557,
      "loss": 4.9004,
      "mean_token_accuracy": 0.18465500473976135,
      "num_tokens": 71827366.0,
      "step": 32780
    },
    {
      "entropy": 5.334942293167114,
      "epoch": 3.253771337832473,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003963736302544942,
      "loss": 4.924,
      "mean_token_accuracy": 0.18169148415327072,
      "num_tokens": 71837896.0,
      "step": 32785
    },
    {
      "entropy": 5.259771013259888,
      "epoch": 3.2542675664946406,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039634379454087814,
      "loss": 4.9195,
      "mean_token_accuracy": 0.18379215747117997,
      "num_tokens": 71850583.0,
      "step": 32790
    },
    {
      "entropy": 5.301876068115234,
      "epoch": 3.2547637951568085,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003963139558181977,
      "loss": 4.9325,
      "mean_token_accuracy": 0.18715383857488632,
      "num_tokens": 71860392.0,
      "step": 32795
    },
    {
      "entropy": 5.298343372344971,
      "epoch": 3.255260023818976,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003962841140871924,
      "loss": 4.912,
      "mean_token_accuracy": 0.18986668139696122,
      "num_tokens": 71871323.0,
      "step": 32800
    },
    {
      "entropy": 5.256008529663086,
      "epoch": 3.2557562524811434,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003962542693486024,
      "loss": 4.8536,
      "mean_token_accuracy": 0.19435556083917618,
      "num_tokens": 71881612.0,
      "step": 32805
    },
    {
      "entropy": 5.260603237152099,
      "epoch": 3.256252481143311,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003962244216031679,
      "loss": 4.9678,
      "mean_token_accuracy": 0.18719144612550737,
      "num_tokens": 71893266.0,
      "step": 32810
    },
    {
      "entropy": 5.304356718063355,
      "epoch": 3.2567487098054784,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000396194570851629,
      "loss": 4.862,
      "mean_token_accuracy": 0.19254924207925797,
      "num_tokens": 71904978.0,
      "step": 32815
    },
    {
      "entropy": 5.389955186843872,
      "epoch": 3.257244938467646,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039616471709472573,
      "loss": 4.9666,
      "mean_token_accuracy": 0.1835562840104103,
      "num_tokens": 71915097.0,
      "step": 32820
    },
    {
      "entropy": 5.369780778884888,
      "epoch": 3.2577411671298133,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003961348603331987,
      "loss": 4.9014,
      "mean_token_accuracy": 0.1803993374109268,
      "num_tokens": 71924672.0,
      "step": 32825
    },
    {
      "entropy": 5.222688055038452,
      "epoch": 3.258237395791981,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00039610500056778817,
      "loss": 4.8132,
      "mean_token_accuracy": 0.19285855889320375,
      "num_tokens": 71936122.0,
      "step": 32830
    },
    {
      "entropy": 5.242972564697266,
      "epoch": 3.2587336244541483,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003960751377992344,
      "loss": 4.8239,
      "mean_token_accuracy": 0.1948391914367676,
      "num_tokens": 71946601.0,
      "step": 32835
    },
    {
      "entropy": 5.349831533432007,
      "epoch": 3.259229853116316,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039604527202827825,
      "loss": 4.943,
      "mean_token_accuracy": 0.18430111706256866,
      "num_tokens": 71958892.0,
      "step": 32840
    },
    {
      "entropy": 5.279014301300049,
      "epoch": 3.2597260817784837,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039601540325566014,
      "loss": 4.8625,
      "mean_token_accuracy": 0.18884371668100358,
      "num_tokens": 71969183.0,
      "step": 32845
    },
    {
      "entropy": 5.283637142181396,
      "epoch": 3.260222310440651,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003959855314821208,
      "loss": 4.8814,
      "mean_token_accuracy": 0.18943305611610411,
      "num_tokens": 71981115.0,
      "step": 32850
    },
    {
      "entropy": 5.350069904327393,
      "epoch": 3.2607185391028186,
      "grad_norm": 1.015625,
      "learning_rate": 0.000395955656708401,
      "loss": 4.9513,
      "mean_token_accuracy": 0.1887938067317009,
      "num_tokens": 71991677.0,
      "step": 32855
    },
    {
      "entropy": 5.337734842300415,
      "epoch": 3.261214767764986,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039592577893524155,
      "loss": 4.9361,
      "mean_token_accuracy": 0.19067416191101075,
      "num_tokens": 72003995.0,
      "step": 32860
    },
    {
      "entropy": 5.330810642242431,
      "epoch": 3.2617109964271536,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039589589816338335,
      "loss": 4.9078,
      "mean_token_accuracy": 0.18433529138565063,
      "num_tokens": 72014052.0,
      "step": 32865
    },
    {
      "entropy": 5.286349678039551,
      "epoch": 3.262207225089321,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003958660143935674,
      "loss": 4.8503,
      "mean_token_accuracy": 0.18489064872264863,
      "num_tokens": 72024756.0,
      "step": 32870
    },
    {
      "entropy": 5.227883291244507,
      "epoch": 3.2627034537514885,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003958361276265347,
      "loss": 4.8736,
      "mean_token_accuracy": 0.1887678548693657,
      "num_tokens": 72035883.0,
      "step": 32875
    },
    {
      "entropy": 5.307349348068238,
      "epoch": 3.2631996824136564,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039580623786302643,
      "loss": 4.8819,
      "mean_token_accuracy": 0.18579447865486146,
      "num_tokens": 72047296.0,
      "step": 32880
    },
    {
      "entropy": 5.30422191619873,
      "epoch": 3.263695911075824,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039577634510378374,
      "loss": 4.8876,
      "mean_token_accuracy": 0.18204347640275956,
      "num_tokens": 72060163.0,
      "step": 32885
    },
    {
      "entropy": 5.320245265960693,
      "epoch": 3.2641921397379914,
      "grad_norm": 1.046875,
      "learning_rate": 0.000395746449349548,
      "loss": 4.9422,
      "mean_token_accuracy": 0.18387205600738527,
      "num_tokens": 72070384.0,
      "step": 32890
    },
    {
      "entropy": 5.313709306716919,
      "epoch": 3.264688368400159,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003957165506010605,
      "loss": 4.8693,
      "mean_token_accuracy": 0.18741924017667771,
      "num_tokens": 72080997.0,
      "step": 32895
    },
    {
      "entropy": 5.353471040725708,
      "epoch": 3.2651845970623263,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003956866488590626,
      "loss": 4.9083,
      "mean_token_accuracy": 0.19338052123785018,
      "num_tokens": 72091161.0,
      "step": 32900
    },
    {
      "entropy": 5.259972810745239,
      "epoch": 3.265680825724494,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003956567441242959,
      "loss": 4.8796,
      "mean_token_accuracy": 0.18476131409406663,
      "num_tokens": 72101807.0,
      "step": 32905
    },
    {
      "entropy": 5.328732919692993,
      "epoch": 3.2661770543866613,
      "grad_norm": 0.96875,
      "learning_rate": 0.000395626836397502,
      "loss": 4.9618,
      "mean_token_accuracy": 0.1857716456055641,
      "num_tokens": 72112800.0,
      "step": 32910
    },
    {
      "entropy": 5.247724485397339,
      "epoch": 3.2666732830488288,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003955969256794224,
      "loss": 4.8691,
      "mean_token_accuracy": 0.187334106862545,
      "num_tokens": 72122391.0,
      "step": 32915
    },
    {
      "entropy": 5.277861022949219,
      "epoch": 3.2671695117109962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039556701197079896,
      "loss": 4.9156,
      "mean_token_accuracy": 0.18787066489458085,
      "num_tokens": 72133145.0,
      "step": 32920
    },
    {
      "entropy": 5.395589828491211,
      "epoch": 3.2676657403731637,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039553709527237344,
      "loss": 4.9599,
      "mean_token_accuracy": 0.1760472521185875,
      "num_tokens": 72146126.0,
      "step": 32925
    },
    {
      "entropy": 5.283221960067749,
      "epoch": 3.2681619690353316,
      "grad_norm": 1.0,
      "learning_rate": 0.0003955071755848876,
      "loss": 4.8522,
      "mean_token_accuracy": 0.19238603711128235,
      "num_tokens": 72157325.0,
      "step": 32930
    },
    {
      "entropy": 5.194116878509521,
      "epoch": 3.268658197697499,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039547725290908347,
      "loss": 4.7731,
      "mean_token_accuracy": 0.1930732861161232,
      "num_tokens": 72168601.0,
      "step": 32935
    },
    {
      "entropy": 5.244743824005127,
      "epoch": 3.2691544263596666,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00039544732724570313,
      "loss": 4.8539,
      "mean_token_accuracy": 0.19718263894319535,
      "num_tokens": 72179845.0,
      "step": 32940
    },
    {
      "entropy": 5.236837530136109,
      "epoch": 3.269650655021834,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003954173985954887,
      "loss": 4.779,
      "mean_token_accuracy": 0.19067852199077606,
      "num_tokens": 72191135.0,
      "step": 32945
    },
    {
      "entropy": 5.262376022338867,
      "epoch": 3.2701468836840015,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003953874669591822,
      "loss": 4.888,
      "mean_token_accuracy": 0.19074117243289948,
      "num_tokens": 72201551.0,
      "step": 32950
    },
    {
      "entropy": 5.277746200561523,
      "epoch": 3.270643112346169,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039535753233752583,
      "loss": 4.8872,
      "mean_token_accuracy": 0.18876436948776246,
      "num_tokens": 72211983.0,
      "step": 32955
    },
    {
      "entropy": 5.3416118144989015,
      "epoch": 3.2711393410083365,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003953275947312621,
      "loss": 4.9384,
      "mean_token_accuracy": 0.17892933040857315,
      "num_tokens": 72223374.0,
      "step": 32960
    },
    {
      "entropy": 5.212268877029419,
      "epoch": 3.2716355696705044,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003952976541411332,
      "loss": 4.8886,
      "mean_token_accuracy": 0.19426627457141876,
      "num_tokens": 72233878.0,
      "step": 32965
    },
    {
      "entropy": 5.259953451156616,
      "epoch": 3.272131798332672,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039526771056788186,
      "loss": 4.9619,
      "mean_token_accuracy": 0.1891522839665413,
      "num_tokens": 72246157.0,
      "step": 32970
    },
    {
      "entropy": 5.3317197322845455,
      "epoch": 3.2726280269948393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039523776401225043,
      "loss": 4.8799,
      "mean_token_accuracy": 0.19663195461034774,
      "num_tokens": 72256805.0,
      "step": 32975
    },
    {
      "entropy": 5.313364791870117,
      "epoch": 3.273124255657007,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039520781447498145,
      "loss": 4.902,
      "mean_token_accuracy": 0.18113366067409514,
      "num_tokens": 72267026.0,
      "step": 32980
    },
    {
      "entropy": 5.283870220184326,
      "epoch": 3.2736204843191743,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003951778619568177,
      "loss": 4.8885,
      "mean_token_accuracy": 0.1908041313290596,
      "num_tokens": 72277989.0,
      "step": 32985
    },
    {
      "entropy": 5.218806505203247,
      "epoch": 3.2741167129813418,
      "grad_norm": 1.078125,
      "learning_rate": 0.000395147906458502,
      "loss": 4.8271,
      "mean_token_accuracy": 0.1929756686091423,
      "num_tokens": 72287989.0,
      "step": 32990
    },
    {
      "entropy": 5.2335611343383786,
      "epoch": 3.2746129416435092,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039511794798077705,
      "loss": 4.8666,
      "mean_token_accuracy": 0.18689028471708297,
      "num_tokens": 72298718.0,
      "step": 32995
    },
    {
      "entropy": 5.297977304458618,
      "epoch": 3.2751091703056767,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003950879865243858,
      "loss": 4.8651,
      "mean_token_accuracy": 0.18669410049915314,
      "num_tokens": 72309217.0,
      "step": 33000
    },
    {
      "epoch": 3.2751091703056767,
      "eval_entropy": 5.169771597203576,
      "eval_loss": 5.093423366546631,
      "eval_mean_token_accuracy": 0.18363502243811203,
      "eval_num_tokens": 72309217.0,
      "eval_runtime": 26.6223,
      "eval_samples_per_second": 1264.692,
      "eval_steps_per_second": 158.101,
      "step": 33000
    }
  ],
  "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": 1.21392089229312e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}